Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of UKEssays.com.
The research area of this paper is a cross-linguistic influence or as otherwise known transfer. Specifically, I propose to examine orthographic transfer and its potential effects on English native speakers’ pronunciation of segmentals in German as their second language.
The fact that orthography can influence second language pronunciation is recognized by many foreign language teachers. Nevertheless, there is not enough evidence for such influence and more research is needed (Bassetti, 2008). Previous research on orthographic transfer has mainly looked at its effects on reading and spelling (Jarvis & Pavlenko, 2008). Cook and Bassetti (2005) argue that writing system transfer is separate from language transfer maintaining that “it is not so much aspects of the language that may be carried over as the attributes of a particular writing system” (p. 29). Jarvis and Pavlenko (2008) concur with the above statement; however they also emphasize the importance of the connection between orthographic transfer and language use, since it relates to the orthographic effects on phonology and vice versa. According to Jarvis and Pavlenko, during the development of L1literacy skills, which uses the phonetic alphabet, important sound-letter correspondences are acquired which later can be transferred to the acquisition and use of the second language. Before proceeding any further, the discussion of two key terms that are extensively used in this paper as well as their definitions as proposed by the researchers in the field of SLA is required. The terms are writing system and orthography.
The first term to define is writing system. Coulmas (1999) defines writing system (WS) as “a set of visible or tactile signs to represent units of language in a systematic way” (p. 560). There are various types of writing systems depending on what type of linguistic units they represent (Bassetti, 2005). Consequently, there are writing systems whose linguistic units are consonants hence the name consonantal WSs (for example, Hebrew and Arabic). Morphemic WSs are represented by morphemes (Chinese) and alphabetic WSs are represented by phonemes (English, German, and Spanish). In addition, there are also syllabic WSs whose linguistic units are syllables (Japanese). This paper will specifically focus on alphabetic writing systems of English and German languages which use the same script – the Roman alphabet.
The term writing system relates to orthography which is the second term and is defined as “a set of rules for using script in a particular language, such as symbol- phoneme correspondences, capitalization, punctuation, etc.” (Coulmas, 2003, p. 35). For example, in English orthography the letter Ë‚sËƒ is read as /s/ and in German orthography as /z/ (Benware, 1986). Thus, the same script- the Roman alphabet- is used differently in the English and German orthographies. Orthographies of the same type of alphabetic writing system and script are also varied in their ‘regularity’ of sound-symbol correspondences (Cook & Bassetti, 2005). For instance, in English, the letter Ë‚aËƒ maps to different phonemes in words such as ‘park, bank, and ball’, whereas in German the same letter Ë‚aËƒ has only one pronunciation /a/ as in German nouns Park, Ball, Bank (Goswami, Ziegler, & Richardson, 2005). English orthography is very inconsistent in terms of its sound-letter correspondences thus it is considered to have deep orthography. Languages such as German, Italian and Spanish have relatively consistent letter-to-sound correspondences, so their orthographies are shallow or as otherwise called phonologically transparent (Cook & Bassetti, 2005). Thus, it should be easier for English native speakers to acquire German sound-letter correspondence rules than for German native speakers to acquire English sound-letter correspondence rules. Nevertheless, there has been observed a considerable number of segmental mispronunciations by English native speakers learning German as a foreign language. Therefore, the question arises whether the knowledge of L1 orthography has an effect on how English native speakers pronounce segmentals in German or are there any other factors such as cognate status or the age of acquisition that contribute to the orthographic transfer?.
The focus of this paper is to explore whether the knowledge of L1 (English) orthography, namely its sound-letter correspondence rules, is transferred to L2 (German) and whether it has any positive or negative effects on L2 pronunciation of segmentals in adult language learners of different proficiency levels. The paper also aims at investigating whether there is a significant difference of orthographic transfer produced by learners at different proficiency levels (beginning, intermediate and advanced). Finally, the study will examine whether a lexical factor such as cognate status has either positive or negative effect on the acquisition and use of German as a foreign language.
A number of studies have examined orthographic effects on either spelling or reading. Many studies also have looked at the effects of L1 orthographic experience on L2 word learning and decoding. However, there has been little research examining the effects of orthography on phonology and vice versa. According to Hayes-Harb, Nicol, and Barker (2010) no research has examined spelling conventions of the native and second languages where both share the same script. Although there is some evidence of the relationship between orthographic and phonological representations in learning new vocabulary, more research is needed to shed light on this phenomenon (Hayes-Harb et al., 2010).
As this study examines orthographic transfer from English to German, the notion of cross-linguistic similarity should be taken into consideration especially given the fact that both languages share a significant number of cross-linguistic similarities in cognate, lexical, phonological, and writing systems areas. Ringbom and Jarvis (2009) discuss the importance of cross-linguistic similarities in foreign language learning. Specifically, they maintain that learners rely on their previous linguistic knowledge and look for similarities rather than differences between the L1 and the TL. Ringbom and Jarvis also state that second language learners refer to their previous linguistic knowledge mostly at the beginning stages of second language acquisition. Another important factor such as phonetic awareness and phonetic sensitivity should not be neglected. Piske (2008) maintains that children develop phonetic sensitivity to speech contrasts specific to their L1 long before they begin read and write and at the same time their ability to perceive non-native speech contrasts (L2) weakens with age. So for adult language learners perceiving non-native speech contrast and sounds that are particular to the target language is difficult. To develop phonological awareness and sensitivity would require an extensive practice and frequent exposure to target language. Considering this fact, Piske states that adult language learners are influenced by writing systems of their L1 during L2 acquisition. Consequently this results in learners pronunciation errors which are linked to the reliance on L1 grapheme-phoneme correspondence rules. Especially this refers to the cases when sound-letter correspondences are inconsistent between the L1 and L2.
Odlin (1989) also provides ample evidence compiled from different studies that demonstrate the clear effect of first language influence on L2 speech production and orthography. Odlin states that similar writing systems in L1 and L2 represent a much easier task for learners acquiring a new language. The same however, cannot be said of pronunciation, as other variables may influence decoding of written symbols in L2. Flege, Freida, Walley, and Randaza (1998) examined whether lexical factors such as frequency, familiarity, cognate status and the age of acquisition have an effect on production. The authors provide an insight into what role the cognate status of words can play and how it effects speech production. More specifically, Flege et al. state that degree of perceived relatedness can vary depending on degree of cross-linguistic similarities the L1 and L2 share. Specifically, they point to the proximity of cognate pairs in L1-L2 in their meaning and especially in sound. This paper will also examine whether cognate pairs influence L2 pronunciation of target sounds. Based on these statements it is also likely that English-speaking learners of German will rely on their knowledge of English sound-letter correspondence rules when speaking or reading in the target language. This may result in either mispronunciation of certain sound-letter correspondences that are inconsistent with English orthography or it may have facilitatory effects on L2 phonology acquisition.
In terms of findings several studies particularly stand out. Young-Scholten (2002)1 provides evidence of orthographic effects on the production of German consonants by English speakers. Another study conducted by Young-Scholten, Akita, and Cross (1999)2 also shows evidence of the effects of written representations on the pronunciation of consonant clusters in learners of Polish as a second language. These findings led the researchers to conclude that there is a relationship between orthographic representations and L2 phonology. Other studies investigating the effects of orthography on second language phonology demonstrate that second language learners’ pronunciation can be influenced by orthographic input (Bassetti, 2007).
Bassetti (2007) examined the effects of pinyin (a romanized version of the Chinese language) on pronunciation in learners of Chinese as a foreign language. The author predicted that orthographic representation of pinyin would result in learners’ not pronouncing the main vowel in the rhymes whose transcription does not represent that main vowel as in rhymes /iu/ which maps to [iou], /ui/ – [uei] and /un/ – [uÓ™n]. The results of the study suggest strong influence of pinyin orthographic rules on the pronunciation of Chinese rhymes. Specifically, learners of Chinese as a foreign language would delete the main vowel, which is not present orthographically; however they would always pronounce the same main vowel in the rhymes /you/, /wen/, and /wei/ (Bassetti, 2006; Bassetti, 2007).
One of the most recent studies conducted by Hayes-Harb et al. (2010) provides some evidence of orthographic influence on the acquisition of the phonological forms of new words. In this study, Hayes-Harb et al. recreated conditions in which subjects experienced learning new vocabulary in a new language. The aim of the study was to investigate whether the presence of the written form of the new words affected the learning of their phonological form. Thus, native speakers of English were divided into three groups. The first group was provided with the written forms of new words that were inconsistent with English spelling. The second group was presented with words that were matching the English sound-letter correspondences. The third group had only an auditory input and no written forms of the new words were provided. The results demonstrated that L1 orthography interfered with students’ ability to learn new words especially when new vocabulary differed from English sound-letter correspondences. While Hayes-Harb et al. (2010) provide some evidence of orthographic effects in the process of learning new phonological forms of words, the authors admit that the method is a novel one and in the future should be controlled by learner’s spelling ability.
Previous studies show clear effects of a relationship between orthography and phonology in second language acquisition. Much of the research has analyzed orthographies that use different writing systems, and as Hayes-Harb et al. (2010) point out very little research has been done investigating the effects of orthographies on L2 pronunciation within the same writing system and script. Therefore, the purpose of this paper is to investigate further the notion of orthographic transfer by analyzing the relationship between orthography and phonology of two languages that share the same script and the following research questions are addressed in this paper:
Is orthographic transfer statistically present in the sample?
Do results vary and are they statistically significant between different proficiency levels?
What is the nature of relationship between error score and students’ perceptions of cognate status?
The participants of this study were 28 undergraduate American English speakers from Ohio University. The students were enrolled in the 100, 200 and 300 levels of German as a foreign language. The levels differ by proficiency. The 100 level is offered to beginners, the 200 level is an intermediate level, and the 300 level is taught to the advanced students respectively. Each proficiency level consists of a three quarter sequence meeting four hours per week with regular homework assignments. Each year sequence of German series focuses on developing all four language skills (listening, speaking, reading, and writing).
2.2 Materials and procedures
The study consisted of two phases: the screening phase and the data collection phase. During the screening phase, students enrolled in all three proficiency levels volunteered to fill out an online language background survey. The purpose of this survey was to select potential participants for the study by controlling several variables that could have had an effect on the results. It was set up in such a way that participants could be screened based on their responses to the first three questions. Thus, those who did not qualify for the study were automatically eliminated without going over all of the survey questions. First of all, it helped eliminate the students with any reading or speech disorders. It also helped exclude those students who were taking another foreign language class along with German. The survey also eliminated the students who had taken a foreign language other than German less than a year ago, which could have contributed to transfer from that language into their knowledge and use of German. In addition, information such as students’ motivation, number of years they have studied German and the time spent in German speaking countries was collected through this survey. As a result, out of 63 students who had volunteered to fill out the survey, 28 were qualified and participated in the data collection phase. The selected participants included nine beginners, eight intermediate-level students, and eleven advanced-level students.
The data collection phase consisted of three tasks administered in the following order: an informal reading aloud task, a formal reading aloud task, and a perception task. All three tasks were performed by the students during individual single session meetings lasting fifteen minutes.
Reading aloud is considered to be the best way to test the knowledge of sound-letter correspondences. It provides good control and allows for comparison of speech samples from different subjects. Reading aloud is also a technique that is widely used in a foreign language class thanks to which participants feel more comfortable during data collection (Madsen, 1983). During the informal reading task, students were asked to read aloud a German text which was adapted in order to target certain German consonants and consonant clusters as detailed in Tables 1 and 2. The inventory sets (see Table 1) include incongruent grapheme-phoneme correspondences students presumably would have had the most difficulty producing.
Table 1. Incongruent grapheme-phoneme correspondences
Sounds exist in both languages, but are represented with different letters
Ë‚wËƒ / [v] e.g., Wind [vÉªnt]
Ë‚vËƒ /[v] eg.: van [væn]
Ë‚vËƒ / [f] e.g., Vogel [foËgl]
Ë‚fËƒ / [f] eg.: fan [fæn]
Ë‚sËƒ / [z] e.g., sinken [zÉªÅ‹knÌ©]
Ë‚zËƒ / [z] eg.: zone [zoÊŠn]
Ë‚sËƒ/[Êƒ] when followed by
e.g., Sport [ÊƒpÉ”rt]
Ë‚sËƒ / [Êƒ] when followed by
eg.: shine [ÊƒaÉªn]
Ë‚sËƒ /[Êƒ] when followed by
e.g., Student [ÊƒtudÉ›nt]
Ë‚sËƒ / [Êƒ] when followed by
eg.: shed [ÊƒÉ›d]
Sounds are different in both languages, but are represented with the same letters
Ë‚tËƒ/ [ts] e.g., Tradition [traditsioËn]
Ë‚tËƒ/[Êƒ] e.g., tradition [trÉ™ËˆdÉªÊƒhttp://sp.dictionary.com/dictstatic/dictionary/graphics/luna/thinsp.pngÉ™n]
Ë‚zËƒ/ [ts] e.g., Zimmer [tsÉªmÉ]
Ë‚zËƒ / [z] e.g., zone [zoÊŠn]
Ë‚chËƒ/[ç] e.g., sportlich [ÊƒpÉ”rtlÉªç]
Ë‚chËƒ/ [tÊƒ] e.g., rich [rÉªtÊƒ]
Also is included a set of congruent phoneme-grapheme correspondences that exist in both English and German languages (see Table 3).
Table 3. Congruent grapheme-phoneme correspondences
Grapheme-phoneme correspondences existing in both languages
Ë‚bËƒ maps to [b]
Ë‚nËƒ maps to [n]
Ë‚lËƒ maps to [l]
Ë‚tËƒ maps to [t]
Ë‚pËƒ maps to [p]
Land [lænd ]
Tag [tæg ]
Park [pÉ‘rk ]
To test students’ pronunciation of the targeted German segmentals in relation to potential transfer effects, each grapheme-phoneme correspondence listed in Tables 1 and 2 was encountered at least twice in the informal reading task and once in the formal reading task. Students’ readings in both tasks were recorded in a soundproof recording laboratory using audio recording software. During the first task, students had one minute to read the text silently before being recorded.
The formal reading task followed the informal one and required the participants to read words with the targeted sounds in isolation. The words in the formal task were a subset of those used in the informal task and some of them were cognates with English. Finally, during the perception task students had to categorize a list of both cognate and non-cognate German words in terms of their similarity to any of the English words. Students had to rate the words on a three-point scale (1=same; 2= similar; 3= dissimilar) relying on their own perceptions. Table 3 below contains examples of words students rated in the perception task.
Table 3 Categorization of cognate and non-cognate words
Results from these tasks revealed whether the reliance on first language orthography had a positive or negative effect on students’ pronunciation of segmentals in German. The mispronunciations in the recordings were quantified, categorized and recorded in a data file. Their performance was then codified and entered in the data file. Finally, student survey responses and identifying information collected through the questionnaire were matched with participants’ reading performance results and the perception task. Once the survey information and performance results were merged, the data was de-identified to ensure the confidentiality of the participants.
Data analysis and results
In order to address the first research question of whether orthographic transfer is statistically present in the sample, the Wilcoxon Signed-Rank Test, the nonparametric alternative to the paired-sample t-test was used to examine whether the participants’ pronunciation errors were statistically more prevalent in cases of incongruent sound-letter correspondences between the L1 and L2 than in cases of congruent sound-letter correspondences. That is, we tested whether the participants’ distribution of errors and successes is keyed to the specific ways in which the L1 and L2 are related in terms of sound-letter correspondences. The results of the test revealed that orthographic transfer is statistically present in the sample (p=.00). Students indeed produced more errors in cases of incongruent sound-letter correspondences and there were zero number of mistakes in cases with congruent sound-letter correspondences.
In order to address our second research question concerning whether there is any difference in the mean scores between proficiency level groups, The Kruskal Wallis test was performed. This test was conducted on the results of the reading tasks of the experiment to test whether the subjects from the three proficiency level groups differed significantly in producing German segmentals [ç, f, ts, Êƒp, Êƒt, z, v]. It was assumed that the distribution of errors would be more prevalent in the beginner level group as opposed to intermediate and advanced level groups. The results demonstrated a significant difference in mean scores between all three proficiency groups for both informal and formal reading tasks (p =.009). The results of the descriptive statistics are summarized below in Table 5.
Table 5 Mean error scores for each proficiency level
Beginners group mean score
Intermediate group mean score
Advanced group mean score
Informal reading task
Formal reading task
Cumulative error mean
Finally, in our last research question we examined the relationship between students’ performance results in both reading tasks and their perceptions regarding cognate status of the words. For this purpose a correlation analysis was conducted to examine what kind of relationship there exist between the number of errors produced in the target sounds and the number of students rating the words as same or dissimilar. As a result, correlation coefficient between error score and a number of students rating the words as same revealed a weak negative relationship between these two variables, with correlation coefficient r = -.47. On the other hand, correlation analysis between error score and a number of students rating the words as dissimilar showed a moderate positive relationship between the variables with r = .57.
The present study is exploratory in its nature yet the results show evidence of orthographic transfer in the sample. By that we mean that students do tend to rely on their L1 knowledge of sound-letter correspondence rules specifically in the cases of incongruent sound-letter correspondences between the L1 and L2. The results also indicate a significant difference in mean error scores at different proficiency levels. Thus, the beginner level students produced the highest number of errors in targeted segmentals than students from intermediate and advanced levels. These findings support the idea Ringbom and Jarvis’s (2009) stated that the second language learners rely more on their previous linguistic knowledge at an early stages of second language acquisition. This especially applies to the languages that share a significant number of cross-linguistic features in cognate, lexical, phonological and writing systems, as in case of English and German.
The results of this study are also in line with previous findings of Piske (2008), which suggested that adult learners are influenced by the writing systems of their L1. A considerable number of errors were observed in producing in seven out of eight German segmentals [ç, f, ts, Êƒp, Êƒt, z, ts] with the exception of [v] which maps to Ë‚wËƒ in German. For example, the highest number of errors students produced was the German consonant Ë‚gËƒ in word final which maps to palatal fricative [ç] when preceded by front vowels as in traurig and consonantal cluster Ë‚chËƒ in word final which also maps to palatal [ç] in words such as Milch, natürlich, sportlich. Students produced these two sounds as [k]. They also confused German Ë‚vËƒ which maps to [f] with its English counterpart [v]. The same was observed with the other two German consonants Ë‚s,zËƒ which map to [z,ts] respectively and were pronounced as English phonemes [s, z]. This can be explained by the fact that all German consonants mentioned above are represented by the same graphemes in English, which resulted in students mispronunciations. Interesting results were obtained from the words that contained labio-dental [v] which in German maps to Ë‚wËƒ grapheme. None of the students had errors producing this sound despite the fact that the same grapheme /w/ maps to an approximant [w] in English. This could be interpreted as following: it appears easier for students to produce sounds that are far apart in terms of their place of articulation than for instance, the sounds that vary only in their voicing. Thus students in this study had difficulty to devoice fricative [f] in German because it is represented by grapheme Ë‚vËƒ, which in English maps to voiced fricative [v]. Participants of this study also struggled producing consonantal clusters Ë‚ st, sp Ëƒ which in German language map to [Êƒt, Êƒp]. Although English has phoneme [Êƒ], it exists in different phonetic environment and is never followed by [p,t] phonemes in onset. Therefore students confused these two consonantal clusters with their English counterparts and produced them as [sp] for example English ‘sport’ or [st] as in English ‘student’.
Lastly, the perception task yielded somewhat ambiguous results. It appears that there might be a relationship between the error score and a number of students rating the words as dissimilar. Nonetheless, this needs to be investigated in the future research using a larger sample as there was insufficient amount of data to strongly support the findings. For more details regarding error score and the students’ perceptions see Table 6 in Appendix A.
The purpose of this study was to find out to what extent English-speaking learners of German rely on the English orthography and whether the presence of cognate words results in mispronunciation in the target language or vice versa helps learners to acquire L2 phonology. Given what was found (a) evidence that English orthography, namely its grapheme-phoneme rules interferes with learners’ pronunciation of German segmentals, (b) evidence that learners were influenced by L1 writing system mostly at the beginning stages of second language acquisition, and (c) some evidence of potential relationship between the errors produced and students’ perceptions of cognate status, we might consider some practical implications for foreign language classroom as well as future research implications.
As Hayes-Herb et al. (2010) suggested language teachers might find it beneficial presenting new vocabulary using initially auditory input only. This may particularly refer to the non-cognates words as students in this study made more errors in words that they rated as dissimilar with English. At the same time, it may be helpful for learners to receive written input of the words that contain German Ë‚wËƒ in word initial, as students produced zero errors, which might suggest that in this particular case written input positively influenced learners’ pronunciation. Therefore it is important for language instructors to be aware of orthographic transfer and its positive and negative effects so that they can address this information accordingly in their classroom. On the other hand more research is needed to support these findings. For this purpose a more comprehensive study with a larger sample would help to expand the current analysis of data and make the findings generalizable for a larger population of students studying German as a foreign language. Teacher surveys might reveal whether language instructors focus on orthographic transfer or is it something new to them as well as to what extent they focus on pronunciation of those specific sounds that might be problematic due to the reliance on L1 sound-letter correspondence rules in their classroom. The results can provide valuable insights that can be applied to the teaching of German as a foreign language.
Furthermore, given the scope of this study only eight German segmentals were investigated. Examining other segmentals whose orthographic representations exist only in German (for example, Ë‚¢Ëƒ, Ë‚ü, ö, ä Ëƒ) or segmentals whose sounds exist only in German but are represented by the same graphemes (for example German Ë‚rËƒ which maps to several phonemes [Ê€, Ê, r] depending on the in word position) would clearly provide better results on the way English and German orthographies interact with German as a foreign language phonology.
To conclude, while the this study attempted to address questions on how orthography of L1 and its sound-letter correspondence rules can be transferred to L2 phonology resulting in mispronunciation of certain segmentals still more research on this topic is needed.
Cite This Work
To export a reference to this article please select a referencing stye below:
Related ServicesView all
DMCA / Removal Request
If you are the original writer of this essay and no longer wish to have your work published on the UKDiss.com website then please: