of 64/64
Dominika Drobniak Hispanics as the Speakers of English - the Perceptibility and Evaluation of the Language Mistakes Made by Hispanics Praca licencjacka napisana w Instytucie Filologii Angielskiej Uniwersytetu im. Adama Mickiewicza pod kierunkiem dra Grzegorza Krynickiego

Ba Dominika Drobniak

  • View
    8

  • Download
    0

Embed Size (px)

Text of Ba Dominika Drobniak

[Your name]

PAGE

Dominika DrobniakHispanics as the Speakers of English - the Perceptibility and Evaluation of the Language Mistakes Made by HispanicsPraca licencjacka napisana

w Instytucie Filologii Angielskiej

Uniwersytetu im. Adama Mickiewicza

pod kierunkiem dra Grzegorza KrynickiegoPozna, 201012 May 2005

OWIADCZENIE

Ja, niej podpisany/a

student/ka Wydziau NeofilologiiUniwersytetu im. Adama Mickiewicza w Poznaniu

owiadczam,

e przedkadan prac dyplomow

pt.

{mozna dodac linijke z kropeczek i dlugi temat wpisac dlugopisem albo skasowac kropeczki i wpisac klawiatura; zalezy od Pani gustu}napisaem/am samodzielnie.

Oznacza to, e przy pisaniu pracy, poza niezbdnymi konsultacjami, nie korzystaem/am z pomocy innych osb, a w szczeglnoci nie zlecaem/am opracowania rozprawy lub jej istotnych czci innym osobom, ani nie odpisywaem/am tej rozprawy lub jej istotnych czci od innych osb.

Jednoczenie przyjmuj do wiadomoci, e gdyby powysze owiadczenie okazao si nieprawdziwe, decyzja o wydaniu mi dyplomu zostanie cofnita.

(miejscowo, data)(czytelny podpis)

Table of contents

3Table of contents

List of tables and figures5Introduction6Chapter 1: Language contact and linguistic variation71.1. Language contact71.2. Bilingualism71.3. Code-switching81.4. Linguistic borrowing91.5. Linguistic variation10Chapter 2: Hispanics in the United States112.1. Terminology112.2. An outline history112.3. The current status of Hispanics132.4. Places of residence152.5. Hispanics as the speakers of English162.5.1. Phonetics and phonology172.5.2. Syntax242.5.3. Semantics262.6. Summary29Chapter 3: Methodology303.1. Subjects and hypothesis303.2. Materials used in data collection313.3. Procedures in data collection313.4. Data analysis32Chapter 4: Analysis344.1. Results344.1.1. The recordings344.1.2. The questionnaire364.2. Discussion384.2.1. The recordings384.2.2. The questionnaire41Conclusion44References46

List of tables and figures13Table 1. Hispanics in the United States

16Table 2. Hispanic population in particular states

35Table 3. The most common language mistakes made by Hispanics

37Table 4. The correlation between CMPW and the scores given by the evaluators

41Table 5. The coefficient of mistakes per word in particular fields of analysis

42Table 6. The average scores and the CMPW of particular speakers

TOC \h \z \c "Figure" 14Figure 1. Educational attainment of Hispanics in the USA

17Figure 2. English proficiency of Hispanics in the USA

18Figure 3. Vowel chart of American English

18Figure 4. Vowel chart of Spanish

16Map 1. The percentage of Hispanics in particular states

IntroductionThe aim of this paper is to discuss the English language spoken by Hispanics (mainly those living in the USA). At the beginning, we will explain the most important terms related to the issue of bilingualism and language mistakes made by non-native speakers of English. The scope of the study is narrowed to Hispanic speakers only, so we will briefly characterize them and focus on the features of Hispanic English with respect to phonetics and phonology, syntax and semantics. The study aims at investigating language mistakes that Hispanics made and measuring the accuracy of evaluation of their utterances by the native speakers of English. Recordings of Hispanics speaking English and an online questionnaire will be the instruments of this research. We will measure also the correlation between Hispanics actual language mistakes and the scores (means of evaluation) given by the native speakers of American English. Furthermore, the results of this study will help investigate which of the three areas of evaluation, namely phonetics and phonology, syntax and semantics, will have the strongest influence on the general score (overall impression of a given recording). The preliminary assumption is that pronunciation mistakes will be most noticeable to the evaluators. Moreover, we will check whether it is possible that the evaluators or speakers characteristics (such as age, gender, state of residence or education) can be acknowledged as influential on the perception of a given recording. It is of utmost importance to emphasize that the utterances will be evaluated with respect to language mistakes, not the speakers proficiency in English.Chapter 1: Language contact and linguistic variation1.1. Language contactAccording to Thomason (2001: 1), language contact is the use of more than one language in the same place at the same time, however, the speakers of these languages do not necessarily have to be fluent bilinguals or multilinguals to interact verbally. This interaction can influence language in several ways, among others: code-switching, code alternation, passive familiarity, negotiation, second-language acquisition strategies, bilingual first-language acquisition, and change by deliberate decision (Thomason 2001: 129). Some of these processes will be of crucial importance for this study.1.2. BilingualismThere are numerous definitions and approaches towards bilingualism depending on different categories, scales and dichotomies (Romaine 2001: 11). According to Bloomfield (1984: 56), native-like control of two languages is the condition of bilingualism. However, linguists are not unanimous as far as the proficiency of a bilingual speaker is concerned. Grosjean (1985: 471-472) claims that bilingualism involves the ability of a speaker of one language to communicate in another in everyday life. The issue of bilingualism can be concluded following Romaine (2001: 11-12):Mackey (1968: 555) concludes that in order to study bilingualism we are forced to consider it as something entirely relative because the point at which the speaker of a second language becomes bilingual is either arbitrary or impossible to determine. He therefore considers bilingualism as simply the alternate use of two or more languages (see also Weinreich 1968: 1). Following him, I have also used the term bilingualism to include multilingualism.1.3. Code-switchingAccording to Franceschini (2002: 51), code-switching means using several languages or language varieties in the course of the conversation. Usage of this mechanism is a more or less conscious decision of a speaker. Intersentential switching and intrasentential switching are the two types of code-switching. The former occurs outside the sentence boundary and the latter takes place within one sentence (Thomason 2001: 132). In many cases, however, it is difficult to distinguish whether an item should be classified as code-switching or a borrowing. Let us consider the following example:(1) It was his idea, t sabes?

It was his idea, you know?In fact, this is a frequent insertion among Puerto Rican and Dominican bilingual speakers used in informal speech (Lipski 2005: 7). However, t sabes is problematic, as it can be treated as either intrasentential code-switching or a borrowing. There are many such ambiguous examples in different languages and linguists are divided with respect to their approaches to the distinction between code-switching and a borrowing. This question is explained by Poplack:

The classification of lone items is at the heart of a fundamental disagreement among CS researchers over (a) whether the distinction between CS and borrowing should be formally recognized in a theory of CS, (b) whether these and other manifestations of language contact can be unambiguously identified in bilingual discourse, and (c) criteria for determining whether a given item was switched or borrowed (2001: 2063).Evidently, the ambiguity of insertions like t sabes depends on the approach of a particular linguist, as differentiation between a borrowing and a single-word switch is conceptually easy but methodologically difficult (Poplack 2001: 2063). Poplack (2004: 594) also mentions the relation between code-switching and the bilingual proficiency the higher the proficiency the more frequent use of code-switches. For the purposes of this study, it has to be emphasized that code-switching and borrowing are not regarded as language mistakes. They will be mentioned in the analysis, as they might have an influence on the respondents perception of the speakers proficiency, but they will not be counted as mistakes.Apart from code-switching, bilingual speakers, especially immigrants, use code alternation. It also involves the use of at least two languages, however, what distinguishes it from code-switching is that code alternation cannot be found within the same utterance or conversation. Code alternation concerns bilinguals who use one of their languages in one set of environments and the other language in a completely different set of environments (Thomason 2001: 136), e.g. Spanish at home and English at school. While code-switching may be applicable in this study, code alternation is not because recordings present single person in a single situation.1.4. Linguistic borrowingAccording to Haugen (1950: 212), a borrowing is the attempted reproduction in one language of patterns previously found in another. As a factor that contributes to the existence of borrowings, Field (2002: 5) gives frequency. The more frequent an item in the source language, the more likely it is to be borrowed. Other reasons include the need of reference to a new object or concept and prestige of a donor language (McMahon 1994: 201-202). There are several types of borrowings (Haugen 1950) and we will describe the ones that are most important for this study. The most common is a direct loan where both form and meaning are borrowed, e.g. Spanish sandwich or ftbol. Another type of borrowing is calque, which incorporates a loan translation and a semantic loan. The first involves word-by-word translation of a foreign expression into the native language, e.g Spanish rascacielos skyscraper. The latter pertains to the extension of meaning of a native word so that it accommodates the meaning from a borrowing language, e.g. Spanish ratn mouse (the original meaning was extended to a computer mouse under the influence of English language). 1.5. Linguistic variation

The concept of linguistic variation is inseparably related to dialects including the ethnic ones. Hispanic American English (HAE) has become a dialect which, according to Ellis (1999: 147), is variation in grammar and vocabulary in addition to sound variations. However, it should be emphasized that there are many sociolinguistic factors that determine ones speech, e.g. the speakers purpose in communication, the relationship between speaker and hearer, the production circumstances and various demographic affiliations that a speaker can have (Reppen et al. 2002: VII). In case of HAE, the most important factors contributing to the emergence of this dialect was the origin of its speakers. For the purpose of this study, we will focus on three of the variations, namely phonological, syntactic and semantic. Phonological variation, also referred to as accent, can signal important information about aspects of speakers social identity about such things as class, age, ethnicity and gender (Eckert 1988: 64). In fact, the most prominent and distinctive features of HAE are the phonological ones (Kvecses 2000: 91). However, syntactic dissimilarities between HAE and Standard American English (SAE) together with differences in lexicon allow to treat them as different dialects. Finally, lexical and morphological variations are also common among Hispanics and although they may have their roots in contact with African American English (Wolfram and Schilling-Estes 1998: 182), they have become a distinctive feature of HAE as well. To conclude, as has been mentioned before, the most important mechanisms for this study include language contact, code-switching, code-alternation and borrowing. The notion of linguistic variations, divided into phonological, syntactic and lexical ones, with respect to HAE will be explained later in this paper.Chapter 2: Hispanics in the United States

2.1. TerminologyA dictionary definition for Hispanic reads: someone who comes from a country where Spanish or Portuguese are spoken, especially ones in Latin America (Mayor 2009: 831) whereas, Latino is a man in the US whose family comes from Latin America (Mayor 2009: 983). Even though, as Ryskamp (1997: 2) notes, the definition of Hispanic makes equal all the Hispanic people living in Europe, Latin America and the United States, as yet, there is no better terminology for such a distinction than calling these people by their nationalities. Notwithstanding, in this paper, we will follow the terminology proposed by federal government of the United States described in The American Community Hispanic 2004 report:

The federal government defines Hispanic or Latino as a person of Mexican, Puerto Rican, Cuban, South or Central American, or other Spanish culture or origin regardless of race. Thus, Hispanics may be any race. The terms Hispanic and Latino are used by the U.S. Census Bureau; hereinafter in this report, the term Hispanic is used to refer to all individuals who reported they were Hispanic or Latino.2.2. An outline history

The presence of Hispanics in the territories of present day United States of America can be divided into three stages, namely, the Spanish era, the Mexican era and the United States era.

The first Spanish explorers came to America at the beginning of the sixteenth century and in 1565 founded the first city on this continent, St. Augustine, Florida. The desire for exploration of new lands and establishment of New Spain led Spaniards through the Southwest of the continent triggering the creation of numerous cities, e.g. Santa Fe, long before the settlement of British pilgrims (Weaver 1994: 43-44). At the same time Spain focused on expansion in Mexico, Peru and the Philippines. Southern states, including Arizona, New Mexico and California were colonized by the end of the eighteenth century (Weaver 1994: 44). Two islands, that are native lands of many Latinos living in United States today, namely Cuba and Puerto Rico, were under the rule of Spain for almost 400 years. During that time, Spanish culture permeated into many aspects of life: Spanish language, the Catholic religion, a plantation, a monocrop economy, Spanish social structure and other institutions (Weaver 1994: 45).The situation of the empire of Spain begun to worsen when in 1821 Mexico gained independence. It was synonymous with the end of European rule in California and the immigration of American settlers to this territory. Additionally, in the 1830s Mexican government allowed Americans to live in Texas and get citizenship, and they soon managed to outnumber the Tejanos (Weaver 1994: 45). The so-called Mexican era was relatively short and ended when the war of 1846 broke out, resulting in the American invasion and Mexico City being taken over in 1847.As a result of the Mexican-American War, the Treaty of Guadelupe-Hidalgo in 1848 and the Gadsden Treaty in 1853, the vast territories once under the rule of Spain and Mexico were partially given and partially sold to the United States. The new American lands were eventually divided into the present states of California, Arizona, New Mexico, Colorado, Kansas, Oklahoma, Nevada, Utah and Wyoming (Weaver 1994: 45). The majority of Hispanics chose to stay and become American citizens, as the government of the United Stated promised them full rights to their lands. However, these promises were never fulfilled and as Weaver (1994: 46) writes in his paper:

Hispanic population became a maligned minority, objects of discrimination, land confiscation and a people with second-class citizenship. The migration of Mexicans for the remainder of the nineteenth century was modest, but it became increasingly larger during the twentieth century. It has always been tied to the periodic economic needs of Mexican workers in Mexico and opportunities in the United States. The situation of Californios (Spanish-speaking residents of California) got even worse during the California Gold Rush (1848-1855). At the beginning, Mexican and Sonoran miners revolutionized mining techniques, which soon brought great benefits to the Californian and Arizonian gold fields (Mendoza and Torres 1994: 74). They were trying to maintain Hispanic traditions and social life, however they soon became overwhelmed by large numbers of settlers who were moving from one place to another in chase of gold. Some American miners, unable to communicate with Spanish-speakers, tried to assimilate them, simultaneously provoking acts of violence (Hill 1999: 1). However, despite discrimination, civil rights violation and language barriers Hispanics managed to maintain their culture and traditions. 2.3. The current status of HispanicsAccording to the data of the U.S. Census Bureau 2008 American Community Survey, there are almost 46.9 million people in the USA who declared to be Hispanic, which constitutes over 15.4% of this countrys population. Over 65.5% of these people are Mexican. Significantly smaller percentage of Hispanics constitute Puerto Ricans, Cubans, Dominicans and people from South America. Below, there is a simplified table presenting the demographics of Hispanics from different countries.

Table 1. Hispanics in the United StatesOriginEstimatePercentage of the US population

Hispanic46,891,45615.4

Mexican30,738,55910.1

Puerto Rican4,216,5331.4

Cuban1,617,0100.5

Dominican (Dominican Republic)1,327,6850.4

Central American3,869,4261.3

South American2,732,4020.9

Other Hispanic2,389,8410.8

The biggest rise in the population of Hispanics in the United States was noted in the 1990s, however it is still growing. According to the newest predictions, the Hispanic population will be increasing to reach 102.6 million in 2050 (Bergman 2004).

According to Current Population Survey 2006, median age of Hispanic US inhabitants was 27.1 among men and 27.7 among women. These numbers are quite low in comparison with an average age of the total of residents of the United States, which is almost 9 years older. This is mainly because among Hispanic immigrants there are many young people who move to the USA with the will to improve their lives and start their families there. In 2004, about 1 in 3 Hispanics was under 18, when compared with 1 in 5 among non-Hispanic whites. What is also related to the above age statistics is the household size. Hispanics live in bigger families than people of other origins in the United States. Over 20% of Hispanic households comprise of four members and as much as 22.5% of five people and more (Current Population Survey 2006).The statistics concerning educational attainment of Hispanics show that for some reasons (probably mainly economic) they are less likely to get a BA/MA degree or even graduate from high school. As it is presented in the graph below, almost half of Hispanics in the USA get a high school degree, however almost 25% have less than 9th grade (Current Population Survey 2006).

Figure 1. Educational attainment of Hispanics in the USAHispanics, like many ethnic minorities, have smaller incomes, partly because of their lower educational attainment. According to Current Population Survey 2006, most of male Hispanics work in construction and maintenance, production and transportation or services. Women mostly occupy jobs in sales and services. An average earnings of a full-time year-round Hispanic worker in 2005 were $25,491, while the total median income was over $11,000 higher. However, even more alarming are the statistics concerning the number of Hispanics living below the poverty line, namely about 22%, which is 9% more than a total for the USA. The same study (Current Population Survey 2006) shows that almost 1 in 3 Hispanic children lives below the poverty line.

Very often, their social status in the United States is, therefore, determined by their economic status. Hispanics live in bigger families, earn less and children are forced to start working instead of going to college or university. It has its effects in low level of educational attainment and the fact that only 17.7% of Hispanics work as professionals, compared with twice as much of all the inhabitants of the United States. 2.4. Places of residenceThe latest immigration of Hispanic population to the United States began in the 1970s. They started to settle in the Southern States (mainly California and Texas) and in New York (Weaver 1994: 17). These trends seemed to continue through the next decades up to now, however, as can been seen in the Map 1 below, Hispanics managed to populate also Central and Western States. According to the data of the U.S. Census Bureau Population Estimates 2008 placed in the Table 2 below, in 2008, the biggest Latino population could be found in California, namely almost 13.5 million of Hispanics. It means that every third inhabitant of California has Hispanic origins. The second state densely populated by Latinos was Texas, the territory which is inhabited by over 8.8 million of Hispanics. Other huge populations of Hispanics could be found in Florida and New York.

As the Map 1 presents, there are four states whose percentage of Hispanic population constitutes over 30% of their overall population, namely California, Arizona, New Mexico and Texas. New Mexico is the state of the highest percentage of Hispanics 44.9%. Besides, Latinos constitute a significant part of the population of Nevada, Colorado, Florida, New York, New Jersey and Illinois.Table 2. Hispanic population in particular statesStateEstimatePercentage of the state population

United States46,943,61315.4

California13,457,39736.6

Texas8,870,47536.5

Florida3,845,06921.0

New York3,250,03816.7

Illinois1,967,12115.2

Arizona1,955,63030.1

New Jersey1,418,54516.3

2.5. Hispanics as the speakers of EnglishAs the statistics indicate (2006 American Community Survey), over 60% of Hispanics living in the United States speak English very well and almost 15% well. Worse abilities of speaking English present 15% (not well) and over 9% of Hispanics is claimed not to speak any English.

Due to the fact that Mexicans constitute the biggest percentage of all the US residents with Hispanic background, the most popular and homogeneous Hispanic American English variety is Chicano English (Kvecses 2000: 91). Beside this, for the purposes of this study, we would also touch upon the question of the characteristics of English spoken by non-native speakers with Hispanic roots. The sections below divide these features on phonetics and phonology, syntax and semantics which are simultaneously the fields of the research described in the following chapters.2.5.1. Phonetics and phonologyThe most distinctive features of Chicano English (CE) and probably the English of non-native speakers with Hispanic origins (HNNE) are in the field of phonetics and phonology. The deviations occur in the realizations of particular sounds, stress patterns and intonation, however, it must be emphasized that very often the differences are arbitrary and depend on an individual speaker (Fought 2003: 62). Taking into consideration the fact that the majority of the characteristics of CE are found also in HNNE which plays a more important role in this study, we will focus particularly on this variety.Figure 3. Vowel chart of American EnglishThe first type of differences between HNNE and Standard American English (SAE) are the realizations of vowels. As Kvecses (2000: 92) writes: SAE has eleven stressed vowel phonemes: /i/, //, /e/, //, //, /u/, //, /o/, //, //, // (Figure 3), whereas Spanish only five /i/, /e/, /u/, /o/, /a/ (Figure 4).

Figure 4. Vowel chart of SpanishSuch differences in vowel charts of these two languages cause the substitution of a few English sounds by one Spanish sound that seems to be the closest (Kvecses 2000: 92). Besides, the Spanish vowel system does not include some features that are basic to the American one, such as the distinction between lax and tense or long and short vowels (Santa Ana and Bayley 2004: 418). What is more, Spanish phonology lacks a set of r-colored allophones of long vowels (2004: 418). As a result of these differences, the speakers of CE and HNNE may not distinguish between: /i/ and // - Many of the speakers of CE substitute // for /i/ in ing morpheme, as in the example (2), but they use // correctly in other contexts (Fought 2003: 81). As opposed to them, the users of HNNE who do not have a vowel // in Spanish, encounter serious problems with this distinction, producing a majority of phonological // tokens as /i/ (Fought 2003: 81). As it is shown in (3), the lack of differentiation between these two phonemes leads to pronouncing distinct words with the same vowel /i/ (Kvecses 2000: 93).(2) movingSAE: /muv/ CE: /muvi/(3) sitSAE: /st/HNNE: /sit/seatSAE: /sit/HNNE: /sit/ // and /e/ - This distinction seems not to be problematic for the users of CE, however, the speakers of HNNE use // and /e/ interchangeably for phonemic // (Fought 2003: 82), as can be seen in (4). (4) handSAE: /hnd/HNNE: /hend/habitSAE: /hbt/HNNE: /hbt/ (Fought 2003: 82) // and // - Fought (2003: 82) notices that, once again, the speakers HNNE encounter some difficulties with distinguishing between these two vowels, inserting // in the place of //, whereas, in CE these two are different phonemes (5). However, this substitution is arbitrary and the same speakers insert // in the right places, as in the example (6) (2003: 83). Finally, the reverse substitution was also sometimes found; some tokens of // are realized as // (7) (2003: 83).

(5) aloneSAE: /lon/HNNE: /lon/(6) someHNNE: /sm/(7) calmSAE: /cm/HNNE: /cm/ diphthong and /e/ - Occasionally, the speakers of Hispanic-English varieties (including CE) and HNNE use /e/ instead of a diphthong, as in the example below (Wolfram and Schilling-Estes 1998: 114).

(8) lakeSAE: /leik/HNNE: /lek/

The second set of differences between HNNE and SAE that will be described includes various realizations of consonants. Although the inventories of consonants in English and Spanish are not as distinct as in the case of vowels, the speakers of HNNE encounter problems with the distinction between the following consonants: stops /t/ /d/ and interdental fricatives // // - Frequently, the speakers of HNNE insert apico-dental stops ([t] [d]) in the place of interdental fricatives (// //), as in (9) (Fought 2003: 67). Obviously, Hispanics are not the only non-native speakers of English who use this substitution, since it is very common among people of other nationalities (2003: 68). It is also a feature which occurs in CE, but only word-initially (Kvecses 2000: 92).(9) somethingSAE: /sm/HNNE: /smtin/thenSAE: /en/HNNE: /den/ (Fought 2003: 68) /t/ and // - This substitution is reciprocal, meaning that the speakers of HNNE substitute /t/ for // and the other way round (10). As far as the CE is concerned, Fought (2003: 82) indicates that all the native Chicano English speakers in my sample has a clear contrast between /t/ for //. However, she noticed this feature in the speech of the older generation of CE speakers. Kvecses (2000: 92) observes that it can be hypercorrection what causes the process of inserting // in words like chicken, since Spanish does not have this sound.(10) shopSAE: /p/HNNE: /tp/

chiefSAE: /tif/HNNE: /if/ /v/ and /b/ - In the speech of CE and HNNE a bilabial stop /b/ may be a substitute for a labio-dental fricative /v/ sound (11) (Kvecses 2000: 92). This may result from the fact that in Spanish v letter is pronounced as a bilabial plosive /b/, as it is shown in the example (12).(11) visit SAE: /vzt/HNNE: /bizit/(12) voy I goSpanish: /boi/ glottal fricative /h/ and velar fricative /x/ - This process occurs exclusively in HNNE and it consists in the substitution of /x/ (a velar fricative sound that appears in Spanish) for /h/, as in the example (13) below (Fought 2003: 83).(13) handSAE: /hnd/HNNE: /xnd/Apart from various substitutions of vowels and consonants, the speakers of CE and HNNE involve into their pronunciation different kind of processes, including:

consonant cluster reduction In fact, this process occurs in many American dialects, however, it is significantly more frequent among the speakers of CE and HNNE, especially in case of /t/ or /d/ deletion (14) (Fought 2003: 68) and besides, Hispanics tend to reduce clusters to a greater extent than many other dialects (Santa Ana and Bayley 2004: 425). Furthermore, there is a process called l-vocalization, which involves the deletion of /l/ sound, when it occurs in a consonant cluster (15) (2004: 425). It may even happen, very rarely however, that a speaker omits an entire cluster (16) (Fought 2003: 69). Kvecses (2000: 92) noted also that sometimes an entire suffix may be deleted, as is the case with the past time morpheme /d/ and the third person suffix /s/ may also be left out (17).(14) restSAE: /rest/HNNE: /res/

(15) coldSAE: /kold/HNNE: /kod/(16) hardwareSAE: /hrdw/HNNE: /hw/ (Fought 2003: 68)(17) savedSAE: /sevd/HNNE: /sev/ loss of final alveolar plosives This process concerns the production of consonants which occur word-finally and are not a part of a consonant cluster (Fought 2003: 69). Usually, it involves plosives which are unreleased in SAE in words such as:

(18) stopSAE: [stp] HNNE: [st]

metSAE: [met]HNNE: [me] no flaps Even though the Spanish consonant inventory includes the flap sound //, the speakers of HNNE often do not use it in English. That is because in Spanish the letter pronounced as // is r, as in the example (19), whereas in American English the realization of /t/ and /d/ as an alveolar flap is very common (Giegerich 1992: 226). In effect, as Fought (2003: 83) observes, HNNE speakers follow this rule only variably and sometimes do not have a flap, as in the following example (20).(19) caroexpensive

Spanish: /kao/(20) personalitySAE: /psnli/

HNNE: /psnlti/ (Fought 2003: 83) epenthetic - As Fought indicates (2003: 83-84), this is a stereotypical feature of non-native English speakers whose first language is Spanish and it involves the insertion of epenthetic // before a consonant cluster beginning with /s/ (21). The reason for such a process is that in Spanish consonantal blends starting with /s/ does not occur word-initially, so in HNNE // sound precedes the cluster (Helman 2004: 94). (21) speakHNNE: /spik/ final devoicing This feature seems to be related exclusively to HNNE, however Fought (2003: 84) points out that it is also noticeable among the older generation of CE speakers. Final devoicing in HNNE can occur with respect to any voiced consonant, yet the devoicing of /d/ and /z/ is most frequent (22). (22) headSAE: /hd/HNNE: /ht/

pleaseSAE: /pliz/HNNE: /plis/

fridgeSAE: /frd/HNNE: /frt/The last issue concerning phonology that will be touched upon is prosody. Although this question is one of the most distinct features of CE and HNNE, it is not well-developed yet (Fought 2003: 70). Due to the fact that some speakers use Spanish and English patterns interchangeably, prosody becomes a very elusive feature (Santa Ana and Bayley 2004: 426). For the needs of this study, we will focus on stress patterns, as they are of primary importance with respect to HNNE.

Word stress shift patterns are very frequent among the speakers of HNNE and a clear influence of Spanish can be noticed in this aspect. In words given in (23), the stress is placed on the penultimate syllable, since Spanish is a penultimate-stress language (Fought 2003: 85). However, words ending with a d letter are stressed on the last syllable, due to the rule transfer from Spanish (24) (2003: 85).(23) `technique

`show up

(24) terri`fied

rea`lized

Sentence stress in HNNE is also affected by Spanish and very often it is shifted to the penultimate word or syllable (Fought 2003: 85). It has to be emphasized that the examples in (25) are not ___________.

(25) Children are that way.

but it aint alright for me to talk to my homegirls? (Fought 2003: 85)In this section, the most significant phonological variations of CE and, more importantly, of HNNE have been described. The examples above seem to be noteworthy effects of language contact and bilingualism, however, some of the aspects still need to be examined more thoroughly (e.g. the issue of prosody).2.5.2. SyntaxSyntax is the next, after phonetic and phonology, field of descriptive grammar full of discrepancies when CE and HNNE are concerned. In this section, we will describe the main syntactic features of these two varieties of English with a special emphasis put on these that are crucial for this study.

3rd person singular forms Spanish is a language full of irregularities of forms, mainly with respect to verbs, and compared to English it seems to have much more complicated morphological patterns. However, inconsistency in using 3rd person singular forms and the lack of the distinction between was/were are very frequent among Hispanics speaking English (Fought 2003: 94). As can be seen in (26), the speakers of CE and HNNE use plural forms of verbs in place of singular ones.(26) He dont want me to end up like my sister.

[If] somebody else just come and take your life, you know.

Everybody knew the Cowboys was gonna win again. (2003: 94)

past tense forms CE and HNNE speakers face some problems with matching an adequate form of a verb to the tense they are using. As a result, very frequently, they create Present/Past Perfect by combining a verb in the past tense form with an auxiliary verb, as in (27). They also ignore some irregularities of verbs and attach ed ending to infinitives of verbs which are not regular (28) (Fought 2003: 94).

(27) I havent wrote in a long time.

() I had came out the hospital before I got shot.

(2003: 94)

(28) Those were the um- most people that I hanged around with. (2003: 94)

pronoun forms This is the next grammar issue which concerns regularization and proves to be problematic for the speakers of HNNE and CE. Fought (2003: 94-95) observes that Hispanics frequently use non-standard reflexive forms consisting of a possessive determiner and a suffix -self/-selves (29).

(29) [They] have to start supporting theirselves at early ages.

() hes a guy, he could take care of hisself.

double negation The origin of this feature is unclear, however it may have its roots in Spanish. Designated by Fought (2003: 97) as negative concord, it involves the usage of two negatives within one sentence (Kvecses 2000: 93), therefore, creating a structure which is grammatically incorrect in English, but fully acceptable in Spanish (30).(30) I dont have no car.

No tengo ningn coche.

embedded questions The structure of embedded questions in English does not involve the inversion of a subject-auxiliary. However, in embedded wh- questions in Spanish, the subject and verb or auxiliary would normally be inverted (Fought 2003: 98) (31). This difference in sentence structure is supposedly the reason why the speakers of CE and HNNE use inversion in embedded questions, as can be seen in the example (32). (31) He told us what it was.Les dijo qu era esto. what was it.

(32) He told us what was it.

modal would Fought (2003: 99) observes that Hispanics use would differently in if- clauses than the speakers of SAE. They insert would have in the clause following if, as in (33), whereas in SAE it is typical to construct this clause using pluperfect tense. In Spanish, this form of a verb is allowed to follow si (the equivalent of if) (34). Therefore, it is supposed that the occurrence of this feature among the speakers of CE and HNNE may have its sources in Spanish.

(33) If Thurman Thomas wouldnt have dropped those fumbles, then the Bills would have won.

(34) Si Thurman Thomas no hubiera perdido la pelota, los Bills hubieran Ganado. wouldnt have

would have

Fought (2003: 99)

prepositions The use of prepositions by the speakers of CE and HNNE is deeply influenced by Spanish and as a result, this feature can be noticed more often among non-native speakers of English. They substitute a proper preposition with the literal translation of the Spanish one. In this way, they insert on in the place of in, because of Spanish en, which means both in and on (35) (Fought 2003: 100-101). Similarly, for is used in the meaning of so that and it is probably derived from Spanish para que, whose literal translation would be for that (36) (2003: 101). This characteristics, however, may vary among the speakers of HNNE and depend on their proficiency of English.

(35) Were really supposed to get out of here on June. (36) For my mum can understand. For she wont feel guilty. (Fought 2003: 100)In this section, the most important syntactic variations were presented. For the purposes of this study, we discussed only those which are probably tied to Spanish, as well as those most frequent among the speakers of HNNE. 2.5.3. Semantics

In this section, we will describe lexical features of CE and HNNE with a special emphasis put on these that can be related to Spanish language. tell In the speech of CE and HNNE tell is often used as ask. As Fought (2003: 105) explains, this can come from an ambiguous use of Spanish decir (meaning say/tell) in embedded questions. The speakers of HNNE when translating them, may insert tell for ask, as in (37). It results from the fact that in Spanish it is correct to use decir in embedded questions instead of preguntar (meaning ask) (38). Fought (2003: 105) supposes that this process could originate in embedded questions in English and then, extend to other constructions.(37) If I tell her to jump up, shell tell me how high.

She was telling my aunt to tell them, you know, what, I mean, whats the reason?

Fought (2003: 104)(38) I asked him if he wanted a car.

Le dije que si quera el coche.

(I) told barely The speakers of CE and HNNE use barely not only in its primary meaning being that something happens immediately after a previous action (Mayor 2009: 118), but also as an equivalent to recently (39). Although, as Fought (2003: 106) observes, this usage is also frequent among the speakers of other non-standard dialects of English, some Spanish impact can also be noticed. The Spanish equivalent word to barely, namely apenas, has two primary meanings. The first overlaps with the English one given above, whereas the second denotes that something happened recently (Fought 2003: 106). In this way, there are two theories of the origins of such an extension of meaning either it was adopted from other English dialects or it stems from the Spanish meaning of apenas.(39) These [treadmills] were expensive when they barely came out.

(Fought 2003: 106)

brothers The discrepancy between the standard meaning denoted by the plural of brother and the meaning in which it is used by Hispanics results from interference with Spanish. The speakers of CE and HNNE use brothers when they talk about the siblings regardless of gender (40). According to Fought (2003: 106), this usage is influenced by a Spanish masculine plural noun hermanos, used for males or for a group of males and females (to refer to females alone they use a feminine form hermanas). Taking into consideration the gender, hermanos is an equivalent to English brothers, however semantically, it has a broader meaning extended to both sexes. (40) To my brothers I usually talk in English.

(the speaker has four sisters and one brother) Fought (2003: 106)

borrow/lend The differentiation between these two words causes problems for many non-native users of English, not only Hispanics, because borrow and lend are marked by the person the action is directed to. Spanish, like many other languages, lacks this distinction, so that the speakers of CE and HNNE use borrow in the place of lend (41) (Kvecses 2000: 93).(41) Can you borrow me you bike?

discourse marker ey This discourse marker is surely borrowed from Spanish, and it can be compared to English yeah. Fought (2003: 106-107) claims that ey should not be perceived as code-switching, since the speakers who use it does not involve any other switches in their speech. (42) If a girls pretty you know and she feels the same for me, ey, I got it right there.

Fought (2003: 106)

The majority of examples of semantic variation given above originate or are borrowed from Spanish. However, as has been mentioned earlier in this chapter, for the purposes of this study we focus our attention on HNNE which is deeply influenced by Spanish. On the other hand, Chicano English also involves the usage of words discussed above, but it cannot be accused of being a dialect created only on the basis of the mistakes made by Hispanics (Fought 2003: 109).2.6. Summary

In this chapter, we characterized Hispanics living in the United States as an ethnic group which has its history and culture. We presented their current status in the USA with a special emphasis on their origins and places of residence. Ultimately, the main focus of this chapter was to describe the varieties of English used by Chicanos and, more importantly for this study, the variety used by all Hispanic non-native speakers of English. This was done with respect to phonetics and phonology, syntax and semantics. We tried to find explanations and influences of Spanish on all these processes and deviations. It has to be emphasised, however, that language change and variation cannot be easily measured and explained, especially with respect to varieties other than CE (Kvecses 2000: 91). For this reason, the above examples cannot be undeniably interpreted by an impact of Spanish, even if for the majority of them, this explanation seems to be accurate and justified. Chapter 3: Methodology3.1. Subjects and hypothesisIn this chapter, we will describe the questionnaire and its respondents as well as the process of data collection.The primary aim of the study was measuring the accuracy achieved by native speakers of American English of the evaluation of language mistakes made by Hispanics. This was done with respect to phonetics and phonology, syntax, and semantics, as these areas cover the most frequent variations in HNNE. The assumption behind such a division was that the most noticeable feature of speech would be phonetics and phonology. For this reason, it was presumed that numerous mistakes in this area will correlate closely to a lower general score for a given recording. Besides, we investigated the influence of the characteristics of the Hispanic speakers and the American respondents on their assessment.Among the secondary aims of this study was the analysis of chosen recordings of Hispanics speaking English also with respect to phonetics and phonology, syntax, and semantics. We also looked for the most common mistakes and investigated in which of the fields of analysis Hispanics made most mistakes. Ultimately, the speakers characteristics (gender, origin, time spent in the USA and the state of residence) were taken into account in order to measure whether they influence the amount of mistakes.3.2. Materials used in data collectionThe data for this study were compiled with the use of a questionnaire. First, ten excerpts from interviews with famous Hispanics speaking English were collected with the use of YouTube (http://www.youtube.com). They became the core part of the questionnaire and the subject of evaluation. Each recording was not longer than one minute in length and contained from two to four sentences. The identity of the speakers in the recordings was not revealed until the end of the questionnaire, so that personal likings would not have any influence on the evaluation. It is, however, possible that some of the voices were recognized anyway. What is more, the proficiency of English of the speakers was varied as well as the time they spent in the USA. The questionnaire was available online for the native speakers of American English who were the target group of this study. Each recording was to be listened to only once and evaluated with respect to pronunciation, sentence structure and vocabulary with the use of 1-10 scale (where 1 is extremely bad and 10 is excellent). The respondents gave also a general score and could leave some additional comments to each sample. They were told to pay special attention to language mistakes, not the variety of vocabulary or grammatical structures. At the beginning of the questionnaire, they provided basic personal data, namely gender, age, state of residence and completed education. It was assumed that some of this information might enable some additional divisions among the respondents. The overall number of respondents was 83.3.3. Procedures in data collectionThe first stage in data collection was the compilation ten recordings of Hispanics speaking English. During the selection process special attention was paid to the variety of speakers proficiency in English and the spontaneity of production. The latter was of exceptional importance since the point was analyzing and evaluating unprepared, natural speech as a determinant of a speakers proficiency in English. Then, the recordings were compiled in an online questionnaire addressed to native speakers of American English. The respondents were not selected, however, the majority of them were linguists or people interested in linguistics. This was due to the fact that the questionnaire was distributed also via the Linguist List (http://www. linguistlist.org). This group may comprise of about 73% of all the respondents. 3.4. Data analysisAt the outset, all the recordings were transcribed and analyzed with respect to semantic and syntactic mistakes. Subsequently, with the use of GATE as an annotation tool, all the phonetic and phonological mistakes were found and marked. Summed up amounts of language mistakes were compiled in a table and the coefficients of mistakes per word (CMPW) were calculated (each recording had four coefficients i.e. for phonetics and phonology, syntax, semantics and the overall one). Such a coefficient was necessary since the recordings contained various amounts of words. It proved to be essential further in this study, namely in the analysis of the questionnaire results. We gathered information about the speakers (such as origin, time spent in the USA and the state of residence) with the ultimate goal of measuring their influence on the amount of language mistakes in the particular recordings.The second part of data analysis process was compilation of the questionnaires results and investigating which of the evaluated components correlate most precisely with the general score. First, all the scores for each field of analysis were averaged without retaining the division on recordings. Then, we calculated the correlation between each component and the general score. At this stage, we did not check the accuracy of evaluation, only the influence of particular scores on the overall one. We used Pearsons correlation coefficient which is defined by the following scale: from +1 (complete positive correlation), through 0 (no correlation), to -1 (complete negative correlation). The next stage was measuring the accuracy of evaluation, meaning the correlation between the average scores for each component and the CMPW. This was done without averaging the results of all the recordings, since the scores and the amounts of mistakes varied significantly among the Hispanic speakers. According to the assumption, this correlation value should be close to -1, meaning that the more mistakes in a given utterance, the lower the score. The third issue we wanted to investigate was the correspondence between the mean of scores given for phonetics and phonology, syntax and semantics, and the general note. This was aimed at measuring if the respondents tend to average the three scores when giving the general one. At the end, we calculated the correlation between the personal characteristics of the respondents and the scores they gave. This was done with respect to gender, age, level of completed education and state of residence. In case of the latter, we asked a question if the amount of Hispanics in particular states has any influence on the notes given by the evaluators. Chapter 4: Analysis4.1. ResultsThe description of findings will begin with the analysis of the corpus composed for the aims of this study. Ten recordings of Hispanics speaking English were examined to find the most common language mistakes the speakers make in spontaneous speech in English. Moreover, we will answer the question in which of the fields of analysis (phonetics and phonology, syntax or semantics) the speakers made most mistakes. Then, we will present how the speakers characteristics influence the amount of language mistakes in their utterances.

In the next part of the resultative section, we will present the results of the questionnaire concerning the accuracy of evaluation of the 10 recordings by native speakers of American English. Then, we will answer the questions: which of the field of research seems to correlate most strongly with the general score and finally, if the characteristics of the evaluators and the speakers had any influence on the scores.

4.1.1. The recordingsThe study shows that it is possible to indicate the most common language mistakes made by Hispanic (Table 3). All the mistakes were divided into three categories, namely phonetics and phonology, syntax and semantics. Table 3. The most common language mistakes made by Hispanics

FeatureNumber of

occurrencesPercentage of all the mistakes in a given field of analysis

PHONETICS AND PHONOLOGY

substitution of /i/ for //5520.1

devoicing4416.1

substitution of /d/ for //3111.3

substitution of /a/ for // or //279.9

substitution of /o/ for //, //, // or //269.5

no sound (the sound is omitted)238.4

insertion of an alveolar /r/155.5

substitution of /n/ for //93.3

no diphthong93.3

substitution of /e/ for // or //72.6

substitution of //for /i/62.2

sound insertion (an unnecessary sound is added)62.2

substitution of // for /z/ or //31.1

voicing31.1

no flap20.7

substitution of /f/ for //20.7

substitution of // for //10.4

substitution of /t/ for //10.4

substitution of // for /n/10.4

wrong sound (not even close)10.4

stress shift10.4

substitution of /t/ for //10.4

SYNTAX

wrong tense indicated by a wrong verb form618.2

wrong verb form515.2

wrong article39.1

unnecessary article 39.1

no verb39.1

no article26.1

wrong tense26.1

unclear structure26.1

wrong form of a structure26.1

wrong preposition26.1

double verb13.0

wrong structure of an embedded question13.0

no preposition13.0

SEMANTICS

wrong word450.0

illogical use of a word or structure223.3

wrong collocation116.7

code switching (not regarded as a mistake)3-

In the first section, it occurred that Hispanics had serious problems with the i/ contrast since the substitution of /i/ for // in words like think, it, this and in the ing morpheme comprises 20.2% of all the pronunciation mistakes. The next most common mistakes were devoicing (16.1%) and the substitution of /d/ for // (11.3%). As far as syntactic mistakes are concerned, most often Hispanics used wrong tense indicated by a wrong verb form, e.g. arrive for arrived or speak for spoke. In the field of semantics the speakers did not make many mistakes of which the most common one was the use of a wrong word. It constitutes as much as 50% of all the semantic mistakes, among which there were also the use of a wrong collocation and illogical structure. The occurrences of code switching were also counted, however as it has been mentioned earlier in this paper, these were not regarded as language mistakes, but as a separate mechanism. The next question we asked was in which of the analyzed language areas Hispanics made most mistakes. With the use of the CMPW, it was proved that phonetics and phonology is the area of linguistics mostly affected by mistakes (the coefficient 0.41). Hispanics made much less syntactic mistakes (0.05) and even less the semantic ones (0.01).As it has been mentioned above, we also checked if there is any correlation between the characteristics of the Hispanic speakers of English and the number of language mistakes in their speech. The characteristics that were taken into account are: gender, the country of origin, time spent in the USA and the state of residence. It was proved that all these characteristics are insignificant. 4.1.2. The questionnaire

The main aim of this study was measuring how accurately native speakers of American English evaluate spontaneous speech and language mistakes made by Hispanics speaking English (Table 4). The results show that the highest correlation between an average score given for each recording and the CMPW occurred in the area of syntax (ab. -0.77). This will prove that the syntactic mistakes were the easiest to notice. The second highest correlation was in the field of phonetics and phonology (ab. -0.52) and the smallest occurred with reference to semantics (ab. -0.2). The correlation between an average general score for each recording and the CMPW is quite high (ab. -0.53). All the coefficients were negative which means that the more mistakes were made, the lower the score was given.

Table 4. The correlation between CMPW and the scores given by the evaluatorsField of analysisThe correlation between CMPW and the scores given by the evaluators

Syntax-0.77

Phonetics and phonology-0.52

Semantics-0.20

General score-0.53

Additionally, we measured which recording was evaluated most/least accurately and we managed to find slight correspondence to some of the distinctive features of the recording (e.g. fluency), but rather no correlation with the characteristics of the speakers (besides the probable personal likings of the evaluators). Moreover, on the basis of the data compiled with the use of the questionnaire the relation between the general score and the scores for particular components was measured. The correlation coefficient between these variables occurs to be very high in the area of both syntax and semantics it reaches about 0.91 and a bit smaller in case of phonetics and phonology (0.88). This proves that syntactic and semantic mistakes had a little more influence on the general score (and therefore, the impression) than the pronunciation ones. However, it turned out that the mean of all scores for phonetics and phonology, syntax and semantics (6.936) is nearly perfectly equal to the mean of all the general scores (6.938). This shows that none of the three language areas has significantly stronger influence on the general score.Additionally, the correlation between the characteristics of the respondents and the scores they gave was measured. The following set of information about the evaluators was taken into account: gender, age, completed level of education and state of residence (specifically the number of Hispanics living in a given state). All these characteristics proved to be insignificant, since the correlation coefficients reached from -0.16 to 0.25.At the end, we measured which of the recordings (therefore, Hispanic speakers) was evaluated most leniently. Some explanations were inferred, however, it should be noted they may be exaggerated since we used relatively little data. 4.2. DiscussionIn this subchapter all the results will be discussed and confronted with the primary hypotheses and other studies.

4.2.1. The recordings

The most common pronunciation mistakes of these Hispanic speakers were substitution of /i/ for // (20.1% of all pronunciation mistakes), devoicing (16.1%) and substitution of /d/ for // (11.3%). As Fought (2003: 81) observed, lack of i/ contrast is one of the most noticeable features of HNNE. In the vast majority of cases, they insert /i/ instead of //. This study confirmed Foughts observations since in as much as 47.8% of cases where // should be used, the speakers inserted /i/. This lack of contrast works also in the opposite direction, however with less frequency (only 13.5%). Such a mechanism can be easily justified on the basis of the vowel charts of English and Spanish. The fact is that there is no // sound in the Spanish inventory of vowels and probably for this reason, the substitution of /i/ for // is the most common pronunciation mistake made by Hispanics. The production of /i/ as // can be therefore regarded as hypercorrection. The second most common pronunciation mistake was devoicing, mostly final (93.1%). Most frequently devoiced sound was /z/ as: a marker of plurality, derivational s and in words like is and was. According to Fought (2003: 84), this process approaches 100 percent. This study shows, however, that devoicing of final /z/ happens in about 74% of cases. It is probably caused by the absence of final voicing in Spanish and is often part of a larger process, the devoicing of word-final obstruents (MacDonald 1989: 229). Therefore, Hispanics tend to devoice final voiced consonants, most frequently /z/ and less often /d/.The study proved that Hispanics have considerable problems with // which they substitute with apico-dental stop [d] or an alveolar stop /d/. Both // and // are problematic for many non-native speakers of English as these sounds occur in few languages. Interestingly, they can be found in many varieties of Spanish and even though, Hispanics do not produce them correctly. The possible explanation for this mispronunciation may be that in Spanish sound // is the realization of letter d in a word like dedo pronounced: /deo/. Therefore, the problem here is not the inability or difficulty of pronouncing sound //, but the letter which is realized as //. In the area of syntax two major problems can be identified, namely wrong tense indicated by a wrong verb form (18.2%) and wrong form of a verb (15.2%). The first type of mistakes is probably caused by either simplification or ignorance of the rules, such as past tense or past participle forms of verbs. Wrong verb form category included such examples as the use of was with reference to the second person singular and is probably related to the reasons given above. This problem of HNNE was also mentioned by Fought (2003: 94). Among other frequent syntactic mistakes there were problems with articles (wrong article or unnecessary article). We found it unusual since Spanish also has a system of articles, which is analogical to the English one. There seems to be no logic in their mistaken use of articles, so it may simply result from speakers carelessness. Some of the features that has been described in the theoretical part of this paper appeared also in our recordings. Among them there was a wrong structure of embedded question supposedly caused by difference between Spanish and English structures of such questions (in Spanish the form of a question is retained, in English it adopts the order of an affirmative sentence).Semantics was the least troublesome field of linguistics for Hispanics and the most frequent mistake was choosing wrong word (50%). We will discuss them all since there were only three such cases. The first one, probably most striking (several questionnaires respondents included it in their comments), was using get worm as worm up. This was probably caused by the fact that a Spanish equivalent for worm up, which is calentarse, overlaps also with the meaning of English get worm. On the basis of such information, the speaker might have assumed that these English words are synonyms and mean to practice, to prepare for something. (43) get worm instead of worm up in:

Now, because I havent speak in English in a long time, you can tell, it take me a while to get warm.

The second word inappropriately chosen was for in:(44) () it was a new set maked very quickly. I mean, for problems and then I couldnt rehearse.

The more appropriate word here seems to be because of. We did not found any logical sources of this mistake in Spanish, but it can be caused by the English use of for for giving reasons, as in (45). In the sentence above, however, because of seems to be a more reasonable choice.(45) He found it increasingly difficult to read, for his eyesight was beginning to fail.

(Mayor 2009: 680)

The last word that was mistakenly chosen was illusion (46). Supposedly, the speaker wanted to say that something was unbelievable and incredible. There are, however, no reasons in Spanish or English for the use of this word here, so it can be treated as a random mistake without any underlying cause. (46) I started when I was nine years old and it was an illusion. I dream about being a singer and an actress.Taking into consideration the number of subjects examined and the length of their speeches (altogether 664 words) all reasoning behind the mistakes presented above, especially in case of syntax and semantics, may be exaggerated. As the syntactic and semantic mistakes were rare, they may prove to be characteristic of a given speaker, not generally assigned to all the speakers of HNNE.As has been mentioned before, there were also several cases of code-switching, but these were not regarded as mistakes. Code switching was used by two speakers (total number of switches was 3). These uses, however, resulted from the ignorance of a particular word or phrase in English rather than high bilingual proficiency.(47) Im the girl of the ochos. (= eyes)

(48) We are influencing the music in our country and in our decada. (= decade)

(49) I mean this kind of kiss with two people that are, you say, con la cara cubierta (= with covered face)

The next aim of this study was determining which of the fields of analysis was mostly affected by mistakes of the speakers of HNNE. The results show that most problems Hispanics have in the area of phonetics and phonology, then much less in syntax and semantics (Table 5). Generally, 0.48 mistake falls on one word, so the speakers made nearly one mistake each two words. Probably it is no surprise that pronunciation mistakes are most common, but the disparity between them and the semantic and syntactic ones occurred to be high. This may be caused by the fact that it is possible to make more than one pronunciation mistake in one word. For this reason, some additional research has been conducted and instead of counting each inappropriate sound, we counted a mispronounced word as one mistake (regardless of the number of mistaken sounds in it). The coefficient of mistakes per word amounted to 0.31.Table 5. The coefficient of mistakes per word in particular fields of analysisField of analysisThe coefficient of mistakes per word

Phonetics and phonology0.41

Syntax0.05

Semantics0.01

Overall0.48

4.2.2. The questionnaire

The first issue that will be touched upon here is the accuracy of evaluation of language mistakes made by Hispanics by the native speakers of American English. As has been mentioned earlier, we used correlation coefficient to measure the accuracy. The exact numbers were presented in Table 4. It occurred that the area of analysis which correlates best with the CMPW is syntax, meaning that the syntactic mistakes were easily detected by the respondents and the scores for syntax lowered due to these mistakes. As far as phonetics and phonology is concerned, the respondents noted the speakers mistakes in this area, but less accurately than with respect to syntax. Therefore, the hypothesis we adopted at the beginning has not been confirmed. It is not pronunciation mistake that is most noticeable, but a syntactic one. The scores given for semantics were proved to correlate poorly with the actual mistakes. This can be caused by the fact that probably many of the respondents evaluated speakers proficiency, instead of only language mistakes (as they were asked to). For this reason, they gave lower scores for semantics even if there were no semantic mistakes in a given sample.The next question we asked at the beginning was: which of the fields of research seems to have the strongest/weakest influence on the general score? It was proved that they all have nearly equal impact on the general score (the correlation coefficients for syntax and semantics - both 0.91, phonetics and phonology - 0.88). Moreover, the mean of all the scores for particular components is nearly equal to the mean of all the general scores, 6.936 and 6.938 respectively. This means that the respondents tend to treat all the three language areas equally and average these three scores while giving the general one.

At the end, we used the compiled data to check which speaker was evaluated most leniently and whom the respondents assessed most harshly. Table 6 below contains the names of the speakers, the CMPW and the mean of the scores given by the respondents. The data in this table is organized according to the average score each of the speakers received (decreasing order).Table 6. The average scores and the CMPW of particular speakersName of the speakerThe average score given by the respondentsThe coefficient of mistakes per word

Salma Hayek8.580.54

Javier Bardem7.900.14

Thala7.830.31

Penlope Cruz7.820.40

Antonio Banderas7.320.51

Julio Iglesias6.780.65

Gloria Trevi6.021.02

Pedro Almodvar5.830.62

Gloria Trevi5.710.71

Pedro Almodvar5.590.39

The results show that the respondents gave Salma Hayek the highest scores, while she made one mistake every four words. According to the CMPW she should be on the sixth place. Pedro Almodvar, whose speeches were evaluated very low, achieved the CMPW very similar to the one of Salma Hayek. However, the difference in the scores they got reached nearly 3 points. The underlying reason for that may be that either her voice was recognized by the respondents and the personal likings increased the scores or not all the mistakes were noticed due to her fluent speech. As far as Pedro Almodvar is concerned, it can be inferred that such low scores may have their sources in his manner of speaking. His utterances were not fluent, he spoke slowly and corrected himself (thus, focused a listeners attention on his mistakes). Interestingly, there were two speakers who used code-switching and it resulted that these are the ones who get the lowest scores: Pedro Almodvar and Gloria Trevi. As has been mentioned earlier, the switches they used did not result from high proficiency, but rather from the ignorance of an appropriate word in English. Supposedly, code-switching might have made a bad impression on the respondents what affected the scores.ConclusionThe primary chapters of this paper treated of the status of Hispanics in the USA in the past and today, as well as discussed the most common features of HNNE and CE. In the methodological part, we described the research whose main aims was the examination of language mistakes made by the speakers of HNNE and the measurement of the accuracy of evaluation of these mistakes by the native speakers of American English. It appeared that the differences in Spanish and English sound inventories play a significant role in the realizations of both English vowels and consonants in the speech of Hispanics. Some of the English phonemes are not present in the English sound system and even if they do, they have different realizations (MacDonald 1989: 216). This study confirmed one of the most popular language mistakes made by Hispanics speaking English are the pronunciation ones, e.g. the substitution of /i/ for //, devoicing (mostly word-finally) and the substitution of /d/ for //. However, the explanations of these discrepancies presented earlier in this chapter are only probable reasons. It is impossible to state clearly and surely that all the differences in realizations of particular sounds originate in Spanish phonology (MacDonald 1989: 216). In the area of syntax, the most frequent mistakes include inserting wrong tense and the wrong form of a verb, whereas semantics caused least and rather minor problems which do not disturb the communication. The characteristics of the speakers proved to be insignificant with regard to their mistakes. The second part of the study was devoted to the analysis of the questionnaire whose respondents included native speakers of American English. One of the worth mentioning results we achieved is that the evaluators detected most easily the syntactic mistakes and least accurately they assessed the semantic ones. Simultaneously, the hypothesis we established at the beginning was not confirmed the pronunciation mistakes were not the easiest to notice. In fact, the scores for phonetics and phonology correlate with the actual amount of mistakes in this area, however, the correlation is not as high as in the case of syntax. It should be emphasized, that this result may be affected by the fact that some of the respondents might have evaluated the proficiency of a given speaker, instead of focusing exclusively on the mistakes (as they were told to). For this reason, they gave lower scores, e.g. for semantics, even if there were no semantic mistakes. We also found that the respondents tend to average the scores for three language components while giving the general note, which proves that they all have similar influence on the overall impression. The characteristics of both the evaluators and the speakers seems not to have any significant influence on the perception of a given utterance. However, there are some underlying reasons for inaccurate evaluation of the recordings, e.g. the fluency of speech might have increased the notes and helped the speaker make better impression. Another vital factor are supposedly personal likings, since the speakers were famous people and their voiced might have been recognized by the respondents of the questionnaire. In conclusion, it should be emphasized that since correlation does not imply causation, it was indispensable to look for other factors that could have an influence on the correlating variables. Practically, it means that both the sociology of language and the sociolinguistic factors, such as the characteristics of the speakers and the evaluators, the status of Hispanics in the USA or personal likings, must have been included in the analysis of the data. References

2000 US National Census, United States Census Bureau (retrieved from: http://www.census.gov, date of access: xx.xx.2010).

2004The American Community Hispanics, United States Census Bureau (retrieved from: http://www.census.gov, date of access: xx.xx.2010).

2006 Current Population Survey, Annual Social and Economic Supplement, United States Census Bureau (retrieved from: http://www.census.gov, date of access: xx.xx.2010).

2008 American Community Survey, United States Bureau (retrieved from: http://www.census.gov, date of access: xx.xx.2010).2008 Population Estimates, United States Bureau (retrieved from: http://www.census.gov, date of access: xx.xx.2010).

Bergman, Mike. 2004. More diversity, slower growth. Washington D.C.: U.S. Census Bureau News. MacDonald, Marguerite G. 1989. The influence of Spanish phonology on the English spoken by United States Hispanics, in: Peter C. Bjarkman and Robert M. Hammond (eds.), American Spanish pronunciation: theoretical and applied perspectives. Georgetown: Georgetown University Press. Bloomfield, Leonard. 1984. Language. Chicago: The University of Chicago.

Eckhert, Penelope. 1998. Gender and sociolinguistic variation, in: Jennifer Coates (ed.), Language and gender: a reader. Malden: Blackwell Publishing Ltd., 64-75.Ellis, Donald G. 1999. From language to communication. Mahwah: Lawrence Erlbaum Associates Ltd.Field, Fredric W. 2002. Linguistic borrowing in bilingual contexts. Philadelphia: John Benjamins B.V.

Franceschini, Rita. 2002. Code-switching and the notion of code in linguistics, in: Peter Auer (ed.), Code-switching in conversation. Language, interaction and identity. London: Taylor & Francis e-Library, 51-75.Giegerich, Heinz J. 1992. English phonology: an introduction. Cambridge: Cambridge University Press.Grosjean, F. 1985. The bilingual as a competent but specific speaker-hearer, in: Journal of Multilingual and Multicultural Development 6, 467-477. Haugen, Einar. 1950. The analysis of linguistic borrowing, in: Einar Haugen, The ecology of language. Stanford: Stanford University Press, 79-109.

Helman, Lori A. 2004. Building on the sound system of Spanish: Insights from the alphabetic spellings of English-language learners, in: International Reading Association, Preparing reading professionals: a collection from the International Reading Association. Newark: International Reading Association Inc., 92-100.Hill, Mary. 1999. Gold: The California story. Los Angeles: University of California Press.

Kvecses, Zoltn. 2000. American English: an introduction. Peterborough: Broadview Press Ltd.

Lipski, John M. 2005. Code-switching or Borrowing? No s so no puedo decir, you know, in: Lotfi Sayahi and Maurice Westmoreland (eds.), Selected proceedings of the Second Workshop on Spanish Sociolinguistics. Somerville: Cascadilla Proceedings Project, 1-15.Macpherson, Ian Richard. 1975. Spanish phonology: descriptive and historical. Manchester: Manchester University Press.Mayor, Michael. 2009. Longman dictionary of contemporary English. Essex: Pearson Education Ltd.

McMahon, April M.S. 1994. Understanding language change. Cambridge: Cambridge University Press.

Mendoza, Rubn G. and Cruz C. Torres. 1994. Hispanic traditional technology and material culture in the United States, in: Thomas Weaver (ed.), Handbook of Hispanic cultures in the United States: Anthropology. Houston: Arte Pblico Press, 59-84.Poplack, Shana. 2001. Code switching: linguistic, in: Neil J. Smelser and Paul B. Baltes (eds.), International encyclopedia of the social & behavioral sciences. Oxford: Elsevier, 2062-2065.Poplack, Shana. 2004. Code-switching, in: Ulrich Ammon, Norbert Dittmar, Klaus J. Mattheier and Peter Trudgill (eds.), Soziolinguistik. An international handbook of the science of language. Berlin: Walter de Gruyter, 589-596.Reppen, Randi, Susan M. Fitzmaurice and Douglas Biber (eds.). 2002. Using corpora to explore linguistic variation. Amsterdam: John Benjamins.Romaine, Suzanne. 2001. Bilingualism. Malden: Blackwell Publishers Inc.Ryskamp, George R. 1997. Finding your Hispanic roots. Baltimore: Genealogical Publishing Co.

Santa Ana, Otto and Robert Bayley. 2004. Chicano English: phonology, in: Edgar Werner Schneider and Bernd Kortmann (eds.), A handbook of varieties of English: a multimedia reference tool, Volume 1. Berlin: Mouton de Gruyter, 417-434.Thomason, Sarah G. 2001. Language contact. Edinburgh: Edinburgh University Press.

Weaver, Thomas. 1994. Latino legacies: Crossing national and creating cultural borders, in: Thomas Weaver (ed.), Handbook of Hispanic cultures in the United States: Anthropology. Houston: Arte Pblico Press, 39-58.Weaver, Thomas. 1994. The culture of Latinos in the United States, in: Thomas Weaver (ed.), Handbook of Hispanic cultures in the United States: Anthropology. Houston: Arte Pblico Press, 15-38. Wolfram, Walt and Natalie Schilling-Estes. 1998. American English: dialects and variation. Malden: Blackwell Publishing Ltd.

Map SEQ Map \* ARABIC 1. The percentage of Hispanics in particular states

Figure SEQ Figure \* ARABIC 2. English proficiency of Hispanics in the USA

PAGE 36