ABSTRACT. Ukrainian-Russian speakers (N=21) residing in Poland with a level of proficiency in L3 Polish at A2-B1 (CEFR) were asked to read out Polish words containing the labiovelar approximant in different contexts resulting in 615 tokens across 5 contexts. The Russian language does not posses a labiovelar approximant. In Ukrainian, it surfaces only as an allophone of the labiodental approximant before back rounded vowels [u,ɔ] (Pompino-Marschall et al. 2017) but can also be devoiced before voiceless consonants (Buk et al. 2008). Other allophones of the labiodental approximant include a voiced bilabial fricative before non-back vowels which can alternate with a labiodental fricative favoured before front vowels (Žovtobrjux and Kulyk 1965). In the syllable-final position, it can be realized as a vocoid [u̯] (Rusanivs’kyj et al. 2004). In Polish, the labiovelar approximant has a voiced, voiceless and palatalized allophone. The voiceless allophone appears word-finally and between consonants of which at least one is voiceless (Dukiewicz and Sawicka 1995). The context for the palatalized allophone was excluded in the study. The hypothesis was that the L3 Polish labiovelar approximant may be problematic for the Ukrainian-Russian speakers as they only possess the sound as an allophone in one of their background languages. It was also hypothesised that the production of the L3 Polish labiovelar would be conditioned by its distribution in Ukrainian (limited to the context of back rounded vowels and before voiceless consonants with other allophones of the labiodental approximant in their respective contexts) and/or by its position in the L3 Polish word. The results show that the Ukrainian-Russian speakers produce the L3 Polish target labiovelar approximant in 61% of the analysed items. The other frequent realization is a velarized lateral which is found in 25% of the analysed words followed by a voiceless labiodental fricative (10%). The production of the labiovelar in L3 Polish does not mimic its distribution in L1 Ukrainian. First of all, the Polish {ł} is pronounced mostly by means of two sounds: the velarized lateral and the target labiovelar approximant. Other variants of the labiodental fricative in Ukrainian, which includes other allophones of the Ukrainian labiodental approximant, are not present in the data. Second of all, the occurrence of the labiovelar approximant is not linked to its distribution in Ukrainian as there are no differences of labiovelar approximant pronunciation across the different contexts. An analysis of the position in the word helps understand which contexts in the word contribute to a greater or lower correctness rate of the labiovelar approximant in L3 Polish. In the word-final pre-vocalic context, the correctness rate for the labiovelar is lower as the rate of the labiodental fricative increases. In the word-initial cluster, the velarized lateral is much more frequent whereas in a word-medial cluster before a voiceless consonant the correctness rate for the labiovelar approximant is much higher. However, the differences between the contexts are small, ranging from 10% to 20%, so the phonotactics cannot be treated as a conditioning factor for the pronunciation of the L3 Polish labiovelar. Another explanation for the variation in L3 Polish labiovelar approximant pronunciation may be based on graphemics. In Polish, the grapheme that stands for all pronunciations of the labiovelar approximant is {ł}. This can be confusing for Ukrainian-Russian learners as it is similar to the grapheme that represents the lateral approximant {l}. This rationale is further supported by the fact that the main non-target pronunciation is a velarized lateral which surfaces as the most common variant for the lateral among the same research group (author 2022). All in all, the production of the L3 Polish labiovelar by Ukrainian-Russian speakers is not conditioned by the distribution of the sound in their background languages but can partly be explained by its position in the Polish word and the graphemic similarity of {ł} and {l}.

ABSTRACT. A number of factors such as age of onset, gender, motivation and type of instruction have been shown to influence the phonological development of foreign language learners in instructed settings (see e.g. Gut 2009 for an overview). In addition, some studies have suggested that language attitudes and metaphonological awareness might also play a role in the development of foreign language pronunciation by classroom learners (e.g. Elliott 1995, Kopečková 2018, Saito 2019, Roehr-Brackin & Tellier 2019). However, no longitudinal investigations have yet been carried out that directly relate these two factors to the learners’ pronunciation of specific sounds of the target language. It is the aim of this study to investigate the relationship between both the learners’ attitudes and metaphonological awareness with the development of their pronunciation skills during the first year of learning a new foreign language. To this end we investigated 21 adolescents (aged 12-13) with L1 German and L2 English, which they had been learning for six years in school, who were just beginning to learn Polish as their L3. Three types of data were collected at the very beginning and the end of the school year: a) questionnaire data on the learners’ attitudes towards learning Polish, learning languages and their pronunciation in general; b) a score of the learners’ metaphonological awareness measured with a Polish accent-mimicry task and c) auditory analyses of their pronunciation of Polish /r/ and vowel reduction in a delayed-repetition task. The results of mixed effects logistic regression modeling show that some of the learners’ attitudes predict their accuracy of pronouncing Polish /r/ and unreduced vowels. Moreover, the learners’ pronunciation of unreduced vowels, but not /r/ in their L3 Polish, improved significantly over the school year and higher metaphonological awareness predicted higher accuracy of unreduced vowels in Polish. The results will be discussed from both a theoretical and pedagogical perspective.

ABSTRACT. The present study investigates the production of /h/ or lack thereof in semi-spontaneous L2 speech amongst two distinct groups of French native speakers learning either English (FE) or German (FG). Second languages that differ in phonotactics and prosodic organization from the first language (L1) present a challenge to the articulatory and perceptual habits of the learners (Flege, 1995, Weber & Cutler, 2006). Both English and German have the voiceless fricative /h/ in their inventory, unlike modern French, whose <h> is only present orthographically. It is therefore considered as a “new” phone using Flege’s (1995) terminology. Previous studies have shown that French learners of English and of German often drop the /h/ or resort to h-epenthesis in hiatuses or vowel-initial words (English: Kamiyama et al. 2011, Exare 2021, Capliez forthcoming; German: Quiehl 1906, Wottawa 2017). John and Cardoso (2009) remark that h-deletion is a “stigmatized” feature of francophone speech in English. As a result, “francophones devote considerable effort to overcoming the problem” (2009, p. 120). Quiehl’s early twentieth-century book on the pronunciation of French accented German also comments on French learner’s difficulty in pronouncing the /h/ in German due to the unpronounced <h> in their L1 (Quiehl 1906, pp. 100-101). While German rarely has instances of h-deletion in grammatical words, English speakers often reduce the auxiliary have to /ǝv/ and pronouns his, her, him to [ɪs], [ǝ], [ɪm]. This may impact French learners’ pronunciation and their underlying representations of /h/ in English vs. German. In addition, John and Cardoso (2009) for English L2 and Wottawa (2018) for German L2 both indicate that task has an impact on the pronunciation of /h/ amongst learners. This study draws on Wottawa’s methodology (2018) using a picture description task by first-year university French learners of German (n=15) to measure proficiency in pronunciation and extends it to learners of English (n=12). Picture description is language independent, so the same picture was used for both learner groups. The key elements in the picture elicited words containing word-initial /h/ in both languages, e.g hopscotch in English and Himmel un Hölle in German. Several studies selected reading tasks to measure the pronunciation of /h/ in L2 speech (Kamiyama et al. 2011, Capliez forthcoming), while others used both reading and informal interviews (John & Cardoso). Fortkamp (2000) and Sample & Michel (2014) use picture description tasks to measure fluency but to our knowledge, little has been made to compare segmental pronunciation difficulties in two L2s in semi-spontaneous speech. Preliminary results of our study show that FE and FG produce about 75% of word-initial /h/ as [h]. A more careful examination of the production indicates that some learners tend to avoid choosing words containing a word-intial [h]. The above-mentioned high percentage may be misleading: the degree of lexical diversity and the strategic selection of words not containing /h/ to describe the picture must also be factored in. The study investigates whether this avoidance is linked to low vocabulary size and thus low L2 proficiency (Ovtcharov et al., 2006) or if it is an individual choice of avoiding segments that are linked to production difficulties, being the result of potential trade-off effects on the working memory (Skehan 2009, Bygate 2009).

ABSTRACT. Prosody is a relevant device for encoding and decoding information structure (Bannert 1985, Büring 2006, Baumann et al. 2015); thanks to its linguistic functions, which operate at the syntactic and pragmatic-communicative level, it plays a fundamental role in the interpretation of utterances. In particular, intonation makes it possible to identify the utterance mode and, depending on the position of sentence stress, to determine the focus of the intonation sentence (Altmann 1989, Selting 1995). Different positions of sentence stress thus reflect different meanings. Along with post-nuclear intonation pattern and focus accent position, there are other prosodic phenomena such as pauses, prepausal lengthening, accent patterns, alignment of accentual peaks with metrical structure, contrastive and emphatic accents that contribute to the correct interpretation of utterances (see. Selting 1994, Braun 2006, Berg 2008, Niebuhr 2010). The correct perception of prosodic patterns and especially of different intonation patterns and accentual positions is crucial for understanding and producing spontaneous speech (Röhr & Baumann 2015). However, since Italian uses different intonational devices when compared to German (see D'Imperio 2002, Missaglia 2012, Moroni 2013), Italian learners of German often do not perceive the prosodic regularities of German correctly and apply Italian prosodic regularities to German. Consequently, the correct perception and production of sentence stress, together with the correct interpretation of an utterance are crucial in a didactic model focusing on German pronunciation for adult Italian learners (see Dieling 1996). On the basis of a perceptive analysis aimed at identifying and describing the differences in the perception of prosodic features between Italian learners of L2-German and German native speakers a teaching proposal is presented. The aim is to bridge the gap in the perception and production of German prosody and allow for a less challenging acquisition of correct intonation. The ways in which intonation is perceived and the extent to which the interpretation of utterances diverges between Italian and German speakers underpin this teaching proposal. Semi-spontaneous speech is useful for investigating the ways in which accentual prominence is assigned in utterances and for analysing the relationships between information structure, pragmatic meanings and prosodic features. A corpus of utterances with different structures and relating to different speech acts will be compiled and used both for the perceptive analysis and as an empirical foundation to illustrate the functions of intonation in relation to communicative aspects. The proposal will be based on visual speech by means of an acoustic analysis with PRAAT together with a ToBI-based transcription of prosody (Grice & Baumann 2002, Kügler et al. 2022). The visualisation of prosody will make it possible to highlight occurrences of mismatch between form, prosodic realisation and communicative intent (see. Lambrecht 1996, Hirschberg 2004, Musan 2010). Finally, these occurrences will be used in order to indicate guidelines and good practices for structuring a Redebeitrag.

ABSTRACT. Previous literature indicates that Second Language Learning, specifically the acquisition of a new phonological system, yields varied outcomes depending on linguistic (systemic) and non-linguistic (extra-systemic) factors. Linguistic factors encompass differences in phonological systems and the presence of new vs. similar sounds (see e.g. Flege, 1987), while extra-systemic factors include age, gender, social networks, learning motivation, and input quality and quantity (e.g., Ramon-Casas et al. 2009, Georgiou 2019). The present study is part of an ongoing doctoral project that examines perception and production in adult and child second language learners of German with Italian as their L1. The focus of this study is on the perception of German vowel categories in Italian adult L2-learners. Maturi (2006) identified nine typical "learning problems" that native speakers of Italian encounter in Italian-German language contact. This study focuses on two of those problems: 1) the perception and realization of /ə/ as /e/ or /ɛ/ and 2) the perception and realization of /y/ as /u/ or /i/. Neither /ə/ nor /y/ exist in the phonological system of Italian. The general hypothesis is that the perception of contrasts involving non-native phonemes by Italian native speakers is less accurate than that of German native speakers due to difficulties with phonological categorization. Figure 1 depicts the expected assimilation patterns. In addition, conducting a detailed analysis of mis-categorizations can provide insights into the specific features on non-native phonemes that influence vowel perception in L2 learners of German. The experimental design is based on a perceptual categorization task described by Darcy & Krüger (2012). Participants listened to a series of contrasts realized in two different word templates: schwa-contrasts presented in the final syllable of pseudowords (e.g., kobə vs. kobe or kobɛ) and /y/-contrasts in the first, stress-bearing syllable (e.g., sypo vs. sipo or supo). The pseudowords were presented auditorily as triads in "same" and "odd" conditions on a computer. Each experimental session ended with a survey of a language experience and proficiency questionnaire (LEAP-Q, German/Italian version). Participants included 31 university students with Italian as their L1 who learned German either at school or at the university. In addition, two control groups of German native speakers and of Italian native speakers without knowledge of German are currently tested. A preliminary analysis of the data revealed a mean response accuracy of 91.67% for German native speakers, 74.7% for Italian native speakers with no German knowledge, and 76.07% for the L2-learners of German. The results on reaction times for different vowel contrasts will be presented and discussed in the talk (see figure 2. for a first overview). In ongoing and future work, the samples will be increased, and production data will be compared with the results on perception.

ABSTRACT. The present paper aims to describe a few salient phonological traits of the interlanguage of L1 Italian learners of L2 Polish, with a focus on syllable structure. The bulk of L2 Polish learners in Poland at present is accounted for by speakers of East Slavic languages (Główny Urząd Statystyczny 2020; 2021). However, due to its economic growth, Poland has also attracted significant numbers of “expats”, i.e. educated Western Europeans who choose to relocate to foreign countries (often more than once in a relatively short time) in pursue of career advances (Di Salvo & Moreno 2017). On the workplace, expats typically rely on English (Hazel 2015; Pavesi & Ghia 2020) or their native language (De Mauro et al. 2002; Gatti 2009; Di Salvo & Vecchia 2019); moreover, since the large European cities in which they are typically based are increasingly plurilingual, learning the local language is not always seen as a necessary step (Dąbrowska 2008; Komorowska 2022). The learning of L2 Polish in the expat community has been largely overlooked so far; concerning phonology, in particular, existing works (e.g. Shoemaker 2015; Bernini 2018) mostly rely on data elicited in experimental settings, such as the VILLA project (Dimroth et al. 2013). Against this picture, this paper presents the results of an investigation of L2 Polish acquisition within the Italian community in Poland. Italian (a Romance language) and Polish (a West Slavic language) are quite distant on several grounds, including the lexicon and grammar. Concerning phonology, the two languages differ especially in terms of phoneme inventory (with particular regard to consonants) and syllable structure, Polish accepting more complex consonant clusters compared to Italian. Because of its distance from the expats’ L1, be it objective or perceived (cf. psychotypology; Kellerman 1983; Nelson et al. 2021), Polish is often described in the expat community as an extremely difficult, or even “impossible” language to learn; native speakers’ self-irony on the difficulty of their language often reinforces such impression. Combined, these attitudes often result in numerous expats not even undertaking the study of L2 Polish. Along with orthography and nominal morphology, pronunciation is frequently mentioned as a source of difficulty and frustration. In order to thoroughly investigate these perceptions and their linguistic justification (or lack thereof), the research is articulated into three sections. In the former, based on the results of a specially designed online survey, the attitude of Italian expats towards L2 Polish is presented. The participants’ responses are largely consistent with the most common stereotypes on the difficulty of Polish, especially as far as pronunciation and orthography are concerned. In the second part of the study, semi-spontaneous speech obtained through an Elicited Imitation Task (Kostromitina & Plonsky 2022) is probed from a qualitative perspective in order to identify the typical traits of the L2 Polish interlanguage of Italian learners. The analysis shows that the most common deviations from the expected target concern various patterns of syllable structure simplification. It will be argued that such adaptations are often interpretable in terms of sonority scales, consistently with the insights of previous research (e.g. Vennemann 1988; Young-Scholten 1998; Carlisle 2001). Finally, in order to attempt a correlation between the participants’ perception of Polish phonology and their actual input, a sample of Polish every-day texts is explored in terms of syllable structure. The results indicate that although the target language does contain complex consonant clusters, these are altogether quite rare among the most common layers of the Polish lexicon. The participants’ negative attitude towards Polish pronunciation, thus, does not seem entirely justified. The paper concludes with a discussion of the implications of these findings for the teaching and promotion of L2 Polish (Miodunka et al. 2018).

ABSTRACT. The quality and quantity of phonetic input are known as the main factors in L2 acquisition. To foster phonetic input, a mediation program through games was created and applied to a group of 12 early childhood Brazilian children in a bilingual (Portuguese/English) school. This program promotes the perception of English sounds through games that are suitable for daily teachers' practices. It connects robust L2 sound acquisition theory (Flege, 1995; Flege; Bohn, 2021; Kuhl, 2011; Sacchi, 2018) with pedagogical practices to benefit children's perception in bilingual school contexts. To observe the program's effects on children's sound perception, a perceptive test composed of 4 tasks was applied at two different moments. First, a pre-test at the beginning of the year, prior to the implementation of the program. Later, a post-test, by the end of the school year. This study presents the results obtained from these perceptive tests. The research was formerly authorized by the children's parents, the school where the data was collected, and approved by the University Ethics Committee. The collected data in the pre-test shows the children's difficulties in perceiving the distinction between voiced and unvoiced English plosives and the /ɛ-æ, i-ɪ, u-ʊ/ vowel contrasts. The post-test results show the mediation program had a positive impact on children’s perceptual discrimination of plosives and pairs of English vowel sounds as observed by the high percentage values. The corpus was chosen due to its similarities to some Brazilian Portuguese sounds (Barbosa; Madureira, 2015; Ladefoged; Disner, 2012), which often causes L2 sounds to be assimilated into L1 counterparts, affecting the speaker's intelligibility (Flege, 1995; Flege; Bohn, 2021; Munro; Derwing, 2015). Aiming at enriching children's phonetic input and reverting the assimilated sounds, the mediation program is composed of games that prioritizes the perception of acoustic cues. It is divided into two parts, namely Sound Awareness and Phonemic Awareness. The Sound Awareness games focus on tuning children's attention to the environmental sounds that surround them. Through various games, children get familiar with the ability to discern and identify sounds. After a semester of playing these games, the second part began. The Sound Awareness program was developed to tune children's attention to acoustic cues. The games were created or adapted to suit teachers' daily practices and call children's attention to acoustic cues. Examples of games are "I Spy", "Sequencing Cards", "Odd One Out", "Scavenger Hunt", "Memory Game" and others. This program proposes an order of sounds to be explored: from the easiest to perceive to the most challenging. This order considers the assimilation process and their L1 phonetic repertoire. Thus, these games are to be played first with consonant sounds that differ from each other, and gradually move to consonant sounds that are similar (/p-b/, /-t-d/, /k-g/). The following step is to play those games with vowel sounds from distinct phonetic categories (/ɛ-i-u/, /æ-ɪ-ʊ/), and later start playing with the lexicon that involves similar vowel sounds (/ɛ-æ/, /i-ɪ/, /u-ʊ/). In the last step, after some time preparing their attention to similar sounds, the games will be played with minimal pairs of such sounds. The results, presented in Table 1, contrast the L2 sound perception in the pre-test and post-test. This study combines L2 sound acquisition with teaching practices to enrich phonetic input in young children. It reveals how phonetics can be worked in a playful way and inserted into the pedagogical curriculum to foster children's perception of challenging L2 sounds.

ABSTRACT. Needs analyses conducted with English-Medium Instruction (EMI) lecturers (Sercu, 2004; Ball & Lindsay, 2013; Helm & Guarda, 2015; Picciuolo & Johnson, 2020) have suggested that it is particularly accentedness and pronunciation that causes major insecurity among EMI lecturers. Studies claim that much of lecturers’ concern for accent and pronunciation is grounded in a hidden “native-speakerism” (Holliday, 2005, p. 6) – i.e. the belief that an individual’s language proficiency is measured by his/her conformity to native speaker (NS) rules – which negatively affect their perceptions of their own language competence. Since these views are often shared by the students (Holliday, 2005), EMI lecturers often worry about students’ criticism of their language skills (Tange, 2010) and, more generally, feel pressured to live up to their students’ expectations. Intelligible pronunciation is undoubtedly an essential part of a lecturer’s communicative competence. However, while intelligibility refers to “the extent to which the speaker’s intended utterance is actually understood by a listener”, comprehension refers to “the listener’s perception of the degree of difficulty encountered when trying to understand an utterance” (Derwing & Munroe, 2005, p. 385). A crucial language factor affecting listeners’ degree of perceived comprehension is related to their judgement of their NNS lecturers’ pronunciation accuracy or accentedness (Clark, 2017; Costa & Mair, 2022; Lindemann & Subtirelu, 2013). Furthermore, Kornder and Mennen (2021) noted that learners’ linguistic experience and language background – e.g. monolingual, bilingual – affect their perception of native and foreign-accented speech, which in turn affects learners’ perceptions of quality teaching (Jensen, 2013). In EMI settings, Clark (2017) found that while international students tended to be harsher in self-evaluation of English language competence and less critical of NNS lecturers’ language competence, second-year students, regardless of their L1, were more tolerant of lecturers’ English and had fewer problems in comprehension. However, more recently Costa and Mair (2022) pointed out that “while Italian listeners who share the lecturer’s first language are likely to be more lenient in their perception of the pronunciation […], the international students may find the local accent a greater impediment to comprehension” (p.11). This paper aims to further contribute to the understanding of EMI students’ perception of their lecturers’ speech, and its impact on their comprehension. To do so, this paper reports on an online semi-structured survey administered to 128 students (including both Italian L1 speakers and speakers with other L1s) attending both first and second-cycle lectures delivered in English by Italian L1 lecturers. Students were asked to assess the speed of delivery, accentedness, intonation and pronunciation of their lecturers, and to evaluate whether these linguistic aspects interfered with lecture comprehension. Responses are compared and discussed on the basis of degree cycle, as well as in relation to L1. Findings indicate that it is students’ past experience of English-taught courses as well as their familiarity with non-native English accented speech that influences students’ opinions of EMI lecturers’ language performance and their perceived comprehension in the classroom. The findings will serve to provide further empirical data which might help to set new pedagogical priorities for training programmes for both EMI lecturers and students.

ABSTRACT. Over the past few decades, English pronunciation instruction has shifted its main goal from sounding perfectly native-like to achieving a high degree of intelligibility. However, the digital age presents new benefits and directions in speech teaching and learning. Due to the widespread exposure to digital content in English from an early age, many ESL/EFL learners may already have a high degree of intelligibility and can aim for higher goals. When given a choice, students increasingly pick sounding persuasive the native-way over sounding simply native-like. This study explores the attitudes of 30 first-year business and economics program students in an EFL setting toward a persuasive speech component added to a regular English pronunciation class, as well as their attitudes toward specific persuasiveness features. While sounding persuasive might appear universal at first glance – just speak louder and pause more – languages, in fact, vary in how they express persuasiveness phonetically. In English, some of the features include consonant—rather than vowel—reinforcement, with a special focus on voiceless stop aspiration (Banzina, 2016, 2021; Niebuhr et al., 2016;); lowered pitch (Strangert & Gustafson, 2008; Guyer et al., 2019; Apple et al., 1979); wide pitch range (Niebuhr et al, 2016; Rosenberg & Hirshberg, 2009); intonation patterns with sharp pitch drops (Jiang & Pell, 2017; Banzina & Niebuhr, 2022), or hyper-articulation (Freeman, 2014). From a language teaching and learning standpoint, persuasiveness as a goal is more feasible than nativeness as the focus is on a select number of phonetic features, which give the learner the best of both worlds—both nativeness and effectiveness. Preliminary survey data show that students see significant improvement in how persuasive they sound after seven pronunciation classes and three individual meetings; if before taking this class they mostly used loudness and stressing key words when aspiring to win over the listeners, the change that they believe contributed to their effectiveness most was slowing down and reinforcing consonants in these key words, in addition to pausing and phrase-final falling intonation patterns. Voiceless stop aspiration was mentioned as one of the features that was a complete novelty to the learners and also the one that was hardest to master. Overall, while such a narrow focus might seem relevant only for students in fields such as political sciences, business or law, in reality, the digital age with the volume of audio/video content generated daily and the importance of online presence marks a shift in student needs in most disciplines, as early as in high school. A persuasive speech component could give them some of the competitive advantage they are looking for.

ABSTRACT. Although much work has already been done on various aspects of the processes of L2 pronunciation acquisition and L2 pronunciation teaching (see a.o. Derwing & Munro 2015 for an overview of research), there is still much work to be done on the issue of L2 pronunciation acquisition and teaching as pronunciation is now known to be key in both language production and language perception, especially in terms of L2 speech intelligibility. In this paper, we report on a survey that has been carried out among both L2 learners and L2 teachers of English in the context of the French-speaking part of Belgium (lower ad higher secondary school). We have some descriptive work on the pronunciation difficulties of Francophone learners of various languages, including English and Dutch (e.g. Hiligsmann et alii 2017, Rasier 2006, 2015). However, little is still known about teachers’ and learners’ attitudes, beliefs, and actual classroom practices as it comes to English as a foreign language in the Belgian context. The target audience for our research consists of a representative sample of both Francophone teachers and learners of English as a foreign language. The survey in which they participated, consisted of a questionnaire, semi-structured interviews, and classroom observations. In the questionnaire and interviews, the learners and teachers had to answer questions on their attitudes, beliefs, and practices regarding the L2 acquisition of English pronunciation. The issues under scrutiny concerned a.o. the respective role of theory and practice, the amount and type of exercises, the role of feedback, the alleged priority of suprasegmentals (e.g. stress, intonation, rhythm) above segmentals (e.g. phonemes) and of intelligibility above correction, the issue of the norm and substandard varieties, a.s.o. The reported answers are then compared to actual classroom observations and practices. All in all, the research presented in this paper provides us with a better understanding and a more complete view of L2 English pronunciation acquisition and teaching in the Belgian context. In our presentation, we will also adopt an international perspective by comparing our results with existing data on other populations of learners and teachers.

ABSTRACT. 'English Phonetics and Phonology' is a year one course of the English Studies degree at Universitat de Lleida (Catalonia, Spain). Students must complete two tests and engage in two projects throughout the course. In one of these projects, students must record themselves reading a short text in English, and then its translation into Catalan and Spanish, and to carry out two tasks: (1) calculate the following fluency measures: mean syllables per run, rate of speech time and speech time ratio (Ginther et al, 2010) for the three resulting audio files; and (2) analyse, using Audacity and Praat (Author, 2021), their rendering of the ten boldfaced expressions in the text. I wrote five texts for the 60 enrolled students, so that each text would be read by a maximum of 12 students. For this presentation, the focus will be on the tasks delivered by the students who were assigned Text 1. This is a text of 138 syllables, and the ten boldfaced expressions in the text are the following: "jazz, beige, jacket, casual, twenty, city, cat, sheep, ship, John". I will divide the insights from the students in charge of Text 1 into two sections. The first concerns consonants. The initial phoneme in "jazz", "jacket" and "John" is the voiced affricate /dʒ/ in English and the final phoneme in "beige" is the voiced fricative /ʒ/. These four expressions are used in Catalan and Spanish with only minor variations: "jazz" is spelt the same in both languages and the original English pronunciation is favoured; Catalan "jaqueta" and "Joan" begin with a fricative in the standard, althoguh there is dialectal variation, whereas Spanish "chaqueta" begins with a voiceless affricate and "Juan" with the characteristic velar fricative [x]; Catalan "beix" and Spanish "beige"/"beis" already reflect in the spelling the voiceless fricatives. Students are only allowed to read the text a couple of times for themselves before recording, so there is no time for them to check on these aspects. Thus, the goal of the task is for them to isolate these expressions in Audacity and to reflect, using Praat, on what sounds they have produced and how these would be categorised in the phonologies involved. The other target consonant of the text is medial /t/ in the words "city" and "twenty", as many students produced the flap allophone [t̬]. The second section deals with vowels. Students are asked to find the F1 and F2 of the /æ/ vowels in "jazz", "cat" and "casual" and to compare them with the measures for connected speech provided by Deterding (1997). They are also to use "haircut" to compare /æ/ with /ʌ/, taking the stress context into account. Finally, students are asked to compare quality and length of /i:/ and /ɪ/ through the minimal pair "sheep"-"ship". I believe asking students to engage in this kind of project has several benefits. It allows them to reflect on their productions and to identify aspects of their pronunciation they may have been unaware of. They can compare their productions with those from other classmates who have read the same text. They learn how to use software like Audacity and Praat, tools which they might want to use later in life in teaching contexts. Ultimately, this leads to discussions on the role of standards (for instance British and American, but also Castilian Spanish or Central Catalan) and dialectal varieties, and concepts like pronunciation accuracy, error, mistake or accent.

References: Author. (2021) Deterding, D. (1997). The Formants of Monophthong Vowels in Standard Southern British English Pronunciation, Journal of the International Phonetic Association, 27, 47-55. Ginther, A., Dimova, S., & Yang, R. (2010). Conceptual and empirical relationships between temporal measures of fluency and oral English proficiency with implications for automated scoring. Language Testing, 27(3), 379–399.

ABSTRACT. Nel clima di rinnovato interesse per la pronuncia come aspetto della competenza linguistica da sviluppare in aula e, contestualmente, per gli studi sulla pronuncia delle L2, s’è sentita la necessità d’esplorare non solo le caratteristiche oggettive degli accenti stranieri, secondo un approccio descrittivo, ma anche gli atteggiamenti dei nativi nei confronti degli accenti stranieri della propria lingua. Tali indagini si rivelano particolarmente interessanti nella misura in cui i valutatori nativi esprimono non solo giudizi di tipo affettivo (“irritabile”, “piacevole”, ecc.) o sulla comprensibilità/intelligibilità di parlato prodotto da parlanti stranieri, ma sono in grado di segnalare in modo preciso e analitico i tratti di pronuncia che sono ritenuti maggiormente responsabili dell’accento straniero. Il presente contributo presenta i risultati d’un’indagine esplorativa sulla percezione e sulla valutazione dell’accento italiano dello spagnolo L2 da parte di parlanti nativi. Il questionario è stato disegnato per valutare quattro dimensioni: “comprensibilità”, “fluidità”, “accento straniero” e “reazione emotiva” (o “percezione affettiva”). Sebbene costrutti quali “comprensibilità” e “intelligibilità” (inteso, il secondo, come parametro binario: comprensione o meno del messaggio) possano esser ritenuti descrittori superflui o di scarsa utilità nel caso di lingue affini, come spagnolo e italiano, s’è deciso d’includere almeno il primo (grado di difficoltà nel comprendere il messaggio) per esser certi di poter escludere che le dimensioni “fluidità” e “accento straniero” esercitino un qualche tipo d’influenza sulla comprensibilità. Le quattro dimensioni vengono valutate attraverso scale di Likert; per la fluidità e l’accento straniero viene altresì richiesto d’esprimere un punteggio per ciascuno d’una serie d’aspetti preselezionati dall’autore. Per una valutazione più fine, per le dimensioni “comprensibilità”, “accento straniero” e “reazione emotiva” sono previste, inoltre, delle domande aperte, per dar modo ai valutatori di motivare i punteggi assegnati o di specificare e commentare in modo più dettagliato gli aspetti che ritengono responsabili dell’accento straniero. La “reazione emotiva” viene indagata in termini di “irritabilità”, intesa come grado di fastidio o di rifiuto che può ingenerare nell’ascoltatore nativo lo scarso dominio della pronuncia spagnola da parte del parlante italofono. Nell’ultima parte del questionario, ai valutatori viene infine chiesto di classificare i problemi di pronuncia rilevati in base al grado soggettivo di gravità. All’indagine hanno partecipato 15 studenti universitari italofoni. Per ciascun informante è stata registrata una breve presentazione (stile “parlato spontaneo”), seguita dalla lettura di frasi progettate allo scopo di far emergere i principali tratti di pronuncia che divergono dall’italiano e che potrebbero esser oggetto d’interferenza; lo stile “lettura” presenta l’indubbio vantaggio di favorire la comparabilità dei campioni audio da valutare, per ottenere risultati statisticamente confrontabili. I valutatori nativi sono spagnoli e hanno tutti una formazione in linguistica (alcuni sono professori ELE). Il contributo si propone un duplice obiettivo: 1) contribuire al filone degli studi sulla percezione/valutazione degli accenti stranieri, colmando una lacuna, dal momento che si tratta della prima indagine condotta s’un campione d’apprendenti italofoni – va sottolineato infatti che, per quanto riguarda la pronuncia dello spagnolo L2, i pochi studi pubblicati fino a oggi sono tutti basati su apprendenti anglofoni – ; 2) sul piano didattico, fornire utili indicazioni per la progettazione d’attività incentrate sulla pronuncia dello spagnolo L2 che privilegino quegli aspetti che, se non adeguatamente curati, potrebbero, se non compromettere la comprensione dei messaggi, deteriorarne, anche pesantemente, la qualità fonica; tutto questo, nell’ottica d’una concezione della fonodidattica che, lungi dal porsi come obiettivo il raggiungimento d’una pronuncia “perfetta” (native-like), orienti i discenti almeno verso un modello che rientri all’interno dei parametri d’accettabilità del parlante nativo (con ricadute positive in termini di successo comunicativo, grazie alla percezione psicologica positiva e agli atteggiamenti favorevoli suscitati nell’interlocutore straniero).

ABSTRACT. Intonation is a key factor in the perception of foreign accents; still, it is a rather underanalysed phenomenon in L2 acquisition, and it often remains unconsidered during L2 teaching as well. This research focusses on the intonational patterns of Hungarian learners of Italian as L2, in the framework of Laboratory Phonology (Pierrehumbert, Beckman & Ladd 2000). Due to its extreme dialectal fragmentation, the intonation of Italian is quite heterogeneous (cf. D’Imperio 2002; Krämer 2009; Canepari 2012; etc.); while Hungarian intonation is among the less studied linguistic phenomena as well (cf. Kenesei & Vogel 1989; Varga 2002; etc.); therefore, the findings of this paper are also useful for the general linguistic description of the two languages. However, our main goal is to identify the typical intonational patterns of Hungarian accented Italian, and to offer solutions for L2 teaching. The research is based on speech recordings made with 15 Hungarian and 5 Italian informants who had to read out loud 15 Italian sample sentences three times. The Hungarian informants (12 females, 3 males) are university students, all fluent speakers of Italian as L2; the native Italian control informants (3 females, 2 males; 2 Northerns, 3 Southerns) are university students (2) and teachers (3). The sample sentences are all typical manifestations of an Italian teacher’s classroom communication, and represent four sentence types: yes-no questions, wh-questions, declarative sentences and exclamations (e.g. «Ci sono parole sconosciute?» ’Are there any unfamiliar words?’). Data were analysed in Praat (Boersma & Weenink 1992) by the measurement of fundamental frequency (F0) and by the classification of all syllables according to two variables: H (high pitch) and L (low pitch). We summarised the results through the ratio of the distribution of H/L and sketched the typical intonational patterns of the two informant groups. Surprisingly, the five Italian speakers produced quite homogeneous results, independently of their dialect, age and sex. On the other hand, the Hungarians’ results were considerably heterogeneous, but even so we managed to elaborate the typical intonational patterns of the Hungarian interlanguage. As a case study, we show in this abstract the results of a yes-no question: «Siete pronti per cominciare?» ’Are you ready to begin?’. The Italian informants all used the intonational pattern LLHLLLLHL; that is, they applied high pitch only on the stressed vowels. Hungarians, however, generally began the sentence with high intonation, which fell in the middle, and rose again at the end: HHLLLLLHH. We show a typical Hungarian intonation in Figure 1, while we also show the overall percentages of high tones in Figure 2. From a methodological point of view the following question arises: could intonation be taught during L2 instruction? A further purpose of the research is to make Hungarian learners of Italian aware of this suprasegmental element in the specific context where the learner is the future teacher of L2. We would like to formulate methodological solutions and excercises in order to improve the intonational patterns of Hungarian speaking Italian teachers.

ABSTRACT. L’intonazione è una delle componenti prosodiche più importanti in termini comunicativi perché permette di veicolare informazioni a vari livelli linguistici ed extralinguistici; per esempio, gli si attribuiscono funzioni grammaticali, emotive, pragmatiche, etc. (Fónagy, 1989). In italiano, così come in altre lingue del mondo, l’intonazione fornisce informazioni socio-indicali e, tra le altre cose, è uno dei parametri che tipicamente rivela la provenienza regionale del parlante (Sorianello, 2006). In prospettiva di didattica delle lingue, l’insegnamento dell’intonazione è spesso tralasciato, o eventualmente relegato a uno schema reputato standard, con poca sensibilizzazione verso le possibilità intonative alternative, che tuttavia caratterizzano il parlato dei nativi con cui l’apprendente è confrontato in situazioni comunicative reali. A partire da queste premesse, ci chiediamo se e quanto la variazione regionale intonativa dell’italiano L1 possa rappresentare una difficoltà nella comprensione da parte di apprendenti di italiano L2, e se ci sia un’interazione con il livello di competenza nella L2. In questo lavoro abbiamo preparato un test percettivo con PsychoPy2 che verrà somministrato a 20 apprendenti di italiano L2 anglofoni e 10 parlanti italofoni nativi di controllo. Gli stimoli sono stati creati a partire da un set di 10 frasi, lette in modalità dichiarativa e interrogativa da 7 parlanti nativi (tot. 10 x 2 x 7 = 140 realizzazioni) di diversa provenienza regionale con pronuncia e schemi intonativi associabili alle varietà torinese, veneziana, fiorentina, napoletana, leccese, palermitana e cagliaritana. Queste varietà presentano infatti schemi intonativi ascendenti o discendenti nella modalità interrogativa. Ai partecipanti viene chiesto di ascoltare ogni frase e determinarne la modalità (dichiarativa o interrogativa). Ci aspettiamo: a) che gli italofoni siano in grado di distinguere correttamente l’interrogativa dalla dichiarativa per tutte le varietà presentate; b) che gli apprendenti di italiano L2 non siano sempre in grado di riconoscere le due modalità perché abituati a schemi più ricorrenti (schema ascendente per la modalità interrogativa, tipico della varietà toscana ad esempio); c) che possa esserci un più alto tasso di riconoscimento della modalità per la varietà fiorentina in quanto viene utilizzata come modello nei corsi di lingua, e di conseguenza è più conosciuta rispetto alle altre. Infine, indagheremo se il tasso di riconoscimento corretto della modalità per varietà di italiano regionale aumenti con un livello di italiano maggiore.