The development of Basque and Spanish in Basque

The development of Basque and Spanish in Basque immersion programmes Idiazabal, Itziar Almgren, Margareta Manterola, Ibon Euskal Herriko Unibertsitatea. University of the Basque Country (EHU-UPV) Research projects: UPV/EHU (9/UPV 00033. 130 -13614/2001), Eusko Jaurlaritza-Basque Government (IT-262 -07), Spanish Ministery of Education (BFF 2003 -05196; HUM 2006 - 11862; FFI 2009 – 13956).

Introduction • Multilingualism is seen as a priority in education, in particular in Europe, where educational programmes such as CLIL are considered effective models to produce multilingual students (Council of Europe, 2001; Coste et al. , 2006; Berthoud & Gajo, 2008) • Multilingualism contributes to cognitive, linguistic and social benefits (Baker, 2001; Cenoz et al. 2001; Kecskes, 2008; ) • Research questions in the Basque context: • Does education in Basque L 2 contribute to linguistic competence in both Spanish L 1 and Basque L 2? Does education in Basque L 2 bring about attrition of Spanish L 1? In the case of immigrant children, do they attain linguistic competence in Basque L 2? Does age of first contact with L 2 matter? • • •

Our research project • • • Our research project aims at studying the development of Basque and Spanish in Basque immersion programmes. Data have been collected on the same subjects at ages 5, 8 and 11 which gives our research a longitudinal as well as a cross-sectional perspective. One immersion study group is constituted by children whose L 1 is Spanish from a predominantly Spanish speaking environment. A second immersion study group consists of children with immigrant background. The reference group is formed by children whose L 1 is Basque in an environment where the use of Basque is very extended. Spanish L 1 and Basque L 1 groups have assisted preschool and school in Basque from age 2 or 3. The immigrant children started school in Basque at different ages. Our analyses contain both discourse and grammatical approaches (Almgren et al. , 2008; in press; Idiazabal & Manterola, 2009; Ezeizabarrena et al. , in press; Aeby & Almgren, in press). The tasks carried out during data collection are related to classroom activities, a fact which renders special relevance to our findings for further research on didactics and teaching of languages in multilingual contexts.

What do we analyse? • • Text genre is considered a relevant research unit for the study of language acquisition and use (Bronckart 1996/2004). In this presentation the text genre chosen is that of oral story-telling, a frequent activity in classroom contexts. The stories produced in Basque and Spanish by our subjects constitute the corpus for the study of L 1 and L 2 development. Our analysis concerns: 1. The level of narrative autonomy 2. The reproduction of the narrative structures of the stories. 3. Verb cohesion 4. Language choice and self corrections.

Subjects: Basque L 2 / Spanish L 1 • The same 37 children participate in this study which refers to ages 5 (preschool) and 8 (primary school). • These subjects were selected according to their family language, Spanish L 1. • Their environment is almost exclusively Spanishspeaking, Basque being reduced to school context. • The children have been educated in early and total immersion in Basque. • Spanish is introduced as school subject during primary education. • Their narrative skills are measured in Basque and Spanish.

Subjects: Basque L 2 / immigrants • 6 children with immigrant background and different L 1 participate in this study. • Their age of first contact with Basque varies from 2 to 9 and their age at the moment of recording from 5 to 11. • The children have been educated in total immersion in Basque. • Their environment is mainly Basque speaking. • Only their competence Basque is studied. in K H R 5; 10 7; 04 7; 06 Z 8; 02 A S 9; 09 11; 10 L 1 A R A B I C P O R T U G U E S E B U L G A R I A N A R A B I C P O R T U G U E S E A. O. 2; 02 5; 00 5; 11 4; 06 6; 01 9; 06 T. E. 3; 08 2; 04 1; 07 3; 08 2; 04

Reference group: Basque L 1 • The same 24 children participate in this study at ages 5 (preschool) and 8 (primary school). • The subjects were selected according to their family language, Basque L 1. • Their environment is predominantly Basque- speaking. • The children have been educated in their L 1 Basque. • These children constitute the reference group for competence in Basque.

Data collection • In groups of five these children were videotaped in a classroom setting while listening to an adult who told the story with the support of a wordless picture-book. • One child then retold the story to a second one, while the others were taken apart. Next, the second child told the story to a third one…until all five children had completed the task. • The children were told that their performance would be videotaped and shown to smaller children who did not know the story. • When data were collected in Basque and Spanish, two different stories with the same structure were used, one for each language.

The story in Basque: Mattin Zaku Classical structure of children’s tale with a happy end, with the canonical 5 narrative phases (Adam 1992) Setting Mattin Zaku and his mother were very poor. Onset Mattin Zaku told her he would go to the castle and ask the king for money. Unfolding He met three helpers (the fox, the wolf and the river) and was faced with three difficult tasks to overcome. Resolution Having solved the problems with the aid of the helpers he obtained the money. Outcome Mattin Zaku and his mother had money and lived happily ever after.

The story in Spanish: Centellita Classical structure of children’s tale with a happy end, with the canonical 5 narrative phases (Adam 1992) Setting Centellita and her mother, who lived in the forest, ran out of fire and could not cook. Onset Mother told her to go and fire at witch’s house. Unfolding The heroine met three helpers (the owl, the dwarf and the fairy) and then was faced with three difficult tasks to overcome. Resolutio n Having solved the problems with the aid of the helpers she obtained the fire. Outcome Centellita and her mother could cook and lived happily ever after.

Narrative autonomy • Many authors consider the age of 5 the reference point for narrative autonomy (Berman & Slobin, 1994), although not for all children (Serra et al. 2000). • In Spanish L 1 the 78 % of our subjects can be considered autonomous, and in Basque L 1, 85 %. • In Basque L 2, the degree of autonomy is even higher: 89 %. • At age 8, all of them are autonomous in Basque L 1, and 92 % in Spanish L 1 and Basque L 2. • Among the immigrant, only the 5 -year-old girl needs help in telling the story. • Although these degrees of autonomy are remarkable, other discourse skills may of course not have been acquired yet (De Weck, 2005; Hickmann, 2003).

Narrative structure • Both stories are constructed according to 5 narrative phases. • The reproduction of all 5 narrative phases reflects the ability of text planning. • Our analyses of children’s reproduction of narrative structures (Almgren et al. 2008; in press) show that phase 3 (unfolding) is always present although with occasional omissions of characters or actions. • However, the symmetry between the setting and the outcome and the onset and the resolution often poses more problems. Children tend to forget to mention the lack of money / fire which is the reason why the whole story happens. Similarly, the success of the enterprise is also omitted more than once. • On this occasion we choose to analyse more closely the symmetries between phases 1 and 5, where most problems seem to appear.

Phase 1 - Phase 5: Table 1 Age 5 Spanish L 1 86 % Age 8 Age 5 Age 8 73 % 92 % Basque L 1 Basque L 2 Age 8 97 % Basque L 2 Immigra nts Age 5 29 % 79 % K 5; 10 H 7; 04 R 7; 06 Z 8; 02 A 9; 09 S 11; 10 X √ √

Phase 1 - Phase 5: comments • This table reflects certain differences between the groups in their text planning and also the influence of age. • In Basque L 2 children perform well already at age 5 although by age 8 they have improved. • This is also true for the same subjects in their L 1 Spanish. • The low results in Basque L 1 at age 5 were unexpected. However, they have improved considerably at age 8 although they remain below the L 2 group. • The results for immigrant children seem to approach at those obtained by Basque L 1 subjects. • In this group, the 5 -year-old child fails to reproduce the symmetry between phases 1 and 5 and so does one of the older subjects.

Verb cohesion: tense maintenance • Discourse analysis has revealed that each genre tends to hold to a basic or dominant tense system. • One of the characteristics of story-telling is the use of past tense to mark the distance between the story and the communicative situation, one of the means by which the narrator establishes a narrative discourse outside the hic et nunc (Bronckart, 1996). • In some cases young children’s stories are anchored in present tense (Bamberg, 1987; Sebastian & Slobin, 1995). • It follows that if unexpected tense switching occurs, this produces problems in the cohesion of the text (Dolz 1990). • However, some changes between past and present tense may not be considered unexpected tense switching but can be seen as a resource for narrative dynamism.

Past tense maintenance Table 2 Age 5 Spanish L 1 78 % Age 8 Age 5 Age 8 89 % 94 % Basque L 1 Basque L 2 Age 8 89 % Basque L 2 Immigra nts Age 5 79 % 67 % K 5; 10 H 7; 04 R 7; 06 Z 8; 02 A 9; 09 S 11; 10 X √ √ √

Examples • The Basque L 2 / Spanish L 1 subjects are quite systematic in both languages, a little less so at age 5. Sometimes unexplainable tense switching occurs as in the following examples: esan zuen otsoa / nora zoaz Ma-kin Zaku? / erregearen gaztelura / esan-ten zu-en Makin Zaku / eta esaten du otsoa / nik e nik oi- e: zalditegia ikusi nahi dut / the wolf said/where are you going, MZ? / to the king’s castle/ said MZ/ and the wolf says /I e I want to see the stable. le dijo la madre / vete. . . / y va caminando. . / y le dice. . . / y le dijo. . . The mother said to him /go / and he walks on / and he says to him / and he said to him • Most immigrant children maintain the story in past tense. Two of them produce unexpected tense switching as above: ta gero esan do / ibaia! / sartu zakoan! // ta gero / Ma- / Matin Zaku esan / esan zon / erregea! dirua nahi dut! and then he says / river! / get into the bag! // and then / Ma- / Matin Zaku sa- / sa- said / king! I want money! /

Comments • Sometimes, however, this tense switching expands over longer segments than single sentences and this rather seems to be a resource used on purpose. da ezin tzun pasa / da da ibaya boltsan sartu zun (4") iristea gaztelura atea jotzen du tan / ta eztote irekitzen / berriro jotzeu / tan / da ireki iteyote / da erreana jutea / (. . . ) ta sateyo zalditegira eramateko / ateatzeu otsua / ta zaldi danak akatzeitu (5") ta san tzun / oin / plaza erdiyan jarriko diogu / sutan / da atea zun ibaya ta su dana itxaldu zun / and he couldn’t pass / and he put the river into his bag (4’’) he arrives at the castle and knocks on the door / and nobody opens / he knocks again / and they open the door / and he goes to the king (. . . ) and he tells him to take him to the stable / he brings out the wolf / and it kills all the horses (5’’) and he said / now / we will put him in the middle of the square / we will burn him / and he took out the river and it put out the fire. • The lower percentage for Basque L 1 at age 8 (as reflected in table 2) can be explained by these alternations, rather than unexpected changes. These kinds of alternation do not happen so often in Basque L 2,

Ellipsis of the auxiliary as a verb cohesion resource in Basque • In Basque the auxiliary verb containing tense, case and person marks can be omitted in certain contexts. • In narrative contexts this is a frequently used resource which brings about swiftness or quickens the story-telling. • At age 5 only a few such examples can be found in Basque L 1. ta / en / zakua hartu –tu (Ø) (4´´) ta jun tzen ya basotikan and / m / pick (Ø) up the bag (4’’) and he walked through the forest • The ellipsis of the auxiliary verb increases at age 8, with similar percentages in both L 1 (in 37, 5 % of the stories), and L 2 (32, 4 %). • In the immigrant corpus we only found one example: ibaia zakun sartu (Ø) eta gaztelura jun zan put (Ø) the river into the bag and he went to the castle • This example appears in the story of S, age 11, whose first contact with Basque was at age 9; 06.

Metalinguistic awareness • We assume a wide definition of metalinguistic awareness that includes both the knowledge about structural components of language and the knowledge about discourse features and language use (Hammers & Blanc, 2000; Schneuwly & Erard, 2005). • We also refer to the fact that the acquisition of an L 2 may bring about an increased conscience of the linguistic forms and functions in the L 1 (Vygotski, 1934/ 1997). • The following examples refer to features such as self corrections or finding the correct word form or tense and also to language choice within the classroom activity.

Metalinguistic awareness • Examples of self corrections as the following show that children are aware of verb cohesion features such as tense and mood correctness: eta esan zion Matxin Zaku e ea nahi baduz- nahi z-nahi ba-zu: duzu / nahi bazuen etorri / eta esan zion baietz (Basque L 2, age 5). and MZ said to him e if you wa- if he wa- want /if he wanted to come and he answered him yes que si quería ayuda que le llamaba- que le llamara a ella (Spanish L 1, age 5) that if (she) wanted help she called- she should call her

Metalinguistic awareness • The following example illustrates both searching for the correct verb form and vocabulary correspondence in Basque: • • • Haurra: dirua nahi dut! / esan zuen / ireki zi- zuen atea eta: // nola esaten da esposa? Heldua: e? / Haurra: esposa / Heldua: lotuta / esan lotuta / Haurra: eta eskuak lotuta jarri zitzaio- / zi: - / zituzten // erregearengana eraman zuten (Basque L 2, age 8) Child: I want money! / he said / (he) op- opened the door and/ // how do you say handcuffs? Adult: eh? Child: handcuff Adult: tied / say tied / Child: and with the hands tied (he) put hi- / flexional morpheme / they put them / (they) took him to the king It seems that the child is more eager to find the correct word in Basque than the adult who does not give a translation equivalent.

Metalinguistic awareness • It is remarkable that the Basque L 2 children always stick to Basque when asking about vocabulary they don’t know in Basque. They never slip into Spanish L 1 seeming to be aware of which is the language to be used in the classroom. • • • Haurra 1: gero / abiatzen zeudenean / e / zorro- / nola da? / lobo bat nola esaten da? / Haurra 2: otso bat / Haurra 1: otso bat ikusi zuten (Basque L 2, age 8) • Child 1: then / when they were on their way / eh / fox in Spanish / how do you say it? / how do you say a wolf? (‘a’ in Basque, ‘wolf’ in Spanish) Child 2: a wolf / Child 1: (they) saw a wolf / • • • On the other hand, now and then these children integrate Spanish words into Basque discourse, since they are aware that the audience understands both languages. • • Haurra: Eta joan zen eta kaminotik, mm, azeri bat enkontratu zuen Child: And he walked along the(in Basque) road (Spanish) um, he met (Spanish lexical verb with Basque aspect marking) a fox.

Metalinguistic awareness • The next example from the Basque L 2 immigrant corpus shows in the first place an example of self correction concerning verb form • Haurra: erregea-ren-gana ta esan zuten / esan zun e / oi / esan zon // e- erraman (=eraman) (helduari begira, laguntza eske) Heldua: oilategira! Haurra: oila-tegira! • • • Child: to the king and (they) said / (he) said eh / uups / (he) said (in local dialect) // ta- take (looking at the adult for help) Adult: to the hen coop! Child: to the hen co-op! • However, the immigrant children rarely ask “how do you say. . . ? “ in Basque. Neither do they use their L 1. Evidently they know that the adult cannot help them there. • Finally, they never integrate words from their L 1 into their stories in Basque, which also shows that they are aware of their own linguistic situation.

Conclusions (1) • • • Returning to our initial questions Does education in Basque L 2 contribute to linguistic competence in both Spanish L 1 and Basque L 2? As our data show, already at the age of 5 the Basque L 2 group have acquired a narrative competence which is comparable to that of the Basque L 1 subjects. They also show a parallel development at age 8. • • Does education in Basque L 2 bring about attrition of Spanish L 1? As to their competence in Spanish L 1, no signs of attrition can be appreciated. • • In the case of immigrant children, do they attain linguistic competence in Basque L 2? As far as our data show these children do not differ substantially from the Spanish L 1 group • • Does age of first contact with L 2 matter? The early contact with Basque among the Spanish L 1 children leads to a high competence in Basque L 2, which was not unexpected. But the immigrant children, with varying ages of first contact with Basque also acquire linguistic skills which are comparable to the Spanish L 1 group. No doubt, not only age, but also the sociolinguistic environment has been relevant for the results in the immigrant group. •

Conclusions (2) • As a whole, as far as narrative skills are concerned, our research shows: • Immersion in Basque contributes to a satisfactory development of both Basque L 2 and Spanish L 1 • Immersion in Basque contributes to a satisfactory development of Basque L 2 in the case of immigrant children. • Bilingual children not only develop their two languages satisfactorily but also gain conscience on formal and functional linguistic aspects.