Phonetic Acoustic Analysis for Faifa Dialect of Saudi Arabic - Coursework Example

PHONETIC ACOUSTIC ANALYSIS FOR FAIFA DIALICT OF SAUDI ARABIC The Arabic language has been rated as one of the most spoken languages in the world in terms of the number of speakers. The language has not received much attention especially from the research community. The essay is specifically concerned with acoustic analysis with relevance to the vowels in Arabic dialect. The first and second formants are analyzed from the Swedish list. During the analysis, the instrument used is the Praat software. The Praat is extensively used to analyze and understand the features of the phonetic vowels. The generated data is the equally analyzed to provide a better understanding of the frequency domains and time. The findings of the research can be used in future research for Arabic speech processing tasks that can involve speech classification, vowel and speech recognition. Arabic language is a Semitic language that is considered as the oldest languages in the world (Versteegh, 2014). Arabic language is also the second most spoken language in the world. There are 36 phonemes in the Modern Standard Arabic, out of which 28 are consonants; two are diphthongs and six are vowels. The first three vowels in Arabic are known as short vowels while the three last vowels are long vowels. Consequently, the duration of the vowel sound in Arabic language is phonemic. The first three vowels are /a, I, u/ while the last three vowels are /a: i: u:/. Most researchers have argued that there are eight vowels in the Arabic language. This is as a result of counting the two diphthongs as vowels especially in the Modern Standard Arabic. Arabic phonemes are made up of emphatic and pharyngeal phonemes. These two classes can only be found in Persian and Hebrew Semantic languages. Arabic dialects have been found to comprise of different vowels, for instance, the Egyptian dialect has extra vowels while the Levantine dialect has an extra two diphthongs; /aw/ and /aj/. There are three non-emphatic s-sounds in the Semitic language. One of the sounds in relation to the cognates is a lateral fricative. The modern Arabic language has three sibilants namely; lateral sibilant, an alveopalatal sibilant, and dental sibilant. During the analysis, I was able to generate the following spectrograms /s/, /š/, /s/. The sound files were generated by a female speaker aged 35 in a single recording session. The beginning of the sibilant occurs to the immediate right of the dotted line inserted in the spectrogram. The s sound as in hiss in the spectrogram has the highest pitch as demonstrated in the darkest part of the spectrogram and its high frequency. Sound /š/ in the spectrogram has the lowest pitch as demonstrated by the lowest frequency at the beginning of the darkest part of the spectrogram that corresponds to the consonant. The Modern Arabic languages have an emphatic. The emphatics is evident in lateralized pharyngealized fricative in most Arabic dialects of a male speaker (Munro, 1993). Certain words from the Arabic dictionary are unique and hence have etymological orientations that affect their pronunciations. A word such as dad can either be pronounced as voiced pharyngealized interdental fricative or a voiced pharyngealized dental stop. The reason is because Arabic was originally described as lughat al-dad, this made most grammarians to consider the sound as unique in the Arabic language. There are several reasons for one to believe that the early dad was realized as either pharyngealized consonant, lateralized verbalized or lateral. The grammarians were very careful in the articulation of the word ensuring that edge of the tongue was due to the molars hence giving a lateral articulation. Previous research on the same has pointed out that the correct articulation of the word in the eighth century as the onset of the tongue and the support of the molars in place of articulation (Newman & Verhoeven, 2002). Arabic is a less researched language compared to other languages such as English, Chinese and Japanese languages. Most research studies that have been conducted on the language has focused on vowels. Most studies focus mostly on Quranic and classical versions of Arabic. Recent studies reported a preliminary study on identification and segmentation of formant transitions that occur in the Quranic Arabic transitions (Newman & Verhoeven, 2002). Procedure Praat Praat was used in speech analysis. The software was used labeling and signal edition. The software is also used for acoustic, duration and formant manipulation and analysis. The software can also be modified for certain specific research purposes. The results can be exported to XL-spreadsheets. Measurements of formants The aim of the research is to measure F2 and F1 because they present certain acoustic properties that can be easily seen in the spectrograms. The vowel quality can be classified and determined from the spectrogram. The first impression was to look at the spectrogram of the stimuli, especially the target vowel. Formants tracks were then automatically computed for the following three formants (F3, F2, and F3). The three formants have a frequency that range from 0 to 3200 Hz and then superposed on the spectrogram. In case of visual mismatch in the spectrogram and the formant tracks the frequency range in the LPC and the model order until the satisfactory match is generated. A segmentation point is then set in a text grid at the offset and the onset of the target vowel and realized the number of formants that can be extracted and a frequency on a different tier. While using a script, the formant and the duration frequencies are then extracted from the offline recordings. The formant values were then generated at the midpoint temporal of the vowel target. The set data was then later analyzed using SPSS. Normalization of vowel The procedure of z- normalization was then applied to the Bark-transformation of F2 and F1 values of native speakers of Arabic. Normalization of a vowel is a statistical operation that was developed to compensate certain specific differences in vocal tract size that results in various resonances of formants. Normalization of the vowel is very important so as to compare different vowels realizations by different speakers. Generally, comparisons include; vowel classification, durations, and formants. In the recent study, normalization is normally used to preserve phonological distinctions among different speaking groups. Normalization can also be used to achieve a crucial improvement of the acoustic output of F2 and F1 on the basis of the z-transformed scores (Munro, 1993). Measurement duration The measurement duration is a difficult task. This is due to the delimitation of units of sound in an acoustic sense that requires segmentation of utterances in which auditory quality and different productive impressions of sounds can make the task of such impression quite complex. Even when it is possible, the duration provided cannot correspond to the linguistic conclusion of length; instance in the long and short vowel. The statistical analysis of acoustic output shows that the dispersion of the English vowels and that is spoken by Arabic speakers applies different distinction categories. Fascinating evidence reveals that English tense-lax vowels pairs are normally closely positioned in the vowel space. Assessment of Intelligibility After the vowel measurement, the data is then compared with the study that was conducted Barney and Peterson in 1952 (Munro, 1993). In the Peterson and Barney study, the dialect analyzed was then referred to as the General American English. For one to record all the vowels for the project, a standard digital voice recorder was used in the study. All the words were then recorded on a hard disk. The recorded data in a mp3 format was later converted into a WAV format with the program known as Audacity. Audacity is a software that can be downloaded in the internet. The recordings were then opened using Praat. As discussed earlier, Praat is a software that can be downloaded from the internet. The acoustic analysis of the vowels was also done using Praat. After finding the F2 and F1 measurement for each vowel, a thorough analysis of the vowel was then made with the data in Blue belonging to the speaker. Readings of Formant A perfect tube can be produced if the vocal tract shape is changed. Consequently, it can be used to change the frequencies of the vibration. Each of the vocal tract frequencies are referred to as the formant (Newman & Verhoeven, 2002). F3 indicates the third formant, and F1 represents first formant while F2 representsthe second formant. The positions of the formant can be changed by moving the lips and the tongue position. When dealing with vowels, F1 can vary from 300Hz to 1000Hz. The lower the frequency, the closer the tongue to the mouth roof. F2, on the other hand, can vary from 850Hz to 2500Hz. The F2 value is proportional to the backness or frontness of the highest part of the tongue when producing a vowel sound. Rounded lips also cause a lower F2 than unrounded lips. F3 is also very significant in establishing the quality of the phonemes in a given speech sound. Other formants that are higher such as the F4 and F5 are very significant in establishing and identifying quality (Munro, 1993). The common syllables in Arabic language are consonant-vowel (CV) consonant-vowel Consonant and (consonant vowel consonant consonant). Arabic consonants only begin with the consonants. The paper aims at analyzing the Arabic vowels formants which will later prove helpful in speech processing tasks and speech recognition tasks. Arabic vowels have always been viewed as patterns. There are important things to note in regard to the intelligibility of the speaker. If an Arabic-speaking individual of a Fiafi dialect were to pronounce the vowels, there would be different pronunciation as reflected below. The Arabic subject would have the high vowels lowered and back vowels centralized. The vowels of the Arabic individual also occur in three major clusters that are likely to cause confusion for those who listen to the speaker. The figure below demonstrates the three relevant clusters. The figure above shows comparison of pronunciation of both the subject and the Assumed Arabic faifa dialect. The data in blue belongs to the English speaking individual while the red color belongs to the assumed Faifa speaking individual. In the next assessment, the focus is on intelligibility that is caused by F1 measurements which takes place between the pronunciation of the assumed Arabic individual and the English Speaking individual. It is important to note that when considering F1 measurement, the median frequency range is 135 Hz. From the analysis, the Arabic individual is likely to have problems with the /I/ while the English speaking individual is likely to have problems with the /e/. The F1 of the Arabic individual are 5Hz and 6Hz apart from the English speaking individual. This means that a listener is likely to have problems identifying when the Arabic speaking individual is using words with /i/ sound. For instance, the sound “weight” may sound as “wit” or “wet.” Another vowel that is likely to cause intelligibility issues for the listener is /o/. The Arabic Individual /u/ is 19Hz different from the /o/ of the English speaking individual. Certain misunderstandings in pronunciations of words such as “show”, the word is likely to sound as “shoe.” Another location with the vowel of the Arabic speaker that may cause confusion is the vowels /e/ and/i/. The two vowels have an F1 that is quite similar to a British English peaking Individual. The Arabic speaking Individual /I/ is 7Hz and 19Hz. The Arabic speaking individual “wane” and “wean” will almost be analogous with the “win” in the British English. The final issue that is caused by the F1 measurement occurs between the Arabic Speaking Individual and the/a/ of a British speaking individual. The difference is only 26Hz. Even though the measurement is greater than the rest, there is a little difference for the ear to distinguish. The pronunciation is likely to cause difficulties in words such as “cot” and “caught” (De Jong & Zawaydeh, 1999). Key Subject: Arabic speaking Individual GAE: English Native speaker Even though the F2 measurement of a vowel does not affect intelligibility issues like the F1 measurement there are significant differences that should be noted. The mean frequency of F2 measurement is 311 Hz. The Arabic Individual /o/ is only 40 Hz more fronted than the British English or English speaking individual /^/. The subject /I/ is only 48 Hz more centralized than the British English or a native English speaker. The table below shows an analysis of measurements of the F1 and F2 of vowels of certain few words. Key Subject: Arabic speaking Individual GAE: English Native speaker Acoustic analysis Implications The data analyzed above provides implications for the teachers of ESL. The implications for Arabic speaker will be discussed below. The issues that were revealed by the Arabic subject individual during the research project is that the individual is a representative of the social network. The Arabic subject is surrounded by low proficiency non-native English speakers or Arabic speakers. The individual receives little or no input that is quite higher than his understanding or proficiency level. And has therefore not taken significant steps in ensuring that he expands his social network in order to include all the inputs that are necessary to build and improve his pronunciation. The Arabic subject requires optimal input level to improve his pronunciation. Optimal input level can be described as the language that is quite higher than the individual language level. If the individual was surrounded by Arabic speakers and English speakers whose proficiency level is not higher enough to meet the definition of optimal input level, then the individual is not receiving the necessary inputs to improve his pronunciation. His accent, therefore, hinders his intelligibility and comprehensibility that will later affect his communication skills. Even though a native kind of accent is not the main goal, the Arabic individual has a lot to do in order to gain the pronunciation that is largely intelligible and slightly accented. In order to better his pronunciation, the individual needs to take an active position in improving his social network. The individual can make this possible by participating in clubs, creating the friendship with native speakers and possibly applying for a job that will provide him with the opportunity to use not only the language but also practice pronunciation. Such efforts will provide him with a higher level of input, which will later improve his communication skills. It is also crucial to take a glimpse at the impact of the implication for teachers of ESL. Previously, the main focus of pronunciation was limited to attaining a native-like accent since there are vast Englishes that are spoken globally and hence the construct has been eroded. The current trend in SLA is to create a distinction in the three main constructs; intelligibility, accentedness, and comprehensibility. The description of the three constructs is conventionally agreed upon, and there is a slight difference in the word choice. Accentedness can be described as the degree in which the speech of the learner is perceived to differ completely from that of the native speaker or the native speech. Comprehensibility, on the other hand, relates to how challenging it is to the speaker to understand the speech of the learner. Intelligibility analyzes whether the message that was intended by the learner was received correctly by the listener (Munro, 1993). Previous research has shown that even with strong accents, certain learners are intelligible to the corresponding listeners. The major two constructs that greatly affect communication are comprehensibility and intelligibility. Due to this, more emphasis should be put on the two construct (intelligibility and comprehensibility) rather than putting much focus on native-like speech. With intelligibility as the main goal, teachers should try and find ways of integrating the opportunities for relevant communication in a social set-up. From the study, it is quite evident that the chances of a student using Academic English is quite minimal, however, when they occur, it only happens in a classroom set-up. There is the need to incorporate activities where the students are not only speaking the language but also gaining the correct level of input. Consequently, the students will gain opportunities and input needed in order to improve their English language skill and pronunciation in general. It is also possible to introduce and educate them on the ways on how they can improve and build their language outside the classroom, such as those discussed above for the individual subject. The learning of a language can be an individualized task where the individual learner may have specific needs that are quite different from other learners. Much focus has shifted to training the student to be autonomous by providing them with the skills and strategies to do so. A teacher is required to act more as a facilitator or a coach rather than just a mere class teacher. In order for the learners to gain from this teacher shift, the teacher should provide the learner with the relevant strategies that will enable the learners to carry out their learning process effectively outside the class (De Jong & Zawaydeh, 1999). A perfect example is the task-based approach to teaching. The approach is based on building authentic opportunities that enable the students to communicate outside the classroom. In most cases, teachers can connect with the general society and, as a result, set up relevant experiences in which the learners can effectively participate. Such experiences can involve going to seek for driving license in a local MDV, renting an apartment, discussing ways of preventing certain ailments with a medical professional. Holding discussions with a counselor regarding a health subject or a life subject. The above ways only provide ways on how to approach the construct of pronunciation both in and outside the classroom set-up. References De Jong, K., & Zawaydeh, B. A. (1999). Stress, duration, and intonation in Arabic word-level prosody. Journal of Phonetics, 27(1), 3-22. Munro, M. J. (1993). Productions of English vowels by native speakers of Arabic: Acoustic measurements and accentedness ratings. Language and Speech, 36(1), 39-66. Newman, D., & Verhoeven, J. (2002). Frequency analysis of Arabic vowels in connected speech. Antwerp papers in linguistics., 100, 77-86. Versteegh, K. (2014). The arabic language. Oxford University Press. Read More