Castilian Adult ESL Learners’ pronunciation of the English phonemes /v/ and /w/
Maria van de Poll S2274019
Pre-Master English Language and Linguistics Supervisor: D. Smakman
Second Reader: J. Grijzenhout Leiden University
Table of Contents TABLE OF CONTENTS ... 2 A LIST OF TABLES ... 3 PREAMBLE ... 3 1. INTRODUCTION ... 5 2. LITERATURE REVIEW ... 6
SPANISH LANGUAGE TRANSFER IN ENGLISH SPEECH PRODUCTION ... 6
ENVIRONMENTS OF /V/ AND /W/ IN ENGLISH ... 7
ENVIRONMENTS OF /V/ AND /W/ IN SPANISH ... 9
STRESS:SYLLABLE-TIMED LANGUAGE VS.STRESS-TIMED LANGUAGES ... 10
RESEARCH QUESTIONS ... 11
3. METHODOLOGY ... 11
VARIABLES: PHONOLOGICAL PATTERNS OF /V/,/W/ AND /B/ ACCORDING TO THEIR ENVIRONMENT ... 11
PARTICIPANTS:SPEAKERS ... 13
SPEECH MATERIAL:SPEAKING TASK ... 14
RESEARCH TOOL ... 15
PROCEDURE, ANALYSIS AND STATISTICAL TREATMENT ... 16
4. RESULTS ... 17
OVERALL FIRST TASK RESULTS... 17
RESULTS TASK TWO ... 19
INDIVIDUAL PROBLEMATIC PRONUNCIATION PATTERNS ... 21
5. DISCUSSION ... 22
6. CONCLUSION... 28
BIBLIOGRAPHY ... 30
APPENDIX I. SPEAKING TASK FOR PARTICIPANTS I: STORY ... 33
A list of tables
Table 3.1. Environment and words used
Table 3.2 Participants’ Language Characteristics Table 3.3. Sample of Data Analysis
Table 4.1. Category of pronunciation of target words.
Table 4.2. Words with target phonemes used by each speaker in Task 2. Table 4.3. Problematic Pronunciation within Speakers
Preamble
As a teacher of English in a language academy in Madrid (Spain), I taught English to adult Spanish learners in companies. These students were highly motivated but struggled to communicate properly with their new bosses whose first language was English. They found it difficult to sound native-like when speaking English because of their influence of Spanish in their speech. Even though they had taken many English classes over the years and their level was advanced, they still felt stuck at a low pronunciation level. It is for this reason that I focused on their speech and found that one troublesome aspect for them in English is the correct pronunciation of /v/ and /w/. Consequently, this Pre-Master Thesis focuses on exploring the problems behind the pronunciation of the mentioned phonemes. In this
preamble I want to thank the participants who made this study possible, Mr. and Ms. van de Poll for helping me collect the data and my thesis supervisor Dr. Smakman for his feedback.
Abstract
This study focuses on the way Spanish ESL speakers pronunce the English phonemes /v/ and /w/. It parts from the hypothesis that the issue starts with language interference from the speakers’ mother tongue (Ellis, 2008). In order to find whether the mispronunciation comes from the mother tongue, the two phonetic systems of English and Spanish were analysed in relation to /v/ and /w/. By doing so, the results showed some overlap between both language systems and possible language transfer. Consequently, the speech of five participants was analysed through two tasks to see whether the theory matched the practice. This research concludes with the statement that indeed there is language transfer at stake, in addition to possible overgeneralization of English rules. These results show the relevance for English teachers of ESL students to spend more time and give more attention to raising awareness of the manner of pronunciation of these phonemes. This can be done by showing the different environments the phonemes can be found and how they can be realized
1. Introduction
Despite the Spanish and English phonological systems being very similar, there are differences which are reflected in Spanish ESL learners’ English speech (Coe, 2001). When listening to a Castilian speaker talking in English, one might depict certain characteristics that set the speaker apart from a native speaker. This second language (L2) speech shows a foreign accent which a L1 (native) speaker can easily recognize (Polyanskaya, Ordin & Grazia Busa, 2017). L2 speakers’ speech differs from native speakers at more than one level: not only do they make various syntactic and lexical errors, but most prominently they pronounce words and utterances differently (Reinisch & Weber, 2012). The different pronunciation patterns lead the learners to speak with a foreign-like and less-understandable accent, with
consequences such as hindering communication (Flege, 1995). The purpose of this current investigation was to find the phonetic contexts (environments of stressed syllables) where the phonemes /v/ and /w/ are pronounced differently in comparison to how a native speaker of English would pronounce them. On the basis of Hill (1970), Noroa, Fein and Obler’s (1988) statement that “foreign accents are not inevitable” (Flege, 1995, p. 236), this study adds insight into the phonetic environments and of the Spanish accent in English when
pronouncing /v/ and /w/ in order to avoid this accent. The target accent here is not the Spanish accent in general, as it does not constitute a negative aspect of language, the focus is on a foreign-like accent which hinders communication between speakers, Furthermore, for the purpose of this study, in order to avoid confusion with the differences between the varieties of Spanish, the focus was on the Spanish variety spoken in the Iberian Peninsula: Castilian. The present study will not only help ESL teachers but also Castilian ESL speakers to sound more understandable and less heavily-accented. Consequently, the ultimate aim of this research was to prove whether the existing theories of the phonetic contexts of /v/ and /w/ and stress
2. Literature Review Spanish Language Transfer in English Speech Production
In order to understand the current level of English in Spain, it is of relevance to look at its historical background. Spain’s level of English, in contrast to other countries in Europe, is low (Kingsley, 2011) due to several reasons. Until Franco’s death in 1975 no languages other than Spanish were allowed to be taught in Spain, and after 1975 the government established English as the primary foreign language. However, this brought problems in the educational system as there were barely any English teachers. Instead, they focused on fostering regional languages such as Catalán (Catalan) and Vasco (Basque). But when Spain joined the
European Union in 1985, there was a drastic change in the focus of language learning, as it shifted to English. Thus, unlike other countries in Europe that have been teaching and learning English for a longer period of time, in Spain English has only been taught and considered important for two decades. This means that there is a large generation of 40/50-year olds who did not receive proper English classes in school and thus find themselves in a language a gap when compared to the English education younger generations have
received/are currently receiving. This lack of education affects knowledge and English skills, which as a domino effect affects their job position as now companies expect them to be fluent in English. Due to this threat, these adult learners are therefore extrinsically motivated to learn English in language academies and are willing to invest in it. But Castilian speaker find English very difficult to learn, specially pronunciation. They carry a natural Castilian accent when speaking English, which hinders communication, making it problematic in the
professional setting and threatening their job positions (Kingsley, 2011).
One of the assumed reasons for struggling with English is the influence of one’s first language (Ellis, 2008). When speaking a foreign language, the manner of pronunciation is
assumed to be linked to the person’s mother tongue (Corder, 1973). According to Flege (1995), this interference is the main factor when producing a foreign-like accent. More
specifically, Canale (2012) mentions a type of interference called ‘interphonology’, which is a set of rules about the pronunciation of a language affecting the foreign language. As speaking with an accent is natural to speakers, the focus is on realizations of accents which challenge communication. Fledge (1995) also adds that the accent could be due to the lack of accurate hearing of the foreign sounds, due to limited amount of exposure and input, making it difficult then to pronounce utterances in a native-like manner it correctly. This could be the case for Spanish speaker as in Spain foreign films are dubbed into Castilian (Kingsley, 2011), and thus reducing the amount of exposure to English.
For this reason, the target group of the current investigation is a group of 40/50-year olds who experience the mentioned language gap, struggle with pronunciation and invest in private courses provided by their employers. It is important that these ESL learners are made aware of the phonological contexts their foreign-like accent becomes more prominent at (Crystal, 1997). By doing so, the ‘mistake’ will be placed in its environment (Shea & Curtin, 2011) providing the ability to explain the (mis)pronunciation. As a consequence, these learners will be helped to improve their pronunciation skills.
Environments of /v/ and /w/ in English
A common and recognisable feature of the Spanish accent (including the Castilian variety) in English is when the utterance includes the phonemes /v/ and /w/. The labio-dental fricative [v] is commonly pronounced as the bilabial plosive [b] or the bilabial fricative [β] characteristic of the Spanish phonetic system; and lastly, the bilabial approximant [w] is pronounced as the velar plosive [g] (Coe, 2001). For this reason, phonetic contexts in stressed syllables of these phonemes were analysed in English and Castilian Spanish firstly, in order to
find where the accent-prone pronunciation takes place. The two word-locations that were analysed were stressed word-initial or stressed word-final syllables.
When one analyses the phonological distribution of /v/ in English it needs to be done so in relation to the phoneme /b/ as it is one of the leading causes for the Spanish accent (Coe, 2001). Both the bilabial plosive /b/ and the labio-dental fricative /v/ can be found in the English phonetic system. The differences between these two phonemes are the place and manner of articulation. In relation to their phonetic context, both phonemes can occur at word initial and word-final (Canale, 2012). Furthermore, as both phonemes have such different features, they are easily distinguishable by native speakers. Additionally, /b/ is voiceless when syllable-initial (Harris, 1994). In English, these phonemes are allophonic, because they create difference in meaning and can occur in any place of the word (Carr, 2013).
The analysis of the distribution of /w/ in English becomes slightly complex due to its dependence on the phonetic context. Depending on the context, this phoneme can be
articulated in different ways. The consonant /w/ is a voiced central approximant, a glide. It is considered to be a labial-velar (Carr, 2013) because of the raised position of the tongue close to the velum in addition to the lips being rounded. Arguably, there are two places of
articulation: the velar place is considered the primary place of articulation, and the labial the second (O’Grady et al., 2011). In relation to vowels, the articulation of the /w/ will adapt according to the one of the following vowels (high/low, front/back, rounded/unrounded). It can also be found in consonant clusters, namely with stop consonants: /tw/, /dw/ /gw/ (Coe, 2001). When found in the cluster of the voiceless stops /p, t, k/ it becomes voiceless when it is preceded by one of these stops (Ladefoged, 2014).
Environments of /v/ and /w/ in Spanish
The analysis of the /v/ in Spanish is tricky due to its relation to the phoneme /b/. This is mainly because the phoneme equivalent to the English /v/ is non-existent in the Spanish phonetic system. Instead, it has two allophones of /b/, [b] and [β] in complementary
distribution which alternate according to the phonetic context they are found at. In addition, Spanish is a syllable-based language, and this stress distribution also affects the pronunciation of the target phonemes. It is “the position in the word and position relative to stress” which are key factors (Shea and Curtin, 2011, p.230). The first allophone, the voiced bilabial stop [b], “occurs post-pause and tends to occur with stress” (Shea and Curtin, 2011, p.230) and “post-nasal consonant” (Canale, 2012, p.158). On the other hand, the bilabial fricative [β] “occurs intervocalically and does not tend to co-occur with stress” (Shea and Curtin, 2011 p. 230). In other words, just as Canale (2012) argues, [β] occurs in all contexts where [b] does not occur because they are in complementary distribution. Interestingly, when the /b/ is found between vowels, they are pronounced more softly and continuously, distancing itself from a stop and nearing the features of the fricative /v/ (Coe, 2011). It becomes clear then, that the pronunciation of the English /v/ can become problematic for Castilian Spanish ESL learners as they might transfer their Spanish /b/ knowledge to use it as bridge towards the English /v/.
Lastly, the pronunciation of the English /w/ is troublesome for Spanish speakers. Firstly, they tend to pronounce it like a /b/ rather than /w/ when found between vowels (Coe, 2001). Therefore, in the context of finding the /w/ in the coda position followed by a vowel nucleus, the Spanish accent will be more prominent as the realisation could sound like a /b/. Additionally, instead of the /w/ sounding like a /b/, a velar plosive /g/ is added, pronouncing would as /gwud/ or /gud/ (Coe, 2001).
Stress: Syllable-timed language vs. Stress-timed languages
Besides the differences between the phonological systems, a factor which makes these two languages very different is their stress pattern. This aspect was considered in this study as one of the possible leading causes of the foreign accent in the pronunciation of /v/ and /w/. English, unlike Spanish, is a stress-timed language, which means that “regular reoccurring beats found in the speech of English speakers fall on stressed syllables which occur more or less at equal intervals” (Carr, 2013, p.99). In English, stressed syllables are pronounced more than unstressed syllables as these often get reduced; the speech rhythm goes according to this opposition of strong and weak syllables (Fear et al., 1994, p. 1893). One of the consequences is that content words are stressed and thus more pronounced than function words (Coe, 2001). This consequence is relevant in terms of how words containing /v/ and /w/ may be
pronounced because Spanish is a syllable-timed language. A syllable-timed language is one in which generally speaking the time spent on pronouncing each syllable takes roughly the equal length of time (Coe, 2001). Unlike English, Spanish does not offer a contrast of
weaker/unstressed counterparts, making no distinction between constituents of greater intensity or longer duration and weak and short (Heuven and Menert, 1996, p. 2439). Therefore, stress can be a transferred item from the L1 creating a string of monotonous stressed utterances in English allocating possible erroneous word-stress. This does not only happen to Spanish speakers as “word-stress errors are naturally very frequent in learning English as a foreign language” (Porzuczek & Rojczyk, 2017, p. 313). For this reason, the focus of this research is on stressed syllables, as they resemblance between English and Spanish is highest. This allows to compare the pronunciation of the syllables containing the target phonemes clearly.
In sum, this study will contribute to the field of Language Acquisition by adding insight into the reasoning behind the articulation of the Spanish accent in English when
pronouncing /v/ and /w/, which will help ESL teachers and consequently enlighten their Spanish ESL students to sound more understandable (especially useful in the working field).
Research questions
The purpose of this present study is to gain insight of the troublesome pronunciation of /v/ and /w/ by Castilian Spanish ESL learners. Therefore, the questions this research aims to answer are the following:
1) What are the expected pronunciation problems for Castilian Spanish ESL learners on the basis of a contrastive analysis of the two phonetic systems in relation to /v/ and /w/?
2) What are the actual problems in the aforementioned learners related to the pronunciation of /v/ and /w/?
3) Are these problems related to word stress assignment in syllable initial or final? In order to answer these questions, an analysis of the pronunciation of adult Spanish learners of English was carried out, followed by data analysis.
3. Methodology
This section explains the procedure of the data collection and data analysis. Firstly, the variables analysed are explained, followed by information about the participants, speech material, research tools and procedure.
Variables: phonological patterns of /v/, /w/ and /b/ according to their environment
The main variables here are the phonemes /v/ and /w/. When looking at the pronunciation of the aforementioned phonemes, it is of relevance to analyse the context within the word. For the purpose of this study, only stressed syllables where the phonemes are found were analysed
because in unstressed syllables the phonemes are more prone to be exposed to syllable reduction or lack of emphasised pronunciation. Therefore, the environments that were
analysed were both stressed syllable initial and stressed syllable final. In addition to the main variables the phoneme /b/ has been included as well. The reason for this is due to its mix-up with the target phonemes /v/ and /w/ in English by Castilian speakers. Regarding the phoneme /w/, although it never takes place as a consonant in syllable final, it has been included for the sake of consistency. The words used to analyse the participants’ speech are the ones shown in table 3.1. Each phoneme has five words in a stressed syllable initial and final position
respectively. Therefore, a total of 30 tokens were analysed (see table 3.1). Table 3.1. Environment and words used.
Phoneme Environment and Condition Example
/v/ Stressed syllable initial
Onset Consonant
Very, value, village, Victor, visibly.
Stressed syllable final Coda Consonant
Love, arrive, leave, dove, curve
/b/ Stressed syllable initial
Onset Consonant
Borrow, bored, business, bye, umbrella
Stressed syllable final Coda Consonant
Globe, tribe, tube, job, Bob
/w/ Stressed syllable initial
Onset Consonant
Wooden, wonderful, weather, winter, overwhelming
Stressed Syllable Final Coda Consonant
Participants: Speakers
The participants selected for this study are five adult Castilian Spanish speakers of English (as a second language). They are all at an advanced level (C1 in the CEFR scale), have been learning English for several years, and still do today (see table 3.2 for participants’
information). They receive classes from the company they work at: in-company classes (teachers who go into the company and give specific business English classes to cater individual or group needs). As mentioned before, the generation of 40/50-year olds have not received proper English education during high school and therefore are at a disadvantageous position in comparison to younger generations entering the working field, creating a threat to their positions in their companies. All their knowledge comes primarily from language school lessons received after secondary/university education. They struggle with pronunciation and independently from their high levels of motivation, discipline and perseverance still produce a Spanish accent when speaking English. The research focus on the English learning process by children and teenagers has received a great deal of attention whereas this ‘gap-generation’ has not. Therefore, this research focuses on this group in order to enhance their pronunciation skills.
Table 3.2 Participants’ Language Characteristics
Speaker Characteristics
1 Male, 40-50 age group, C1 level
2 Female, 35 – 45 age group, C1 level
3 Female, 40 – 50 age group, C1 level
4 Male, 40 -50 age group, C1 level
Speech Material: Speaking Task
The participants who took part in this research were asked to complete two tasks. In the first one, they were asked to read the following story:
My uncle John has a very good job. He’s a university professor, actually, and very intelligent. He is a man of proper manners. He is the kind of person who will always arrive on time, never leave without saying good-bye, will always give a hand to someone who needs it and who is visibly kind.
As I said, my uncle is very intelligent. I don’t know how, but he seems to know all countries and each tribe on the globe. I guess it is because he is constantly, working reading and learning new things, which is why he is never bored. However, he would never show off as he is very humble.
The strange thing is though, he’s always losing things. It’s quite extraordinary. For example, last Thursday, a cold winter day when the weather was terrible he
accidentally left his umbrella on the London tube. He was on his way back to the
village from his business trip. It must have fallen on the ground in a steep curve
without him noticing. It’s a rather special one, with red and yellow stripes and a
wooden handle. It was a present from his sons Victor and Bob for his birthday one
year. Therefore, it has special value to him and lets no one borrow it. You can often hear him say: ‘I love this wonderful umbrella so much, I will never throw it away!’ It
was quite an overwhelming situation for him to realize he had lost it. The next day, as
soon as he was free, he flew as fast as a dove to the Lost Property Office to ask about it. Fortunately, it’s in the next street to his house. He’s no stranger to the people there. They know him quite well.
This text is an adapted version of the picture-story from the book Learner English by Swan and Smith (2001). The adaptations were made to meet the aims of this current study and consequently added more words containing /v/, /b/ and /w/ in different syllable locations (stressed syllable initial and final). As can be seen, the text provides different words that contain /v/, /b/ and /w/ in different positions of the word. This task is controlled speech: they just had to read the story, without having to create the story themselves. Afterwards they were told to complete the second task. This task required them to finish the story they had read in the first task following given pictures (see figure 3.1). The focus of the second task was on the speakers’ free speech, with the pictures as an aiding tool to give them somewhat of a
allowed them to tell the story and describe the pictures freely, without focusing on the pronunciation. The purpose of the second task was twofold: firstly, allowing the participants to use free speech and secondly, to analyse how recurrent these phonemes are in order to determine the urgency of the problem. Both tasks focus on connected speech. They were not asked to read the words aloud in isolation, to analyse natural speech instead of planned speech, as the pronunciation could be affected by it.
Figure 3.1. Picture-Story for Task 2 Research Tool
The research tools that were used are the following: the aforementioned speaking task (adapted version of the picture-story by Swan and Smith, 2001), the audio recordings and the analysis table. The speaking task was used during the data collection, as the main research
material. The audio recordings have a length of between a minute and a half and two minutes. They were the tool to find the (in)correct pronunciation of the words containing /v/ /b/ and /w/ and the analysis table was completed accordingly. This table was used twice for each participant: once for the analysis of the first task (controlled speech) and another for the second (free speech) task. The data analysis table (see table 3.3) is an example of how the analysis was approached.
Table 3.3. Sample of Data Analysis. Participant X
Phoneme Environment in
stressed Syllables
Clear
pronunciation
Problematic pronunciation
/v/ Initial Victor, visibly, village, value,
very
Final Dove, arrive, love Curve, leave
/b/ Initial Umbrella,
business, borrow, bored, bye
Final Tube, job Tribe, Bob, globe
/w/ Initial Wonderful,
weather, winter, would
Wooden, Overwhelming
Final Know, how, throw,
show, flew
Procedure, analysis and statistical treatment
As explained previously, participants were asked to complete two tasks, which were audio-recorded, transcribed and categorised. When categorising the words according to the location of the phoneme within the word, they were placed either in the boxes ‘clear pronunciation’ or
‘problematic pronunciation’. This revealed the location within the word (syllable initial or final) where most of the errors take place. It also revealed consistency throughout the analysis of the five participants, whether there were common error patterns. The consistency was binary: on the one hand amongst all participants and on the other hand within the speech of a single speaker (comparing the first task to the second).
4. Results
The purpose of this study was to find the pronunciation problems Castilian Spanish ESL learners experience in relation to the phonemes /v/ and /w/. In order to answer this, the speech of five adult Spanish speakers was analysed according to the environment of the /v/, /w/ and /b/ (stressed syllable initial or final) and located the place of the mispronunciation considered problematic and/or hindersome. Additionally, the factor of stress was taken into consideration to assess whether it affects the mispronunciation.
The results can be categorised into three sections: first the overall results of task number will be shown, followed by the results of the second task, to be concluded with individual results (comparing both tasks).
Overall first task results
Despite showing individual differences, there are some results in the first task which all speakers had in common. Firstly, after analysing the pronunciation of the /v/ in syllable initial position, out of the five target words two were found to be mispronounced incorrectly by all five speakers (very and Victor), pronounced as a [β] instead of a [v]. Furthermore, the three remaining words were mispronounced by four of the five speakers. None of the selected words were pronounced correctly with the /v/ in stressed syllable initial position. In the section of /v/ in the stressed syllable final position the most often mispronounced word was
dove. It was pronounced as [dLf] instead of [dLv] by four of the five speakers whereas its minimal pair love was pronounced correctly by four speakers (see table 3.1).
The phoneme /b/ was included in order to analyse whether it caused interference with the pronunciation of /v/. The results show that in stressed syllable initial position, no
mispronunciations took place, whereas a different outcome can be seen in the stressed syllable final position. All the words from this section were mispronounced by at least one speaker. The most prominent mispronunciation was produced in the word tube, placing the word tribe as the second most mispronounced word as four of the five participants struggled with it. Additionally, four of the five speakers also mispronounced the final /b/ in the proper noun Bob whereas its minimal pair job was pronounced correctly by four out of five speakers. In both words the /b/ was pronounced as the voiced labio-dental fricative [f] instead of the correct voiced bilabial plosive [b] (see table 4.1).
The results of the analysis of the pronunciation of the /w/, show most struggles can be seen in the stressed syllable initial position: all speakers did not pronounce the word overwhelming correctly, creating a mid-word pause to pronounce the word and yet
mispronounced it, assimilating the /w/ and focusing on the pronunciation of the /h/ instead. Moreover, more than half of the participants mispronounced the words wooden and would, adding an initial /g/ in front of /w/ producing the cluster /gw/. Lastly, when finding the /w/ in syllable final position, no significant mispronunciations occurred, which was an expected result as here it acts more like a vowel, but was still included for consistency.
Table 4.1. Category of pronunciation of target words. Phoneme Syllable
Position
Word No. Speakers
with Clear pronunciation No. Speakers with Problematic Pronunciation /v/ Initial Very 0 5 Value 3 2 Village 1 4 Victor 0 5 Visibly 1 5 Final Love 4 1 Arrive 5 0 Leave 4 1 Dove 1 4 Curve 3 2 /b/ Initial Borrow 5 0 Bored 5 0 Bored 5 0 Business 5 0 Bye 5 0 Umbrella 5 0 Final Globe 2 3 Tribe 2 3 Tube 0 5 Job 4 1 Bob 1 4 /w/ Initial Wooden 1 4 Wonderful 5 0 Winter 5 0 Weather 5 0 Overwhelming 0 5 Final Know 5 0 Show 5 0 Throw 5 0 Flew 3 2 Yellow 5 0
Results Task Two
The second task the participants had to complete was free speech: they were to finish the story following the pictures. Due to this it was possible to see how often these phonemes were actually used, revealing the urgency of the problem, but this will be discussed in the
second task, the total is forty-six, an average of nine per person in comparison to thirty 30 per person that the first task provided. Most tokens found were those containing /w/ and /v/ in syllable initial position. There are three words which re-occur in both tasks, in both they are mispronounced: very, tube and wooden (see table 4.2). The discrepancy between the two tasks – mainly the reduced number of tokens used in the second task, show the urgency of the problem.
Table 4.2. Words with target phonemes used by each speaker in Task 2. Phoneme Syllable
Position
Speaker Clear Pronunciation Problematic
Pronunciation /v/ Initial 1 Very 2 Very 3 4 Very 5 Very Final 1 Have 2 3 Give 4 Have, leave 5 Gave
/b/ Initial 1 Umbrella, book
2 Umbrella, but, back
3 Umbrella
4 Umbrella, be, but
5 Umbrellas, book, bye
Final 1 Tube
2
3 Tube
4 Tube
5 Describe
/w/ Initial 1 Was, weeks,
paperwork
2 Well, was, will
3 Wooden
4 Was, when
5 Went, was, where,
when, what
Final 1 Wooden Show
2
3 Yellow Saw
4 Yellow
Individual problematic pronunciation patterns
There are a number of mispronunciations which show systematic consistency at it happened in in both tasks. For example, a speaker mispronounced wooden in both tasks, others mispronounced tube and very in both. In contrast, some speakers used the same word in controlled speech and free speech but mispronounced it in one, showing intraspeaker
variation: for example, a participant in controlled speech mispronounced wooden whereas he correctly pronounced it in free speech. Yet another participant used the same word in both tasks but the results were opposite. Furthermore, another speaker pronounced leave correctly in controlled speech and pronounced as [li:f] in free speech. Finally, the word tube (which was one of the most troublesome words) was pronounced correctly in controlled speech but incorrectly in free speech by another speaker (see table 4.3).
Table 4.3. Problematic Pronunciation within Speakers Speaker Free and Controlled Speech Problematic Contexts
1 Very (problematic in both contexts)
Tube (problematic in free speech)
Wooden (problematic in controlled speech)
2 Very (problematic in both contexts)
3 Tube (problematic)
Wooden (problematic in free speech)
4 Tube (problematic in both contexts)
Very (problematic in both contexts) Leave (problematic in free speech)
5 Very (problematic in both contexts
The most outstanding results show, therefore, that the troublesome location for the /v/ is at stressed syllable initial position whereas for the /b/ is it stressed syllable final position. Lastly, stressed word initial position (and not in syllable final) of the /w/ was also problematic. In order to explain these mistakes in further detail, the three most troublesome words were selected and analysed: very, tube and wooden.
5. Discussion
This research aimed to answer three main questions in the field of linguistics and second language acquisition theories. The first question pondered upon the main pronunciation problems in relation to the phonemes /v/ and /w/ that Castilian Spanish adult ESL learners struggle with. In order to get an answer an analysis of the environments where the phonemes were realized was carried out. The two phonetic systems (Spanish and English) were
compared revealing some overlapping pronunciations which was suggested to be a leading cause to error. After knowing what problems learners could face when speaking English, the second research question was explored. The speech of five participants was analysed to evaluate what problematic phonological environments the participants struggled with. In other words, the correlation between the theory and practice was analysed. Lastly, the factor of stress was analysed in order to see whether this caused interference and lead to make mistakes.
As discussed in previous chapters, the expected problems for Spanish ESL speakers of /v/ and /w/ were revealed in the contrast of phonemic systems of both the languages in question. The research question parted from the hypothesis that these errors could arguably be the result of mother tongue interference (Ellis, 2008). In point of fact, some of the errors in specific word locations could be explained through language interference theories. The clearest
example of this interference can be seen in the participants’ use of /v/ and /b/ interchangeably. As the Spanish system lacks the phoneme equivalent to the English /v/, the transfer of the Spanish phonemes [β] and [b] cause interference. Additionally, in the results of the analysis the complication of pronouncing the /w/ as a consonant proved itself problematic. As the results show, the most representative errors showed in the words very, tube and wooden, which is why these three were selected and analysed to provide answers to the main research questions.
To answer the question about the problems with the /v/, the word very was taken as an example. Even though very can be considered a function word in English, it is a degree adverb modifying, adding emphasis to an adjective or an adverb (Cambridge Dictionary, 2019). Thus, speakers pronounced it with emphasis and stress as if it were a full lexical word, making it the most mispronounced token in the selection of words in this research. This could be due to different reasons. Firstly, as has been clarified, the /v/ phoneme is non-existent in the Spanish phonetic system and as becomes apparent through the researches’ data, it has not been adopted by the Spanish ESL participants, subsequently pronounced as a [b]. When looking at this in more depth, the Spanish bilabial stop [b] occurs in a very specific
environment: with main stress and post-pause (Shea and Curtin, 2011), which would explain why all participants pronounced it as a stressed and lexical-like word, after a forced pause. Secondly, in relation to stress, this stress mark could be justified as follows:
“When speakers are deliberately trying to articulate clearly; they pause at word boundaries preceding weak syllables but not at word boundaries preceding strong syllables; i.e. they mark precisely those boundaries which the observed listener behaviours would not detect” (Fear et al, 1994, p. 1893).
In other words, the word very in the context of the example sentence (a very good job), is found between a reduced syllable (a) and a lexical word (good). As the word very both follows a weak syllable and precedes a lexical word with main lexical stress it could either have been a reduced word or a strong word, yet a pause was created in order to speak clearly and audibly (Fear et al, 1994). Consequently, the word boundary was exaggerated, leaving room for the stress to be exaggerated after the reduced article and thus emphasising the fall of stress on the first syllable of the following word (containing the consonant under investigation ‘very’). Another fact to point out in terms of mother tongue interference, is that in Spanish, a syllable-based language, has a bias towards syllable initial stress (Coe, 2001), which could be another argument explaining the participants’ pronunciation. What’s more, this can also be seen in learners whose first language is not Spanish. For example, studies carried out by
Porzuczek and Rojczyk (2017), suggest that Polish speakers also tend to stress the first syllable rather than the last syllable. It could be suggested then, that just like Polish ESL learners, Spanish ESL learners also reflect a stress bias in their English speech.
Not only is the mother tongue an influence but also the instruction about stress assignment they might have received. These participants could have been informed about the typical English stress position such as “in English the typical stress position is word initial” (Heuven & Menert, 1996, p. 2439); and English uses “contrastive stress as a cue to word
identification” (Porzuczek & Rojczyk, 2017, p. 313). Therefore, there are two factors that play a role in their pronunciation: learners could overgeneralize their knowledge about English lexical words receiving stress and the word initial stress tendency and secondly, the learners could carry the knowledge of the different environments where the Spanish [b] takes place transferring it into different English contexts. With these factors in mind, it is explicable why Spanish ESL speakers pronounced very with the initial plosive emphasized [b].
When analysing the pronunciation of /b/ the troublesome aspect does not occur at syllable initial as one might have expected; rather, it takes place in stressed word final. The word all participants mispronounced here was tube pronouncing it as [tjuf]. The pronunciation varied between /f/ and /v/ which contrasts with their pronunciation of very as /vɛri/. A possible explanation for this could be that in Spanish, when the /b/ occurs intervocalically it is pronounced more softly, getting articulatorily closer to an English fricative /v/ (Coe, 2001). Even though the /b/ is located at the end of the syllable, and the phonemic transcription places it in the coda position, the spelling suggests it precedes a final vowel, so when reading the word out loud the spelling of such could suggest the reader to read it with the understanding of it being placed intervocalically, and therefore producing the sound of a fricative instead of a plosive.
Furthermore, according to Heuven and Menert (1996, p. 2240) the final syllable, if it is followed by a silence, which was the case for the word tube, “should be longer than the other syllable(s)”. Therefore, since the ending of the word tube (which is monosyllabic) ends in a stop, it makes the syllable shorter, stopping abruptly. Stops are less sonorous and shorter than fricatives (Carr, 2013), so it could be argued that the speakers chose to use the fricative to extend the sound, making the word longer with the ultimate aim to make themselves
understood more clearly. Other students with different mother tongues also struggle with this issue, such as the Dutch. Dutch speakers have the tendency to “often pronounce syllable-final voiced obstruents in English as voiceless” (Reinisch & Weber, 2012, p. 1165), making the gap towards a voiceless fricative smaller. Spanish speakers could share this tendency with the Dutch. In conclusion, the two main possible causes for mispronouncing the /b/ in word final position could be to the mother tongue interference when the target consonant is placed between vowels and when it is located before a long pause.
Lastly, wooden was the word that caused most problems with the /w/, which is why it was selected as the most representative mistake. When this word was pronounced, the velar plosive /g/ was added in the coda position which converted it into the cluster of /gw/. Therefore, they pronounced the /w/ with the properties of a vowel. This was an expected result based on the theory, as Coe (2001) explained that the /w/ can be found in clusters with stop consonants such as /tw/, /dw/ /gw/. A possible explanation to this is connected to the special features of the /w/. As a glide, /w/ is more sonorous than other consonants and can be realized with the features of a vowel depending on its context such as word final in now (Carr, 2013). The glide /w/ and the back vowel /u:/ share most of their features: [-consonantal, +sonorant, +continuant, +voice, oLABIAL, +round, oDORSAL, +high, +back], except the syllabic feature, /w/ being [-syllabic] and /u:/ being [+syllabic] (O’Grady et al, 2011) This is because vowels can be the nucleus of a rhyme within a syllable but glides such as /w/ can’t.
Hence, when found next to a consonant it will take the properties of the following vowel, but when found preceding a vowel or a set of vowels such as wood or would Spanish learners may opt to pronounce it in two ways: either pronounce the /w/ with the features of the glide connecting them to those of the vowel, or they assimilate regressively adding a velar /g/ in the onset position creating a /gw/ cluster. Learners of English as a second language are made aware of the fact that the properties of the /w/ are similar to those of a vowel. As a
consequence of this instruction, learners could feel the need to include a consonant in the coda position, adding the /g/ and making a CV (simple consonant + vowel) syllable structure. As Carr (2013) explains, the CV structure is the most basic syllable structure in human languages and can easily be accepted as valid in new languages. There is a general wider range of allowance of different consonants that can be placed in the onset position, making it easy for Spanish ESL speakers to add a velar in front of a glide, and therefore the glide acts similarly to the round back vowel /u:/. By doing so, the CV structure is completed successfully. In sum, due to the vowel-like properties of the /w/, Spanish learners might tend to add a consonant in front of it and pronounce /w/ as a vowel instead.
Using these three words (very, tube and wooden) as examples of the problematic contexts for Spanish ESL learners, it becomes clear that the mistakes made by them are rooted into the difficulty of finding the right way to pronounce them but they can simultaneously be
explained in phonological terms and language interference theories. Some of the cases are clear L1 interference as explained before, yet others might be due to the overgeneralization of stress pattern rules they have learnt about English.
All the phonemes under investigation were placed purposefully in stressed syllables to avoid syllable reduction of weak syllables. Stress was thus a factor that played a role in these contexts. Because of this, the learners pronounced the syllables with emphasis, making the target phonemes audible. All the speakers are at an advanced level of English (CEFR C1) and
yet they mispronounced the phonemes, which raises the question whether proficiency, is the key solution to the mentioned pronunciation problems. Instead, in order to enhance the speakers’ pronunciation skills, these students should receive more detailed instruction as to how to pronounce it in different contexts. Porzuczek & Rojczyk (2017) in their research found that the learners with higher-proficiency did use the correct stress patterns more often than otherwise. However, it was also mentioned that it was not exclusively due to their high levels of proficiency that this accuracy took place. The participants in their study had previously “received explicit instructions and training in word stress patterns in English” (Porzuczek & Rojczyk, 2017, p. 320). Therefore, instructing learners to mark stress appropriately and teaching them how to pronounce and distinguish the phonemic systems are keenly important to sound more native-like. As mentioned by these researchers, “proficiency and phonetic training are significant predictors of English word stress realization” (Porzuczek & Rojczyk, 2017, p. 320). This correct usage of stress helps the speech rhythm which enhances native-like accents, because “the differences in speech rhythm between utterances produced by L2
learners at different proficiency levels make a greater contribution to perceived foreign accent than differences in speech rate” (Polyanskaya et al., 2017, p.350).
In sum, the three main problems these participants struggled with were the pronunciation of the /v/ in stressed syllable initial position, /b/ in stressed syllable final position and /w/ in stressed syllable initial position. These problems can be traced back to mother tongue interference through ample evidence in existing theories and studies. Another source of the difficulty could be the overgeneralization of acquired rules about English. These problems can be fixed with the help of the ESL teachers in raising awareness of the phonetic environments where the problems occur.
6. Conclusion
The current study aimed to find the reasoning behind the problems Castilian Spanish speakers face when pronouncing the English /v/ and /w/. When comparing it to other studies and existing theories it becomes apparent that the problem can be traced back to mother tongue interference and overgeneralization of learnt rules about English. Therefore, these problems can be fixed by raising awareness to learners about this issue, investing and prioritizing this aspect. The practical applications are therefore the ones mentioned above. Teachers of English, besides focusing on increasing the students’ levels of other skills and structures (i.e. reading, writing and grammar and vocabulary), time should be spent
explaining to the students how and when to pronounce the /v/ and /w/ respectively. This will help the students sound more native-like level of English, which will avoid communication hindrance and increase their confidence levels when approaching professional situations in which English is needed.
This research has limitations which should be addressed in order for this study to be used as a reference for future research. Firstly, to clearly analyse where the mispronunciation of the target phonemes took place, they were only placed in stressed syllables. For future research it could be of interest to include the phonemes in weak and strong syllables (stressed and unstressed) to examine whether there is a different pattern in pronunciation when the
phonemes are found in weak or strong environments. Secondly, both tasks used to collect data were focused on connected speech, as it reflects natural communication and speech. However, it could be useful to analyse the speech of participants using both isolated and connected speech. For example, a participant could be asked to read independent words out loud containing the target phonemes (isolated speech), followed by reading a story (connected speech) to explore whether there is a difference in pronunciation between the two types of speech. This strategy could reveal the level of pronunciation awareness of the speaker and
whether there is intraspeaker variation. Lastly, in the second task the participants were asked to finish the story without specific guidelines related to the phonemes /v/ and /w/ and the results showed that there was a significant decrease in the frequency of usage of the
phonemes, meaning that either the speakers chose to avoid using words containing /v/ and /w/ or perhaps the usage in daily speech is not very frequent, which would decrease the urgency of the problem. This research could be used as a pilot study to research more problematic English phonemes for Spanish speakers, such as vowels and the consonant /h/.
Bibliography
Best, C. (1993). Emergence of language-specific constraints in perception of non- native speech: A window on early phonological development. In Developmental
Neurocognition: Speech and Fce Processing in the First Year of Life, ed. B. de Boysson-Bardies, S. de Schonen, P. Jusczyk, P. MacNeilage, and J. Morton. Dordrecht, the Netherlands: Kluwer Academic Publishers.
Cambridge Dictionary (2019) Retreived in May 2019 from:
https://dictionary.cambridge.org/es/diccionario/ingles/very
Canale, G. (2012) La Adquisición de /v/ en la interfonología espanol-inglés: un estudio sociolinguistic de estudiantes montevideanos. Elia (Estudios de lingüística inglesa aplicada), 12 (2012) pp.153 – 180).
Carr, P. (2013) English Phonetics and Phonology. An introduction. (2nd Ed.) Singapore,
Wiley-Blackwell
Coe, N. (2001) Speakers of Spanish and Catalan. In: Swan, M. & Smith, B. (2001) Learner English: a teacher’s guide to interference and other problems (3rd ed.) Cambridge: Cambridge University press.
Corder, S.P (1973) Introducing applied linguistics. London: Penguin Education. Crystal, D. (1997) The Cambridge Encyclopedia of the English Language. Cambridge,
Cambridge University Press.
Ellis, R. (2008). The Study of Second Language Acquisition. Oxford: Oxford University Press Flege, J. E., and Eefting, W. (1987). The production and perception of English stops by
Spanish speakers of English. Journal of Phonetics 15:67-83.
Flege, J. E., Munro, M., and MacKay, I. (1995). Effects of age of second-language learning on the production of English consonants. Speech Communication 16:1-26.
Flege, J.E. (1995) Second Language Speech Learning; Theory, Findings, and Problems. In “Speech Perception and Linguistic Experience”: Issues in Cross-Language Research. Timonium, MD: York Press,
Hardy, J. E. (1993) Phonological Learning and Retention in Second Language Acquisition. John Benjamin’s Publishing Company.
Harris, J. (1994) English sound structure. Oxford: Blackwell.
Heuven v. V. J., and Menert, L. (1996) Why stress position bias? The Journal of the Acoustical Society of America 100, 2439 (1996); doi: 10.1121/1.417952
Hill, J. (1970) Foreign Accents, language acquisition, and cerebral dominance. Revisited Language learning 20:237-48
Kingsley, J. (2011) English Language Market Report: Spain. A report prepared for the UK’s English Language sector. British Council. www.britishcouncil.org
Ladefoged, P. & Johnson, K. (2014) A Course in Phonetics. (7th Ed.) Boston, MA: Thomson
Wadsworth.
Noroa, L., Fein, D., and Obler, L. (1988). Talent in foreign languages: A case study. In The Exceptional Brain, Neuropsychology of Talent and Special Abilities, ed. L. Obler, and D. Fein. New York: Guilford Press.
O’Grady, W., Archibald, J., Katamba, F. (2011) Contemporary Linguistics: An Introduction (Learning about Language). Pearson Education Limited Longman.
Polyanskaya, L., Ording, M., Grazia Busa, M. (2017) Relative Salience of Speech Rhythm and Speech Rate on Perceived Foreign Accent in a Second Language. Language and Speech 2017, Vol. 60(3) 333–355
Porzuczek, A. and Rojczyk, A. (2017) English Word Stress in Polish Learners’ Speech Production and Metacompetence. Research in Language, 2017, vol. 15:4 DOI: 10.1515/rela-2017-0018
Reinisch, E. and Weber, A. (2012) Adapting to suprasegmental lexical stress errors in foreign-accented speech. The Journal of the Acoustical Society of America 132, 1165 (2012); doi: 10.1121/1.4730884
Shea, C. E and Curtin, S. (2011) Experience, representations and the production of second language allophones. Second Language Research 27(2) 229–250
Swan, M. & Smith, B. (2001) Learner English: a teacher’s guide to interference and other problems (3rd ed.) Cambridge: Cambridge University press.
Appendix I. Speaking task for Participants I: Story Task 1. Read the story aloud
My uncle John has a very good job. He’s a university professor and he’s very intelligent. He is a man of proper manners. He is the kind of person who will always arrive on time, never leave without saying good-bye, will always give a hand to someone who needs it and who is visibly kind.
As I said, my uncle is very intelligent. I don’t know how, but he seems to know all countries and each tribe on the globe. I guess it is because he is constantly working, reading and learning new things, which is why he is never bored. However, he would never show off as he is very humble.
The strange thing is though, he’s always losing things. It’s quite extraordinary. For example, last Thursday, a cold winter day when the weather was terrible, he accidentally left his umbrella on the London tube. He was on his way back to the village from his business trip. It must have fallen on the ground in a steep curve without him noticing. It’s a rather special one, with red and yellow stripes and a wooden handle. It was a present from his sons Victor and Bob for his birthday one year. Therefore, it has special value to him and lets no one borrow it. You can often hear him say: ‘I love this wonderful umbrella so much, I will never throw it away!’ It was quite an overwhelming situation for him to realize he had lost it. The next day, as soon as he was free, he flew as fast as a dove to the Lost Property Office to ask about it. Fortunately, it’s in the next street to his house. He’s no stranger to the people there. They know him quite well.
Appendix II. Speaking task for Participants II: Picture Story Task 2: Finish the story as detailed as possible
