Identification of Thai students’ level of phonetic sensitivity in the imitation of Russian syllables

. The article deals with the role of phonological sensitivity in the development of skills in a foreign language. In psycholinguistic terms, verbal communication in non-native language is a language contact. It is important that, in learning environments, the interaction of contacting languages should not be spontaneous, it should be taken into account in the modeling of verbal communication. Since the formation of mechanisms that ensure the speech activity of an individual in the language under study occurs under the influence of interference, it is necessary to study the manifestation of the specifics contacting linguistic systems in the perception of speech in a non-native language. To make the right decision – whether or not an incorrect perception of the phonemic composition of words occurs in the natural conditions of communication in a foreign language – is not possible. This requires a special phonetic experiment. The authors describe the results of an experiment aimed at revealing the phonetic sensitivity of native speakers of Thai. The material of the experiment was 260 syllables having the structure CV (consonant + vowel). Subjects who had not previously studied the Russian language had to listen to audio recordings of syllables pronunciation and reproduce them. Thus, the implementation of Russian vowels and consonants in the composition of syllables was analyzed. The authors succeeded in revealing the similarity and difference in the articulatory characteristics of the interacting languages’ sounds. It was found that when imitating Russian syllables the Thais are not aware of the opposition of consonants on the grounds of "hardness-softness" and "voiceless-voiced". In this connection special work is required to produce soft sounds and to develop skills for distinguishing sounds from voiceless-voiced. During the study, it was confirmed that the influence of the native language is clearly manifested if it is functionally prevalent.


Introduction
In the 1990's an active study of the role of phonological sensitivity in the development of reading and spelling skills began, which led to a wide application of this term by many researchers in various fields: education, speech therapy and psychological clinical practice [9]. sensitivity" are equivalent and denote the individual's ability to distinguish, identify, and encode the heard sounds of a foreign language. In our opinion, these terms are narrower than "linguistic sensitivity," using of which, as applied to the phonetic aspect of communication, is not entirely justified. To clarify the term "phonetic sensitivity" we turned to Englishlanguage sources. It has been found that researchers use two terms: "phonetic / phonological sensitivity" and "phonetic / phonological / phonemic awareness," with preference given to the term "phonological awareness." In the work of G. T. Gillon, it was noted that the term "phonological awareness" appeared in the academic literature in the late 1970's and early 1980's and was interpreted as the individual's comprehension of the sound system or the phonological structure of the sounding word. This term was used to describe the process of forming reading skills among children. In earlier works it was argued that understanding the sound system of a word enables children to encode (or pronounce) a printed word [4; 6; 7; 15; 16 et al.]. Then in 1980, A. J. Marcel was one of the first to establish a direct link between phonological sensitivity and spelling error, proving that an individual who tolerates such errors has a low level of phonological sensitivity [18].
However, an opinion about the difference between the terms 'phonological awareness' and 'phonological sensitivity' could be found. In the article "A Developmental Continuum of Phonological Sensitivity Skills" Lisa A. Pufpaff notes that the term 'phonological sensitivity' was introduced by K.E. Stanovich (1992) to describe a set of skills related to phonological processing that is used in the reproduction of speech sounds [20,679]. K. E. Stanovich believes that "phonological sensitivity should be viewed as a continuum from 'deep' sensitivity to 'shallow' sensitivity. Tasks indicating deeper levels of sensitivity require more explicit reports of smaller sized units [23, p.317]. In other words, the author believes that phonetic sensitivity should be viewed as a continuous command from "deep" sensitivity to "superficial". A deeper level of sensitivity means having accurate information about the simplest units. Thus, according to the author, the term 'phonological sensitivity' has a wide meaning, embracing also 'phonological awareness', since 'phonological awareness' most often means only the ability to identify and reproduce the heard sound (allophone), but not to reveal the differential signs of the phoneme. In the following discussion, we will use the term "phonetic sensitivity" to denote the individual's ability to produce phonological processing of the heard sound and implement it in accordance with the results of such processing.

Subjects
To reveal the level of phonetic sensitivity of the Thais, we conducted an experiment in which 10 native speakers of Thai language (5 men and 5 women) took part. The age of the participants is from 18 to 28 years.
All subjects are residents of Bangkok. 5 subjects work in areas not related to the usage of foreign language, while others are students of non-linguistic universities.
All participants in the experiment have certain command in English as their first foreign language, which is varied from A2 to C1 level of English as foreign language. One subject speaks fluently not only English but also Japanese and is currently studying German.
In order to obtain reliable results, the Thais, who had not previously studied the Russian language, were specifically selected for the group of subjects.

Experiment Procedure
The experiment was conducted distantly. All participants in the experiment were sent an electronic audio recording of Russian syllables' pronunciation. The audio was supplied by an instruction in Thai: "Listen to the syllables and repeat them after the announcer in pauses. Record your pronunciation using the voice recorder of your mobile phone. You can listen to each syllable only once". Then the testers had to send their audio recordings to the experimenter by an e-mail.

Research Methods
During the experiment, the simulation was used with no written support. According to a number of researchers, an imitation, being understood as "the repetition of sounds, words, phrases after a teacher or announcer in sound recording" [1], is an almost ideal way of obtaining a large volume of phonetic information, since under these conditions the speech behavior of an individual is more automated compared with the letter record heard [5,10]. An obvious advantage of this method of research is that the recipient is not "imposed" by an artificial phonemic classification [21]. Important is the dependence of the simulation results on the interaction of the sensory and motor levels [14, 19, etc.].

Findings
Thai-Russian interference is clearly manifested at all levels of the Russian language's phonological component, i.e. on segmental as well as on suprasegmental. This is due to the typological dissimilarity of Thai and Russian.
An analysis of the results of our experiment shows that Thai subjects had difficulty reproducing Russian sounds. Violations of the differential and integral phonemes' signs were detected.
Let us discuss the results of auditory phonetic analysis of the Russian syllables' imitation by speakers of the Thai language.

/ɑ/
Position /a/ in combinations with prepositional hard consonants does not cause particular difficulties for Thais. However, in combination with sizzling /ž/, / š':/ and affricates /č'/, phonetic disturbances (3.6% of the total number of syllables) are found. In these combinations, the phoneme allophone [a] was recognized as an open vowel in the front row of the middle ascent [ae] (3.2%) and as a diphthong [ua] (0.4%).
In the position after soft consonants, systemic violations of the vowel / a / are noted, since in this position it is characterized by diphthongoidity (heterogeneity) -[ia], which is not characteristic of the Thai phonetic system. However, according to the experimental data, there is a possibility that the Thais can normatively realize this sound (3.3% of the subjects were reproduced by the diphthongoid vowel). But in most cases allophone  It is known that in the absolute end of the Russian syllable / word in the vowel /ɨ/ appears an [i] -shaped sound. It could be assumed that in this position the Thais will pronounce the diphthong [ɯi], however in our experiment such implementations were not noted. We assumed that in the position after soft consonants the utterance of the allophone [iu] would be difficult for the Thais. However, despite the fact that diphthongoidity is not inherent in the Thai phonetic system, cases were noted when subjects correctly implemented this diphthongoid vowel in syllables with sonants /n'/ and /m'/ (6.5%). Most often, the allophone The analysis of the results of the phonetic sensitivity test of Thais during the imitation of Russian vowels in syllables is summed up in Table 1. It was expected that the greatest difficulties for the subjects would arise in the implementation of allophones of Russian vowels combined with soft consonants. In some positions deviations from the norm of vowels' pronunciation were noted. Another problem was the non-discrimination of articulatory similar sounds, for example, in syllables with sibilant consonants: /š/ and /ž/, /š/ and /š':/, /č'/ and /c/. Most of the participants could not implement hard /r/ and /l/. In this regard, it can be argued that the phonetic disturbances in the imitation of Russian sounds by the Thais were manifested in the implementation of both vowels and consonant sounds, which ultimately led to the incorrect pronunciation of Russian syllables.

Consonants
In most cases, the consonants of the Russian language were replaced by the corresponding Thai sounds. An analysis of the experiments' results shows that the most problematic for the Thais sounds are hissing, back-lingual and all soft consonants. When the soft phoneme /p'/ was realized, more variants were noted than when the solid /p/ was pronounced. A weak palatalization was observed in the reproduction of this phoneme (62%). 18% of the subjects realized the corresponding hard consonant, 7% -hard voiced [b], 2% -soft [b']. It should be noted that the soft /p'/ was pronounced correctly in 11% syllables, which means that with the required amount of training exercises, Thais can learn the articulation of this sound.

Libial stops noisy consonants
Errors of Thais in the imitation of lip stoppers are associated with a violation of the differential signs of these consonants: hardness-softness, voiceless-voiced, as well as aspiration of a voiceless hard / ph /, characteristic of the corresponding Thai sound (see table  2). Phoneme /f/ subjects were correctly implemented in 87% of cases. Minor impairments are associated with the realization of a palatalized sound (6%) and English labial-dental sound /v/ (7%).
Pronouncing the voiced /v/ and /v'/ caused considerable difficulties for the participants in the experiment due to the lack of sounds in the Thai language. The majority of the subjects realized the English dental [v], the peculiarity of articulation is that the tongue is rounded, its tip touches the upper edge of the lower teeth  Table 3). The greatest number of deviations from the phonetic norms of the Russian language is fixed in the articulation of sonorous /v/ and /v'/. Thai bilabial [w] is realized instead of the fricative labial-dental [v] both in English and in Russian speech. Probably those subjects who have developed the skills of pronouncing the English lip-dental [v], could reproduce it in place of the Russian labial fricative sound.

Front stops noisy consonants
The phoneme /t/ was realized by the corresponding Thai sound in 52% of cases. Pronouncing [d] instead of /t/ (26%) indicates the non-discrimination of these two consonants with native speakers of Thai. The phoneme /t/ was also represented by the Thai front aspirate [th] (12%), the mid-tongue occlusal [ʨ] (6%), the bilibial [w] (2%) and the fricative [f] (2%) (see Table.

4).
In place of the phoneme /t'/, the subjects reproduced the mid-lingual stopper /ʨ/ in 71% of cases. Only 6% of cases were noted when the Thais realized the sound It should be said that in accordance with the symbols of the international phonetic alphabet, the Russian voiceless soft affricate has the same designation as the Thai mid-lingual stopper /ʨ/ (จ). In our experiment, 71% of the subjects on the site of Russian affricates /č'/ implemented a half-lingual slit [ʨh]. The rest of the participants pronounced Thai apical whistling [s] (17%) and averaginal blind ( [11]), and, for unclear reasons, Central European sonant [1] (1%).
The reason for the implementation of the mid-lingual stopper /ʨ / in place /č'/ is the absence of palatalized /t'/ and /š'/ in Thai, so subjects could perceive mild /č'/ as based on "false similarity". In order to realize the given consonant, subjects pronounced a sound similar to them, i.e. also closed, but medium-lingual. Articulatory descriptions of Thai /ʨ/ and Russian /č'/ are fundamentally different from each other. N.A. Lyubimova characterizes the affricate /č/ as follows: "consonant /c'/ is in Russian an antero-lingual double-focus voiceless soft affricate" [17, p.119]. B. Aymongkhon describes a similar Thai sound as a "non-voiced and unaffected" semicolonical consonant /ħ/ [2, p. 67]. According to the classification of Thai consonants, proposed by M.R. Kalaya Tingsabadh and Arthur S. Abramsom, consonant [ʨ] -"post-alveolar affricate", i.e. post-alveolar affricate, which is opposed to the consonant [ʨh] only by the presence-absence of aspiration [25, p. 25]. The difficulty in determining the articulatory properties of a given sound can be explained by the reason for its realization by the Thais instead of the Russian stop, slit and affricate.
According to Aymongkhon B. [2, p.88], due to the absence of the Russian sonorous hard whistling /z/ in the Thai language, the subjects correlated this Russian phoneme with the English-speaking frontal slit agreeing to it (31%). However, most often the Thai apical nonpolarized sound [s] (66%) was realized. In addition, a mid-lingual stopper /ʨ/ (3%) was also introduced.
Simulation of the phoneme /z'/ caused certain difficulties for the subjects. This phoneme was realized either as a weakly palatalized voiced consonant [z'] (10%), or as a velarized [z] (16%). In the remaining 19% of cases, the phoneme was represented by the Thai apical whistling /s/.
Instead of the phoneme /š/, most subjects produced a Thai-language slit [ʨh] (70%), due to the lack of sizzling in the Thai language.  As you know, in the Thai language there is no hissing phoneme /š'ː/. In most cases, it was implemented as a Thai-language slit [ʨh] (49%). Instead of the Russian soft hissing long, the subjects pronounced the apical sound [s] (31%) and the Thai mid-lingual stop [ʨ] (20%). This is due to the non-distinction between single-focus and two-focus consonants.
The soft phoneme /g'/ in most cases was also realized by the stopper deaf [k] (62%). In addition, a sonorous weakly palatalized posterior-lingual consonant was found (38%).
Aymmongkhon B., referring to the opinion of A.V. Ventsov and A. Martine, write about the difficulty for the Thais to realize the voiced back-lingual abrupt sounds: "the formation of the binder-collar bones in comparison with the front-lingual and labial ones is closer to the glottis. Therefore, at the time of the formation of their bows, the resonator is filled with a smaller amount of air and this prevented the difference between intraoral and subglottic pressure, which is necessary for the realization of the voiced consonant" [Cit. on 2, p. 94]. The absence of a slotted posterior phoneme /x/ in Thai means the fact that the subjects were pronounced Thai pharyngeal [h] (50%). In other cases, the Thais realized explosive stoppers: aspirated [kh] (44%) and non-respiratory [k] (4%). At the same time, 2% of the subjects were able to pronounce the Russian back-lingual fricative [x].
The phoneme /x'/, also unmatched in Thai, was most often realized as a pharyngeal [h] (82%). 18% of the participants in the experiment reproduced the aspirated aspirate [kh].

Sonants
As expected, the phoneme /m/ did not cause any special difficulties for the subjects, because in Thai there is a corresponding consonant, so they realized this sonant correctly in 91% of cases. 9% of violations -the implementation of a labial sonant with weak palatalization (see Table 8).
Despite the fact that the soft phoneme /m'/ is absent in Thai, 11% of its normative realizations are noted. In most cases, /m'/ was realized as a slightly palatalized sound (73%). 16% of the subjects reproduced non-palatalized [m].
Pronouncing the nasal anteroposterior /n/ also was not difficult for many participants in the experiment. This sonant was replaced by the corresponding Thai consonant (82%), which, according to acoustic signs, almost does not differ from the Russian nasal sonant. The realization of /n/ as a weakly palatalized consonant was found (4%). In other cases, reproduction of the labial sonant /m/ (13%) was noted, which indicates the possibility of nondecreasing the phonemes /n/ and /m/.
With the implementation of the phoneme /n'/, more variants of utterance were recorded. The soft nasal sonant /n'/ was realized mainly as a poorly palatalized consonant (51%), nonpalatalized (9%) and labial sonant [m] (11%). It is important that soft /n'/ has a greater number of correct realizations than other soft consonants (18%). In addition, some subjects pronounced the mid-lingual sonant [j] in n'a-n'u syllables (11%).
The apical velarized /ɫ/ was reproduced as the corresponding Thai slotted lateral sonant [l] (63%). In 31% of cases, a trembling sonant [r] was realized. In addition, the pronunciation of retroflex slit English [R] (6%) was found.  Our experiment confirmed the idea that the source of phonetic errors in a foreign language can be a previously learned non-native language (in our case, English).