The present work reports on a study on acquisition of pragmatics in Italian as a foreign language, in a guided learning context. The study focuses on realization patterns of three speech acts (namely requests, complaints and apologies) following the classifications suggested by Nuzzo (2007), with special attention to the form of address and the use of internal and external modifiers. After a brief but substantial review of the previous literature on the topic, including a summary of the main theoretical assumptions linked to Brown and Levinson’s politeness theory and Austin and Searles’ classifications of speech acts, this study will try and compare native and non-native realization patterns and examine if different variables such as social distance, language proficiency level and contacts learners may have or have had with Italians make any difference in terms of address form selection, sentence lenght and use of modifiers. The methods used in the study consisted of a Sociolinguistic Questionnaire about learners’ sources for language learning, contacts with Italian native speakers and trips to Italy, and of a written Discourse Completion Test (DCT) made of 21 scripted situations (7 situations for each of the three speech acts), with different levels of social distance and intimacy between subject and addressee. The situations were set out orally, explained in detail by a native speaker, following Yuan’s (2001) findings on how orally elicited data seem to be closer to natural speech; participants were asked to write down what they would have said in the very first place. The subjects were Dutch native speakers attending an Italian course, 23 at a B1-level and 20 at a B2-level, and 20 Italian native speakers. Around 60% of the subjects are female. The Belgian subjects are generally older than the Italian subjects. The DCT outcome was used for building a dataset on which Conditional Inference Trees (ctrees) were plotted in R. As a non-parametric method for data classification, based on recursive binary partitioning, ctrees choose the predictor that is most strongly associated with the data partitions as the basis of the split. This method is used in the paper for determining which of the mentioned variables have significant effects in the learners data. The generated plots suggest that learners tend to be more polite to unknown people, while they pay less attention to social distance: they tend to overextend the polite form to informal situations with a younger or socially equal addressee, and to prefer the familiar form while talking to known professionals (e.g. doctors or teachers), where native speakers always select the polite form. Learners with a higher language proficiency seem to be slightly better at grammar, better at selecting the polite form when social distance is higher and intimacy is lower, and to use a native-like number of words and modifiers per sentence. Contacts with Italians seem to be significant for B1-level learners only: those who have had some are slightly better at grammar and make use of a larger number of internal modifiers.
Il presente lavoro si inserisce nel contesto degli studi sull’acquisizione di pragmatica in italiano come lingua straniera, in un contesto guidato. Nello specifico, lo studio che comprende si concentra sui pattern di realizzazione di tre atti linguistici (ovvero richieste, proteste e scuse) secondo la classificazione suggerita da Nuzzo (2007), con particolare attenzione alle forme di allocuzione e all’uso di modificatori interni ed esterni. Dopo un breve ma esaustivo riassunto della letteratura precedente sull’argomento, compresa una menzione approfondita dei principali assunti teorici legati alla teoria della cortesia di Brown e Levinson, e alle classificazioni degli atti linguistici di Austin e Searle, questo studio tenterà di comparare i pattern di realizzazione di parlanti nativi e di apprendenti, e di osservare se diverse variabili come la distanza sociale, il livello linguistico o il contatto che gli apprendenti possono avere o avere avuto con italiani faccia differenza in termini di selezione della forma di allocuzione, lunghezza delle frasi e uso dei modificatori. I metodi utilizzati nello studio consistono di un Questionario Sociolinguistico sulle fonti per l’apprendimento della lingua usate dagli apprendenti, sui contatti con parlanti madrelingua italiani e sugli eventuali viaggi fatti in Italia, e di un Discourse Completion Test (DCT) composto di 21 situazioni create ad hoc (7 situazioni per ciascuno dei tre atti linguistici), con diversi livelli di distanza sociale e intimità tra soggetto e interlocutore. Le situazioni sono state spiegate nel dettaglio oralmente da un parlante nativo, secondo le osservazioni di Yuan (2001) su come i dati elicitati oralmente siano più vicini al parlato spontaneo; i partecipanti dovevano scrivere che cosa avrebbero detto in quelle situazioni. I soggetti sono parlanti nativi di nederlandese, iscritti a un corso di italiano a Leuven, 23 a livello B1 e 20 a livello B2, e 20 parlanti nativi di italiano. Circa il 60% dei soggetti sono di sesso femminile. I soggetti belgi sono generalmente più anziani di quelli italiani. I risultati del DCT sono stati usati per costruire un dataset che, inserito in R, è stato il materiale per dei Conditional Inference Trees (ctrees), che, in qualità di metodo non-parametrico per la classificazione dei dati, basato sul recursive binary partitioning, i ctrees scelgono come predittore la variabile che è più fortemente associata alle divisioni dei dati come base della divisione stessa. Questo metodo è usato nello studio per determinare quale delle variabili menzionate abbia effetti significativi nei dati degli apprendenti. I plot generati suggeriscono che gli apprendenti tendono ad essere più cortesi con persone sconosciute, mentre prestano meno attenzione alla distanza sociale: tendono a sovraestendere la forma di cortesia in situazioni informali con una persona più giovane o con un interlocutore di uguale livello sociale, e a preferire la forma familiare rivolgendosi a professionisti (come dottori o professori) che però conoscono, mentre i madrelingua italiani scelgono sempre la forma di cortesia. Gli apprendenti con un livello linguistico superiore, sembrano essere lievemente più bravi da un punto di vista della grammatica, nonché nella selezione della forma di cortesia quando la distanza sociale è maggiore e l’intimità minore, e nell’uso di un numero native-like di parole e di modificatori per frase. I contatti con italiani sembrano significativi solo per gli apprendenti di livello B1: quelli che ne hanno o ne hanno avuti sono lievemente più bravi in grammatica e fanno uso di un numero più alto di modificatori interni.
Allocuzione e cortesia in richieste, proteste e scuse in italiano: uno studio su apprendenti di italiano LS e parlanti nativi
MAROCCHINI, ELEONORA
2016/2017
Abstract
The present work reports on a study on acquisition of pragmatics in Italian as a foreign language, in a guided learning context. The study focuses on realization patterns of three speech acts (namely requests, complaints and apologies) following the classifications suggested by Nuzzo (2007), with special attention to the form of address and the use of internal and external modifiers. After a brief but substantial review of the previous literature on the topic, including a summary of the main theoretical assumptions linked to Brown and Levinson’s politeness theory and Austin and Searles’ classifications of speech acts, this study will try and compare native and non-native realization patterns and examine if different variables such as social distance, language proficiency level and contacts learners may have or have had with Italians make any difference in terms of address form selection, sentence lenght and use of modifiers. The methods used in the study consisted of a Sociolinguistic Questionnaire about learners’ sources for language learning, contacts with Italian native speakers and trips to Italy, and of a written Discourse Completion Test (DCT) made of 21 scripted situations (7 situations for each of the three speech acts), with different levels of social distance and intimacy between subject and addressee. The situations were set out orally, explained in detail by a native speaker, following Yuan’s (2001) findings on how orally elicited data seem to be closer to natural speech; participants were asked to write down what they would have said in the very first place. The subjects were Dutch native speakers attending an Italian course, 23 at a B1-level and 20 at a B2-level, and 20 Italian native speakers. Around 60% of the subjects are female. The Belgian subjects are generally older than the Italian subjects. The DCT outcome was used for building a dataset on which Conditional Inference Trees (ctrees) were plotted in R. As a non-parametric method for data classification, based on recursive binary partitioning, ctrees choose the predictor that is most strongly associated with the data partitions as the basis of the split. This method is used in the paper for determining which of the mentioned variables have significant effects in the learners data. The generated plots suggest that learners tend to be more polite to unknown people, while they pay less attention to social distance: they tend to overextend the polite form to informal situations with a younger or socially equal addressee, and to prefer the familiar form while talking to known professionals (e.g. doctors or teachers), where native speakers always select the polite form. Learners with a higher language proficiency seem to be slightly better at grammar, better at selecting the polite form when social distance is higher and intimacy is lower, and to use a native-like number of words and modifiers per sentence. Contacts with Italians seem to be significant for B1-level learners only: those who have had some are slightly better at grammar and make use of a larger number of internal modifiers.È consentito all'utente scaricare e condividere i documenti disponibili a testo pieno in UNITESI UNIPV nel rispetto della licenza Creative Commons del tipo CC BY NC ND.
Per maggiori informazioni e per verifiche sull'eventuale disponibilità del file scrivere a: unitesi@unipv.it.
https://hdl.handle.net/20.500.14239/7772