Prosodic Features (Electroacoustic Analysis) of the French Language in South India

Lagutina Anna PhD in Philology specialty "Romance languages" 108814, Russia, Moscow, Linden Park str., 5 LagutinaAV@mail.ru Other publications by this author

Lalova Tat'yana PhD in Philology Associate professor at the Department of Russian and Foreign Languages of Russian University of Transport (MIIT); 127994, Russia, Moscow, Obraztsova str., 9, building 9 t_lalova@mail.ru Other publications by this author

To date, one of the most populous countries with a population of about one and a half billion people is the Republic of India.  A huge state located in South Asia is of interest from the point of view of various scientific fields. One of these areas, of course, is linguistics, since this region is known not only for its history and archeology, but also for the combination of many cultures and languages existing here.

The origins and diversity of modern languages and dialects common in India, the nature of the culture and mentality of the inhabitants of a given country or its separate region can cause various kinds of features that arise when they study foreign languages as a result of interference "Interference is usually understood as a failure when switching from one language code to another, in which the transfer occurs the rules of one language system to the rules of another, which entails a violation of the norm in the receptor language" [Rebko A.V., 2018], when the rules of one language or dialect, which is native to the learners, are transferred to the rules of another, being studied. In this regard, during our experiment, studying prosodic ways of implementing text semantics, conclusions were drawn that can be considered reliable for French language learners from a certain region of southern India – the state of Tamil Nadu.

The recording of the analyzed text was made by a student from Chennai, one of the largest cities in the country and the capital of the specified state. 

In this study, the task was to analyze the pronunciation characteristics of Indians studying French in this region in terms of the ways of intonation of the semantics of texts using the main prosodic characteristics (pitch frequency, intensity, duration and pausing).

The experiment described in this article was carried out using the PRAAT software, which allows you to calculate the absolute values of the main acoustic characteristics of speech.

Using the example of the text under study and its transcription, we will present the course of the instrumental analysis carried out.

Avant de commencer le cours de fran?ais en ligne, j’?tais un peu nouveau, je pensais que ce serait difficile ? comprendre, car je n'avais jamais eu d'exp?rience comme ?a. C'est la premi?re fois que j’ai un cours en ligne ? l’Alliance fran?aise.  Mais apr?s le d?but du cours, c’?tait facile ? comprendre. Je travaille bien. Maintenant, j’aime les cours de fran?ais en ligne. ?a marche bien. J’ai un tr?s bon professeur. Elle parle bien le fran?ais, elle m’apprend tr?s bien, elle m’explique tr?s bien les r?gles de grammaire. Elle r?pond ? toutes mes questions et d?cide mes doutes sur le sujet. Et elle donne les exercices sur mon alliance et je l’ai fait tous les jours. C'est un moyen s?r d'apprendre le fran?ais en ligne depuis chez soi et de rester ? l’?cart du coronavirus.

[a-`v-d?-`k?-m-?se-le-?ku:? -d?- f-`s?--?li?|?e-?t?- -?p? -nu-?vo|-p-?s?- k?- `s?-s?--di-fi-?si-la-k-?pd?|?ka?--`na-v?-`?a-m?-?y-d?ks-pe-j:s-k?m-?sa||s?-la-`p-mj-?fwa-k?-e--?ku:--?li?-a-la-?lj:s-f-?sz||m?-`a-p-`l?-de-?by-dy-?ku|se-?t?-fa-?si-la-k-?pd?||-`t?a-vaj-?bj||m-?tn-m-le-?ku:?-d?-f-`s?--?li?||sa-`ma?--?bj||`?e--`t-b-`p-f?-?s||?l-`pa?-l?-?bj-l?-f-?s?|?l-ma-?pr-t-?bj||?-l?ks-?plik-t-?bj-le--?l?-`d?-a-?m||?l-?e-?p-a-tu-`tme-k?s-?tj|e-de-?sid-me-?dut-sy?-l?-`sy-||?l-?d?n-le-`z-z-?sis-sy?-`m-na-?lj:-se-`-le-?f?-tu-le-u||s?-t-`mwa-j-?sy?-da-?p-d-l?-f-?s?--?li?-d?-`p?i-?e-?swa-e-`d?-s-?te--le-?ka: ?-dy-`k?--`na-vi-ys||]

The meaning of the above text is conveyed by contrasting semantic lines: the difficulties faced by auditors in the course of learning French are contrasted with the results achieved, the available methods of teaching and the positive assessment of the work of teachers. After analyzing the meaning of words and combining the vocabulary of the specified text into two semantic complexes expressing semantic lines, we deciphered the sound file taken and calculated the values for four prosodic characteristics of speech: pitch frequency (CHOT), duration, intensity (amplitude of sound pressure) and pause.

One of the main acoustic characteristics is the pitch frequency. According to linguists L.P. Blokhina and R.K. Potapova, frequency indicators "act as the leading ones in the general complex of acoustic parameters" [Blokhina, Potapova, 1982, p. 4].

In order to determine the average pitch frequency level in hz, the following formula was applied [Blokhina, Potapova, 1982, p.15]:

(? CHOT without.sl.)/n = CHOT cf.

here - ? CHOT without. sl. – the sum of the indicators of the values of the CHOT of all unstressed syllables,

n – represents the total number of unstressed syllables,

CHOT cf. – the average frequency level. This indicator acts as a normalizing unit necessary for the conversion of absolute values (hz) into relative ones.

The normalization consisted in comparing the magnitude of the highest frequency level on a particular word or a group of words united by one meaning with the average.

Next, let's move on to the definition of duration – one of the key characteristics of sounding speech. To identify reliable results, the authors of the study normalized the available absolute values. The average sound duration calculated according to the formula was taken as a unit of rationing [Lalova, 1986, p. 97]:

(?t ph)/n = t cf.

here - ?t ph is the total duration of the spoken sounds,

n is the number of pronounced vowel sounds,

 t cf. – the average sound duration.

The rationing process consisted in comparing the taken duration values (in msec) with a medium-sound duration.

Sound pressure or intensity is an excretory characteristic of speech. The procedure for determining the normalization of this indicator was carried out according to the following formula [Lalova, 1986, p. 99]:

(?w_ph)/n = Wsr.

 here - ?w_ph is the total intensity of vowels and sonants,

n is the number of pronounced vowels and sonants,

W cf. – the average sound pressure level.

The average sound pressure level is defined as a normalizing unit, the ratio to which was calculated for all analyzed sound pressure values.

The next stage of our research is the analysis of pausing – an additional means of highlighting vocabulary. Within the framework of this experiment, intra-phrasal pauses located inside the syntagma and inter-phrasal pauses taking place at the end of completed semantic units were noted. The results obtained in absolute values (msec) were correlated with the duration of a single pause with a normalizing unit. These calculations were carried out according to the following formula [Lalova, 1986, p. 101]:

(?P)/n = Pcr.

here - ?P is the sum of the duration of all inter–phrase or intra-phrase pauses,

n is the number of inter–phrase or intra-phrase pauses,

Pcr. – the average pausal duration, which was taken as a unit of rationing.

Next, the calculation of fluctuations from the average values of the main acoustic parameters, which determine the words included in the lexico-semantic complexes taken for research, was implemented.

Then the total rate of speech was calculated, measured by the ratio of the total utterance time to the number of linguistic units, as well as the absolute rate, determined by the ratio of the time spent on utterance minus pauses to the number of linguistic units.

The table below shows the results of the acoustic data analysis (Table 1). The first column lists the lexical units of the semantic complex taken by the authors. In the second, the results of the ratio of the duration of each pause to the normalizing unit are demonstrated. In the third column, the data of the correlation of the sound duration and the normalizing value are marked. The fourth column shows an increase or decrease in the pitch frequency of a certain word in relation to the average value. In the final – fluctuations in sound pressure. All the data obtained are presented in relative units.

           Table 1

The replica in question contains 21 lexical units that are part of two lexical-semantic complexes that are opposite in meaning: the difficulty of learning French and the positive results achieved during training.   As can be seen from the above table, the frequency of the main tone is marked by 21 lexical units, intensity – 21, pause – 3, duration – 21.

The results of the analysis indicate that the lexical units with the greatest deviation of the pitch frequency from the average are defined as: avant (1.63 times), mais apr?s (1.62 times), ? l'?cart du coronavirus (1.58 times), les exercices (1.57 times), jamais (1.57 times times).

The maximum deviations from the average sound pressure were revealed in the words: tous les jours (1.14 times), la premi?re fois (1.13 times).

The greatest amplitude of duration relative to the average value is observed in the following lexical units: tr?s bien (2.61 times), toutes (2.07 times), mais apr?s (1.98 times).

The duration of the pause, as much as possible exceeding the average, is marked before the tokens: chez soi (0.50 times), ? l'?cart du coronavirus (1.24 times), un tr?s bon (0.14 times).

As a result of the analysis by the PRAAT software, the total speaker's speech rate was 0.3017 units, the value of the absolute speech rate was 0.2400 units.

The data of the performed instrumental analysis of the text under study indicate that the marking of vocabulary, which is part of two semantic complexes opposed to each other, is carried out in most cases by means of the frequency of the main tone and duration. A pause usually acts as an additional excretory agent, while complementing other prosodic characteristics of speech.

Nevertheless, the role of duration as a secondary means of intonation is the least informative, because it is associated with the presence of excretory stress in the studied vocabulary layer.