Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction: The Educational Context of the Research

The Spanish Bilingual Sections programme in Poland is based on the May 1997 agreement between the Ministry of National Education and Sport for the Republic of Poland and the Ministry of National Education and Science for Spain, and the later 2005 and 2010 appendices, as its legal basis. These relate to the creation and running of Spanish Bilingual Sections in middle and high schools in the Republic of Poland. To be admitted onto the programme, candidates have to perform linguistic ability tests, which examine general understanding and command of the language. The selected students then follow a programme reinforced with classes in Spanish. At the end of this programme, in addition to getting the Polish baccalaureate certificate, they have the option to obtain the Spanish baccalaureate certificate, after passing the corresponding exams and complying with the requirements necessary to issue said documentation.

The Spanish Bilingual Sections programme is divided into two educational stages. The first stage, during which students devote 630 h of classes to studying Spanish as a Foreign Language (SFL), is completed during Middle school or at the beginning of high school. The second stage is a specific programme consisting of “Spanish Language and Literature” and “Spanish History and Geography”. Students dedicate approximately 1,100 h to classes, which are given in Spanish, over the three baccalaureate years.

The Bilingual Sections model in Poland implements what is usually referred to as “Content and Language Integrated Learning” (CLIL), as part of the 2004–2006 European Commission Action Plan to promote language learning and linguistic diversity in the European Union. Through this type of teaching, students learn curriculum contents whilst exercising and perfecting their linguistic competences, combining contents and language. CLIL emerges “with the aim of better preparing students for life in a Europe in which mobility is becoming increasingly widespread and should be within reach of everyone” (Eurydice 2006: 3).

With regards to learning a foreign language, CLIL objectives in Poland are threefold: (i) promoting the values of tolerance and respect towards other cultures through use of the CLIL target language (socio-cultural objectives); (ii) developing linguistic abilities with an emphasis on efficient communication, motivating students to learn languages by using them for real, practical purposes (linguistic objectives); (iii) developing subject-related knowledge and learning ability, stimulating the assimilation of content by means of different and innovative methods (educational objectives) (Eurydice 2006: 22).

As to the educational objectives, the idea of an innovative methodological approach is intrinsic to CLIL. Both the language and the non-linguistic content constitute teaching tools, with neither one being predominant over the other. “Achieving this twofold aim calls for the development of a special approach to teaching in that the non-language subject is not taught in a foreign language but with and through a foreign language (Eurydice 2006: 7)”. Arnau (2001) highlights the following characteristics of this innovative approach: (1) language is learnt within a contextualised use; (2) the learner employs the language in a meaningful way, using it to communicate himself or herself; (3) the student learns forms whilst using them, and uses them whilst learning them.

Therefore, the CLIL methodology responds to the need for students to be exposed to situations that require authentic communication, because “learning a language is learning to communicate oneself” (Ellis 1992). This point is related to the CLIL linguistic objectives’ concept of effective communication, which promotes the teaching of a foreign language by means of a communicative approach.

Regarding CLIL socio-cultural objectives, it should be highlighted that in the Spanish Bilingual Sections model in Poland, language is strongly related to culture. In addition to achieving high-level linguistic capabilities, a great amount of knowledge about Spain is also acquired in bilingual classes. Therefore, language is not only a tool for communication, but also, and most importantly, an instrument to relate and convey culture (Tatoj et al. 2008).

1.1 First Stage: Year 0 vs. Middle School (Gimnazjum)

Taking the CLIL approach in the bilingual programme into consideration, it is imperative that the students gain a level of linguistic competence in Spanish in order to tackle the non linguistic subjects with success. For this reason, when defining the linguistic competence objectives and contents at different stages of the bilingual programme, each stage of the Bilingual Sections curriculum is related to the Council of Europe’s Common European Framework of Reference for Languages (CEFR) scale. Furthermore, basic functional, grammatical and socio-cultural contents required at different stages of the Cervantes Institute Spanish Diploma qualifications are also integrated within the programme. This means students should reach the level of ‘independent user’ (B1) by the end of the first stage, and have achieved the level of B2+ by the end of the bilingual programme, approaching the stage classified by the CEFR as ‘competent user’ (Table 9.1).

Table 9.1 Comparison of curriculum levels to Spanish sections
Full size table

Within this context of language learning, an emphasis is placed on the first stage of the bilingual programme, with 630 teaching hours devoted to studying Spanish as a Foreign Language (SFL). The objective of this stage is to prepare students for learning non-language subjects in Spanish, and the language is treated as an essential vehicle for acquiring knowledge in Literature, History, Geography and Spanish Art History. This stage also gives teachers the chance to evaluate the linguistic level achieved by students in depth. It also allows them to evaluate the application of this knowledge in both language and non-language subjects during the following stages.

Poland’s Spanish Bilingual Sections programme offers two learning modalities for this initial stage:

  1. (a)

    Modality I: Year 0 – language immersion, with intensive Spanish as a Foreign Language (SFL) classes, to which at least 18 h are dedicated per week. This course is given in high schools in Year 0, before the first year of the Baccalaureate.

  2. (b)

    Modality II: A 3-year course of extensive SFL classes. These classes are taught in gimnazjum (Middle school), and the 18 weekly hours of classes seen in Modality I are shared out over the 3 years of Polish Middle school, therefore becoming six teaching hours a week.

In Year 0, (high school), classes are taught by three or four teachers, both Polish and native Spanish speakers. With regard to the 3-year Middle school course, students have one main Polish teacher, supported by a native Spanish speaker who gives 1 h of conversation classes per week and two additional hours of Introduction to Spanish Literature classes from year 2 onwards.

Concerning continuity in the programme, there is normally one class in Year 0, where students continue to follow the bilingual programme in the first year of high school, except in very special cases. In Middle school, there are normally two classes, reduced to one in high school. This reduction is on one hand due to students voluntarily dropping the subject after the 3 years of Middle school and on the other, due to a selection process based on linguistic competence in Spanish gauged by a final exam and the student’s academic performance shown in their school report. Students who voluntarily leave the course mostly do so because either they choose to study non-humanities subjects or they have a purely linguistic interest in the programme. To a lesser extent, students leave the programme as a result of the difficulties they experience in their learning of the Spanish language (Table 9.2).

Table 9.2 Bilingual programme stages in the Spanish sections in Poland
Full size table

Regardless of which modality is chosen, the student must complete the same SFL curriculum objectives, content, tasks, etc.). The SFL curriculum, which serves as a reference for “Year 0” and the three Middle School years, was developed by the Education Office of the Embassy of Spain in Poland (Consejería de Educación de España en Polonia 2005).

However, each method has pros and cons, and various factors (time, funding, motivation, effort, or psychological development) tend to favour one modality or the other. These factors were analysed by Tatoj et al. (2008) in the Evaluation of Poland’s Spanish Bilingual Sections Report. Based on interviews carried out with teachers in Bilingual sections, the authors concluded that students who followed the Year 0 course showed a higher linguistic and learning level than those who had followed the bilingual Middle school programme. The report states that teachers highlighted the fact that choosing the bilingual course in high school is a more conscious decision, and as such, students are more motivated to study. According to the authors, in general, high school students who take the Year 0 course gain extra time to concentrate almost exclusively on the study of the Spanish language. Furthermore, it has been claimed students can learn the grammar more easily at the age of 16 than when they are 13 years old (Tatoj et al. 2008).

On the other hand, in their opinion, bilingual classes during the 3 years of Middle school does not adequately prepare students for the demands of a bilingual class in high school. Students have neither sufficient knowledge nor linguistic abilities to allow them to participate fully in classes given entirely in Spanish, by a teacher who does not speak Polish (Tatoj et al. 2008).

Evidently, these conclusions have been fiercely challenged by teachers and educators working in bilingual Middle schools in Poland. In defence of Middle school teaching, it has been argued that both the Council of Europe and the European Union encourage bilingual teaching from the beginning learning stages (pre-school and primary school), and that as a result of the early start and partial immersion method, these programmes have been incredibly successful in terms of language performance in Canada and America (Eurydice 2006). The official stance of Poland’s Ministry of Education is to favour the homogenisation of the first stage of the Bilingual Sections around the Middle school model, and as such, is opting for the abolition of Year 0 and the establishment of collaboration ties with “satellite” Middle schools, where future students would be prepared for the second stage of the bilingual programme.

2 Theoretical Foundations: Lexical Availability and Evaluation of Lexical Competence

As it has been shown, it is the first stage of the bilingual programme which develops and establishes the basic linguistic abilities in Spanish necessary to tackle non-language subjects in the bilingual Baccalaureate curriculum. Lexical competence is one of these competences.

Even with the limitations of the methodology (Hernández Muñoz 2006; Higueras García 2008), a study of lexical availability is suggested as an exceptional tool to evaluate the control of fundamental vocabulary which ensures the ability of basic communication in a foreign language.

As it is well known, in addition to basic and common linguistic expressions related to our physical surroundings (head, window, food) or basic conceptual distinctions (sleep, leave, enter), the fundamental lexicon also includes other, more abstract terms expression of possibility, how close or far away the concept is from reality, etc. The fundamental lexicon includes two easily-distinguishable lexical sub-groups (Michéa 1950, 1953): (a) Basic lexicon – commonly- used and non-subject specific. Mainly grammatical words and words which continuously appear in any conversation or written text, regardless of the topic being discussed (e.g., to, the, not, many, there is, give, person, put, etc.), and (b), Available lexicon – topic-related, comprising specific semantic content and words that whilst commonly-used, are only employed in relation to a topic (e.g., frying pan and fork in relation to ‘kitchen’, or letter and stamp in relation to ‘post’).

It is precisely the need to select words which must be taught in foreign language classes which gave rise to the birth of lexical availability, as explained in the introductory chapter to this book (Chap. 1).

From a pedagogical perspective, the benefits of studying lexical units – lexías, using Pottier’s terminology (1971) – must be highlighted. This helps to evaluate adequate learning of vocabulary as well as to determine the group of widely-available words which shape the active lexicon. Effectively, lexical availability is designed to evaluate the school’s efficacy in its aim to educate the students in their command of the fundamental lexicon, both in L1 and L2. This evaluation has been carried out by means of the monitoring of students’ development of lexical competence at specific points in the learning process.

In the Spanish-speaking world, the results of such an evaluation of the native tongue have provided researchers with different results. López Morales (1973, 1978) in San Juan de Puerto Rico, and López Chávez (1993, 1995) in Mexico measured the development of lexical availability in primary education. Both parties detected significant irregularities in the pupils’ lexical competence and a lack of gradual qualitative and quantitative progression in the lexical acquisition process. However, Román-Morales (1985), in Dorado, Puerto Rico, Mena Osorio (1986) in Concepción, Chile, Echeverría (1991), in Chile, and Alba (1995), in the Dominican Republic found positive results. After conducting an analysis of lexical availability at three different stages of primary school, they all noticed a fairly regular, gradual increase in the number of word types, as well as in the average of word responses as school grade increased.

Similarly, lexical availability in L2 also allows for the examination of different phases of the lexical learning-acquisition process of Spanish as a Foreign Language. This was done by Carcedo González (1998, 2000) in his studies devoted to studying Finnish students’ lexical availability in Spanish. In his conclusions, Carcedo González (2000: 213–216), recorded a very uneven development of vocabulary in different subject areas, and a gradual evolution of lexical richness parallel to the rise in the level of study, with a qualitative leap from high school to university level.

Carcedo González’s monograph (2000) looks at the lexical availability of a sample of 350 Finnish students, learners of Spanish as a Foreign Language. Without doubt, this constitutes the work of reference for any exhaustive analysis of learners’ lexical availability and of the effect that extra-linguistic variables can have on it. The variables considered by Carcedo González are ‘type and course grade’ (4th and 8th year of high school, first and second year of university), ‘gender’, ‘mother tongue’ (Finnish and Swedish) and ‘knowledge of other Romance languages’.

Following this, Samper Hernández (2002) devoted a monograph to the examination of the lexical availability of 45 students of different nationalities who attended Spanish courses at the University of Salamanca, adopting methodological guidelines similar to those of Carcedo González (2000). In her study, she found out a clear decrease in learners’ lexical development once students reached the highest level. This was explained by students being poorly grouped according to their command of the Spanish language, or by the belief that the use of more complex or less common lexical units – and not only the number of words – implies a better mastery of a foreign language Samper Hernández (2002: 85–86).

Using these studies as a model, I conducted an analysis of the development of lexical availability in Polish students, learners of Spanish attending bilingual sections in Poland (López González 2010). In this study, using two identical samples of 120 students studying Spanish in Middle school and High school, I found an evident enrichment of lexical competence in Bilingual Sections students as they advanced in their studies, both in total words (+27.8 %) – quantitative – and in different words (+ 49.6 %) – qualitative – together with the existence of a solid common base in both educational levels with regard to easily-available vocabulary.

Lexical availability studies therefore allow for the identification and understanding of the vocabulary which is actually available to a group of language learners. As such, they become an instrument for the evaluation of lexical competence in a foreign language as well as for the study of the effect of educational methods on the development of lexical knowledge.

Germany and Cartes (2000) carried out a study to determine the effect of the factor ‘type of educational setting bilingual, (private, state school) on the lexical availability of learners of English as a foreign language in Chile. Using a sample of 60 students in the first year of Middle school, and working with three cue words, ‘Body’, ‘Food’ and ‘House’, they found out that the teaching methodology used in each institution proved to be decisive. Students in the bilingual educational proved to have a higher degree of lexical availability than students in other educational institutions. This was so because they used the target language as a means of communication in 80 % of the core subjects in the curriculum. Behind the bilingual school, it was the private school, in which English was taught by way of a communicative methodology based on functions of the language. The state school appeared in last position, with lower results due to vocabulary being taught out of a communicative context and following a traditional programme based on the teaching of grammar rather than on a communicative approach.

With regard to Spanish as a Foreign Language, the study by Higueras García (2008) is outstanding. This author looks at the lexical availability of 43 adult students learners of Spanish as a Foreign Language in the metropolitan zone of Madrid, in six semantic categories (‘Body parts’, ‘Clothing’, ‘Food and drinks’, ‘The kitchen and its utensils’, ‘Games and entertainment’, and ‘City’). For this purpose, initially, she follows Carcedo González’s (2000) and Samper Hernández’s (2002) methodology, bearing in mind the extralinguistic factors of ‘sex’, ‘age’, ‘socio-cultural level’, ‘mother langue’ and ‘knowledge of other languages’. However, given the main characteristic of the group – Intermediate level (B1) students from two Official Language Schools in Madrid, in a programme of immersion in the Spanish language and culture – this author includes two new variables: the ‘Teaching-learning method’ (regulated methods versus non regulated methods – both methods) and ‘Years of study of the Spanish language’ (with intervals of a year). Regulated methods include universities, official language schools, or private language academies; non regulated methods are non systematic methods and self-taught learning.

In the results, “the variable ‘Years of study of the Spanish language’ has a significant impact […], to such a degree that a general directly proportional relationship between the years of study of the Spanish language and a larger number of words provided by informant can be noted” (Higueras García 2008: 202), with “a consequent upward trend between the average number of responses given by the informant and the number of years that the informant has learnt Spanish”, up to 3 years of study (Higueras García 2008: 205). In the variable ‘Teaching-learning method’, in every semantic category “the highest quantitative rates […] were provided by those students who had not followed any official system of learning and teaching Spanish”, noting a slight superiority in the informants who had combined both methods regulated and non regulated) (Higueras García 2008: 203).

3 Research Objectives

As seen above, at the end of the first bilingual stage, students are required to have a good command of the Spanish language, reaching B1 level of the CEFR at least. There are two modalities for the first stage, and a controversy, when it comes to designating which modality should prevail in the model of bilingual teaching. However, the preference for one or the other is based more on subjective notions than on objective data.

Given this state of affairs, the present study aims to ascertain which of these modalities obtains better results. With this purpose in mind I set out to achieve the following objectives: (1) to provide objective data, based on lexical availability tests and on a lexical-statistical study; (2) to describe quantitatively and qualitatively the available lexicon of learners of Spanish in the two instructional programs in the 16 sampled semantic categories; (3) to compare quantitatively and qualitatively the available lexicons from Middle school and Year Zero; (4) to analyse the influence of the type of school on the lexical availability of two instructional programmes; (5) to determine the structure of the most available lexicon (active vocabulary), distinguishing both common and exclusive lexicon for each type of instructional programme; (6) to analyse the composition of learners’ active vocabulary, according to the level of difficulty of the lexical units.

4 Methodology

This study adopted the methodological steps followed in previous research in Spanish as L2. We also adopted a quantitative and qualitative approach to the study of lexical competence in representative samples of students in the two instructional programmes within the initial bilingual stage. As such, it follows the methodological guidelines of the PanHispanic Project on lexical availability, supervised by López Morales. Material was gathered by way of a written semantic fluency task in which the informants have to produce all the words that they come to their minds about a specific topic – also known as centre of interest or semantic category- for 2 min. The number of categories in the test was up to 16 categories. These categories were: (1) ‘Parts of the human body’, (2) ‘Clothing’, (3) ‘Parts of the house’, (4) ‘House furniture’, (5) ‘Food and drink’, (6) ‘Objects on the table for the meal’, (7) ‘The kitchen and its utensils’, (8) ‘School furniture and materials’, (9) ‘Heating and lighting’, (10) ‘The city’, (11) ‘The ‘Countryside’, (12) ‘Means of transport’, (13) ‘Farm and Garden Work’, (14) ‘Animals’, (15) ‘Games and entertainment’, (16) ‘Jobs and professions’.

The study was carried out in six bilingual Polish schools: Poznan, Lublin, Wroclaw (Middle school); Bydgoszcz, Warsaw-Cervantes, Warsaw-Marti (Year 0). The tests were administered at the end of the school years 2005/06 and 2006/07 in the case of the Middle school and 2010/11 in the case of Year 0. An additional group of 30 tests is scheduled to be done in the Lodz Bilingual Section for the year 2012. That will make the sample for Year 0 up to 120 participants. The distribution of the sample under studied is shown in Table 9.3 as follows:

Table 9.3 Sample distribution
Full size table

In relation to the terminology used, it is important to point out that when speaking of words, in practice it is lexical units that are being dealt with, which can consist of more than one word. An example is found in the answers produced by informants in respond to prompts in lexical availability tasks (e.g., flat plate, pull up weeds, mow the lawn, etc.), an observation that had already been made by Dimitrijévic (1969).

In a statistical study of lexical availability as the present one it is important to distinguish between token and type. Token refers to all the informants’ computable word occurrences, whereas type refers to each different lexical unit.

In line with lexical availability studies in Spanish L1 and L2, we followed the editing criteria suggested in Samper Padilla (1998). The data was electronically processed and recorded on the website Dispolex.com, which provided us with the tools needed to carry out the most common calculations in lexical availability studies: total number of words (tokens) and different words (types) counts, the average number of responses given by informant, and the lexical availability index. This last measure accounts for the number of informants who generated a given word within a semantic category and the position in which they produced the word.

We also provide the cohesion index by applying Max Echeverría’s formula (1991). This index relates the values obtained in tokens and types by dividing the average number of responses given by the informant in each centre of interest by the number of different words. In this way, it can be determined which semantic categories are compact (or closed); that is to say, the degree of coincidence in informants’ word responses.

In order to carry out a qualitative comparison of the Middle school and Year 0 lexicons, I have restricted the comparable lexical units to those with an availability index (a.i.) equal to or higher than 0.1, as was done by Carcedo González (2000). These units are those which, being mentioned more frequently by the participants and being placed higher in the lists, correspond to widely available lexicon: active vocabulary. The need for this limitation is justified by Samper Padilla (1999: 554) in the following way:

After a determined point on the records of availability, a group of words appear which are mentioned by very few participants and which, furthermore, are included in the answer sheets in positions of little relevance. Therefore, we could find ourselves comparing phenomena particular to the language of a specific group of participants or even only one informant, rather than comparing general facts.

As a result, I have obtained lists of widely available lexicon (active vocabulary) from both Year 0 and Middle school. I have listed the common types with an a.i. of >0.1, in the initial ten positions (more active vocabulary), the remaining common lexicon which falls into this interval of availability, and the types with an a.i. of >0.1, (those found on one list are not repeated on the other).

For each centre of interest I classified word responses according to the CEFR levels. To make this classification, I contrast the types obtained to the guidelines included in the chapter of “Specific notions” of the Niveles de referencia para el español, belonging to Plan curricular del Instituto Cervantes (Instituto Cervantes 2007), and the glossaries of Aula Virtual de Español, AVE, del Instituto Cervantes http://www.ave.cvc.cervantes.es), at the levels A1-A2 and B1-B2. When none of these words are found on these lists, I have contrasted specific materials to teach lexicon, in which lists of words according to the CEFR levels are offered, such as Vocabulario, Elemental A1-A2, and, Medio B1, by Baralo et al. (2008, 2009), and textbooks used at the first bilingual stage, such as Club Prisma A1, A2, A2-B1 and B1 (Equipo Club Prisma 2008/2010).

5 Results

5.1 Overall Results: Lexical Availability in the Spanish Bilingual Sections

The general statistical results counted 62,990 words in total, giving an average of 190 words per student and more than 3,936 per category of interest. However, distribution relating to center of interest shows some important differences. ‘Food and drink’ (6,404) and ‘The city’ (6,055) are the most productive categories in terms of tokens, followed by ‘Animals’ (5,233), ‘Body parts’ (5,192) and ‘Jobs and professions ‘(5,033); students retrieved the least amount of words in the centers of interest ‘Heating and lighting’ (2,092), ‘Objects on the table for the meal’ (1,957) and ‘Farm and garden work’ (1,247) (Graph 9.1).

Graph 9.1
figure 1

Total words in the general distribution

Full size image

Now taking types, or different words, into consideration, the sample produced a total of 4,545 types, with an average of 284 types per center of interest. The distribution varies according to the specific category. Those with the highest number of types are ‘Games and entertainment’ (559), ‘The city’ (553), ‘The ‘Countryside’ (488) and ‘Jobs and professions’ (446); with a lower number of types appearing in ‘House furniture’ (160), ‘Body parts’ (140), ‘Objects on the table for the meal’ (122) and ‘Parts of the house’ (120). Furthermore, we can also observe the closeness of the interval dispersion of lexical richness below the average (91 units) as opposed to the widespread interval above the average (208 units). We also can see the categories containing a low degree of lexical richness, ‘House furniture’ (141), ‘The kitchen’ (160), ‘Heating and lighting’ (183), ‘Means of transport’ (184) and ‘Animals’ (187) (Graph 9.2).

Graph 9.2
figure 2

Total of word types in the general distribution

Full size image

Comparing tokens and types, it can be seen that, in line with other studies, the ranges which correspond to the different categories of interest for total words and different words are not proportionate. In this sense, it is striking that categories which are very productive in terms of tokens – ‘Animals’ (3), ‘Body parts’ (4), have considerably smaller type ranges – 8 and 14 respectively; or that ‘Farm and garden work’, with a range of 16 for tokens increases to a range of seven for types (Table 9.4).

Table 9.4 Comparison of ranges for tokens and types
Full size table

The cohesion index indicated that the most compact category is ‘Body parts’ (0.11), followed by ‘Parts of the house’ (0.8), and ‘Animals’, ‘Clothing’ with (0.7). On the other hand, the most open, that is to say, the category with the widest variety of replies is ‘Farm and garden work’ (0.01), followed by ‘Games and entertainment’ (0.02), and a wide group (‘The ‘Countryside’, ‘Jobs and professions’, ‘The city’, ‘Heating and lighting’) with an index of (0.03) (Table 9.5).

Table 9.5 Cohesion index according to centre of interest
Full size table

The words which appear most frequently among Polish students – mentioned by more than 75 % of the participants – are, in decreasing order: coche ‘car’ (97.3 %), perro ‘dog’ (97.3 %), gato ‘cat’ (96.9 %), ojo ‘eye’ (96.4 %), bicicleta) ‘bike/bicycle’ (92.7 %), plato ‘plate’ (92.4 %), (auto)bús ‘bus’ (91.8 %), lámpara ‘lamp’ (90.6 %), cocina ‘kitchen’ (89.7 %), cabeza ‘head’ (89.4 %), mesa ‘table’ (89.4 %; 84.6 %; 76.4 %), profesor ‘professor’ (89.1 %), mano ‘hand’ (88.2 %), pantalón ‘trousers’ (86.4 %), silla ‘chair’ (85.8 %; 81.6 %), camisa ‘shirt’ (85.2 %), cama ‘bed’ (83.4 %), nariz ‘nose’ (83.1 %), avión ‘plane’ (82.2 %), calle ‘street’ (79.5 %), caballo ‘horse’ (77.6 %), and pierna ‘leg’ (75.2 %). All of these words – 22 in all- represent things immediately connected to daily life. By center of interest, the representation is the following: 5 words from ‘Body parts’ (ojo ‘eye’, cabeza ‘head’, mano ‘hand’, nariz ‘nose’, pierna ‘leg’), 4 from ‘Means of transport’ (coche ‘car’, bici(cleta) ‘bike/bicycle’, (auto)bús ‘bus’, avión ‘plane), 3 from ‘Animals’ (perro ‘dog’, gato ‘cat’, caballo ‘horse’) and 3 from ‘House furniture’ (mesa ‘table’, silla ‘chair’, cama ‘bed’), 2 from ‘Clothing’ (pantalón ‘trousers’, camisa ‘shirt’) and 2 from ‘School’ (mesa ‘table’, silla ‘chair’), and 1 from ‘Heating and lighting’, ‘The city’, ‘The kitchen’, ‘Objects on the table for the meal’, ‘Parts of the house’ and ‘Jobs and professions’. More open fields – ‘Food and drink’, ‘The ‘Countryside’, ‘Games and entertainment’ – and more specialised fields – ‘Farm and garden work’ – are not represented.

5.2 Comparison of Year 0 vs. Middle School

5.2.1 Quantitative Analysis

The statistical analysis applied to the data yields a total of 20,407 words in Middle school and 16,498 in Year 0. However, given that the samples compared are not identical – 120 Middle school students to 90 high school students, the average of responses per student in both groups is compared.

Table 9.6 Comparison of word average in the initial bilingual stage
Full size table

As can be gathered from Table 9.6, the average per category of interest is higher in Year 0 (11.46) in comparison to Middle school (10.63) by nearly one word more. In general, to a greater or lesser extent, word production is higher in Year 0 than Middle school in 13 of the 16 centers of interest. The biggest differences favouring Year 0 can be found in “Body parts” (+4.15), ‘Clothing’ (+1.68), ‘Objects on the table for the meal’ (+1.67) and ‘Jobs and professions’ (+1.57). The three centers of interest with a higher production in Middle school are ‘The city’ (−0.34), ‘Food and drink’ (−0.71) and ‘School’ (−1.45). This last category shows a significant difference.

The distribution of the centers of interest coincides quite a lot, as in both educational modalities, the centers of interest found above and below the average are the same, with the ranges coinciding in seven categories: ‘Food and drink’ (1), ‘The city’ (2), ‘Means of transport’ (10), ‘House furniture’ (11), ‘Clothing’ (12), ‘The kitchen’ (13) and ‘Farm and garden work’ (16). In the remaining cases, the ranges are very close, with differences of between 1 and 3 points, although the case of ‘Body parts’ should be highlighted, with a range of 7 for Middle school and 2 for Year 0 (Table 9.7).

Table 9.7 Comparison of word type production in the initial bilingual stage
Full size table

If we focus on a similar comparison of word types, the average per center of interest is also higher in Year 0 (134.24) in comparison to middle school (120.47). The vocabulary is therefore more varied in Year 0. However, some changes must be noted. Here, the number of categories with a higher production in Year 0 reduces to 11, with the biggest advantages in ‘The city’ (+43), ‘Animals’ (+36), ‘Clothing’ (+30) and ‘Objects on the table for the meal’ (+30). After these, ‘The kitchen’ with +25 and the three categories ‘Body parts’, ‘Food and drink’ and ‘The ‘Countryside’ with +24. The advantages favouring Middle school are minimal and of little significance ‘School’ (−8), ‘House furniture’ (−4), ‘Farm and garden work’ (−4) and ‘Heating and lighting’ (−2).

The distribution of the categories also coincide here, as the center of interest which are found above and below the average is the same. With regard to ranges, they are identical in half of the cases: ‘Games and entertainment’ (1), ‘The city’ (2), ‘Jobs and professions’ (3), ‘The ‘Countryside’ (4), ‘Food and drink’ (5), ‘School’ (6), ‘Parts of the house’ (15), and ‘Objects on the table for the meal’ (16). Only the ranges in ‘Clothing’ (12–8) and ‘The kitchen’ (13–9) are different, by 4 points, favouring Middle school, and ‘Means of transport’ (7–13) by 6 points, favouring Year 0.

Until now, an advantage towards the Year 0 group has been demonstrated, but are the differences shown based on a homogenous distribution of data within each group? As can be seen in Graph 9.3, this is not the case. Whilst Middle school gives a normal distribution, around an average of 9.94 words with a standard deviation of 0.89, with a maximum of 10.88 (Poznan) and a minimum of 9.12 (Wroclaw); in Year 0, with the average being 10.90 words, the standard deviation is 1.96, with a maximum of 13.15 words (Warsaw Cervantes) and a minimum of 9.52 (Bydgoszcz).

Graph 9.3
figure 3

Word average according to educational context

Full size image

As such, in relation to the word average, a much higher production can be observed in Warsaw Cervantes with regard to Year 0, and a notable advantage in Poznan as well, in the Middle school. If we disregard these two schools, the other four Warsaw Marti, Bydgoszcz, Wroclaw and (Lublin), either Middle school or Year 0, fairly similar results are found, with an average of around 9.62 words. Two of the schools with Year 0 have a lower average than that of Poznan Middle school with the highest average, (and one of them, Bydgoszcz), has lower results than two of the Middle schools.

A similar situation can be seen regarding the word types in relation to centers of interest and city. Here, the distribution continues to be normal in Middle school, with an average of around 71.3 types and a standard deviation of 7.14, with a maximum of 79.56 (Poznan) and a minimum of 67.25 (Wroclaw). In Year 0, the average is 78.06 types and the standard deviation is 12.31, with a maximum of 90.5 (Warsaw Cervantes) and a minimum of 65.88 types (Bydgoszcz) (Graph 9.4).

Graph 9.4
figure 4

Word type production according to educational context

Full size image

In the case of type distribution in relation to the educational stage, the situation is found to be quite similar to that seen for the word average. More than one division in the two groups, Middle school and Year 0, can be observed, with three heterogeneous groups distinguished: a trio formed by Bydgoszcz (Year 0), Wroclaw and Lublin (Middle school), with around an average of 78.68 word types, and significantly above the others, Warsaw Cervantes with an average of 90.5.

5.3 Qualitative Analysis: Structure of Active Vocabulary

As I have already mentioned in the methodology section, in order to analyse the structure of the most active vocabulary of the two groups under study as well as the convergences and divergences within lexical inventories of educational stages, I have restricted comparable lexical units to those whose availability index is equal or higher than 0.1, discarding any lexias of inferior values. Thus the standard-related facts are compared.

After collecting all available vocabulary with the indices exceeding 0.1, the proportion of types to be compared is relatively small. As a result, the total number of 2,048 and 2,282 lexical units which form the respective inventories of Middle school and Year Zero has been reduced to 286 (14.0 %) and 297 (13.0 %) units (Table 9.8).

Table 9.8 Distribution of word types (a.i > 0.1), according to centre of interest
Full size table

As a result of the examination and one-by-one comparison of lexical units which form the inventories of highly available vocabulary: (active vocabulary) in Middle school and Year Zero different lists were obtained. These lists present common types with a.i. >0.1 up to position 10 (the most active vocabulary), the remaining common vocabulary within this interval of availability as well as the types with a.i. >0.1, which are present on one list and not on the remaining ones- highly available vocabulary which is exclusive to Middle school and Year Zero. All those lists are presented as appendix attached to this chapter. Table 9.9 below displays a quantitative summary of the lists included in the appendix.

Table 9.9 The most available word types on the lists of Middle school and Year Zero
Full size table

A close inspection of such Table reveals a strong uniformity in the lexical units which produce the highest levels of availability for each educational stage; this provides us with a solid common base for comparison. A large amount of these available words coincide on both models for the initial stage, as can be observed in the column “Top 10 common types”. Here, word types completely coincide for ‘Clothing’ (10) and are very similar for ‘Body parts’, ‘The ‘Countryside’, ‘Means of transport and ‘Animals’ (9 types), and ‘House furniture’, ‘Food and drink’ and ‘Games and entertainment’ (8 types). This convergence of lexical repertoires is equally evident in “Remaining common types with an a.i. of >0.1”, with 18 common word types in ‘Food and drink’ and 10 in ‘Body parts’, ‘The city’ and ‘Animals’.

Within common vocabulary with a.i. > 0.1 results show important differences in productivity for the different categories. As opposed to categories with a varied vocabulary – ‘Food and drink’ (26), ‘Body parts’ (19), ‘Animals’ (19), ‘Clothing’ (18), ‘The city’ (17), ‘The Countryside’ (15) and ‘Jobs and professions’ (15) – others present a very limited widely available vocabulary, and even the absence of common types in the positions after the initial 10 with an a.i. of >0.1. – ‘The kitchen’ (10), ‘Heating and lighting’ (6), ‘Objects on the table for the meal’ (5), ‘Farm and garden work’ (2).

In both groups – Middle school and Year Zero–, and for each subject stimulus, the types with an a.i. of >0.1 specific to one of the educational modalities appear, as well as the richest lexicon given by one group or the other, according to category of interest. Year 0 leads in eight categories, showing a significant difference in six of these: ‘The kitchen’ (9), ‘House furniture’ (6), ‘Farm and garden work’ (6), ‘Body parts’ (5), ‘Clothing’ (5), ‘Parts of the house’ (4). On the other hand, the widely available vocabulary in Middle school is higher in five categories: ‘Animals’ (10), ‘The city’ (9), ‘Jobs and professions’ (9), ‘School’ (6), ‘Means of Transport’(4). In each of these cases, the difference is significant, though fields with more varied associations like ‘The city’ and ‘Jobs and professions’ tend to favour a higher number of word types also specific to Year 0. The same circumstances influence on the seven exclusive types in both groups for the category ‘Food and drink’.

If we look at the data from the perspective of the percentage of students who retrieved each word, we find that in Middle school 22 types were produced by more than 75 % of the participants; in the case of lyceum this number increases to 27 types. The types common to both groups are as follows: perro (dog), gato (cat), coche (car), ojo (eye), mesa (table), bicicleta (bike/bicycle), cabeza (head), plato (plate), (auto)bús (bus), profesor (professor), mano (hand), pantalón (trousers), lámpara (lamp), cocina (kitchen), camisa (shirt) and cama (bed). As can be inferred from the above, these were produced by more than 75 % of informants. Next to these word types other five exclusive category types appear on the Middle school list, such as silla (chair), mesa (table), habitación (room) and calle (street); while in the case of lyceum other ten types have been listed, such as avión (plane), nariz (nose), dormitorio (bedroom), libro (book), oreja (ear), camiseta (t-shirt), cuchillo (knife), bolígrafo (ball-pen), agua (water) and caballo (horse).

The percentage of highly available word types retrieved by 75 % or more informants is quite low: 7.7 % in Middle school, and 9.1 % in Year Zero, with significant advantage on the part of Year Zero. However, if we extend the studied interval up to 50 %, the percentages become equal, as a result of which both in Middle school and in Year Zero half of the students produced a little more than a quarter of the active vocabulary: to be exact, 26.5 % in Middle school and 26.6 % in Year Zero.

All in all, it should be remembered that many types are repeated in different categories. Therefore, if we consider the totality of the vocabulary with a.i. >0.1, in the case of Middle school we can find 79 entries corresponding to 35 actual types (27.6 % of the active vocabulary), while in the case of Year Zero we can find 83 entries corresponding to 39 truly different types (27.9 % of the active vocabulary). Therefore the percentage values of the repeated vocabulary within semantic categories are similar.

We can distinguish three groups within such active repeated vocabulary:

  1. (a)

    Common repeated types

    (auto)bús (bus), árbol (tree), caballo (horse), casa (house), cerdo (pig), cine (cinema), coche (car) cuchillo (knife), frigorífico (fridge), gallina (hen), gato (cat), lámpara (lamp), lavadora (washing machine), mesa table), ordenador (computer) pájaro (bird), perro (dog), plato (plate), pollo (chicken), puerta (door), silla (chair), teatro (theatre), televisión (televisión), tienda (shop), tranvía (tramway), vaca (cow), ventana (window).

  2. (b)

    Repeated types exclusive to Middle school:

    armario (wardrobe), bicicleta (bicycle), bufanda (scarf), discoteca (discotheque), flor (flower), fregadero (sink), gente (people), iglesia (church).

  3. (c)

    Repeated types exclusive to Year Zero:

    agua (water), baño (bathroom), cocina (kitchen), cuchara (spoon), escritorio (desk), gallo (rooster), jardín (garden), jardinero (gardener), libro (book), metro (subway), nevera (fridge), olla (pot), oveja (sheep), tenedor (fork), tren (train).

The repeated vocabulary can be grouped in the following categories: 18 word types in the category ‘The ‘Countryside’, 14 in ‘The city’, 13 in ‘The kitchen’, 12 in ‘House furniture’, 10 in ‘Animals’, 10 in ‘School’, 7 in ‘Means of transport’, 6 in ‘Games and entertainment’, 5 in ‘Parts of the house’, 5 in ‘Objects on the table’, 3 in ‘Heating and lighting’, 2 in ‘Food and drink’, 2 in ‘Farm and garden work’, 1 in ‘Jobs and professions’ and 1 in ‘Clothing’.

As for the characteristics of the highly available word types, their comparison with the inventories of “Specific notions” listed in Niveles de referencia para el español (The levels of reference for the Spanish language) under Plan curricular del Instituto Cervantes (the Curriculum Plan of the Cervantes Institute) the Cervantes Institute (2007), as well as their comparison with the vocabulary lists of the most frequently used Spanish language manuals (Equipo Club Prisma 2008/2010) at the initial bilingual stage and with specific materials for teaching vocabulary (Baralo et al. 2008, 2009) shows that this classification is a reliable reflection of the complexity level of active vocabulary and can serve as a manner of establishing correspondence between the level of planned and real lexical command.

With reference to the above, with the exception of ‘Objects placed on the table for the meal’ and ‘Farm and garden work’, in the other 14 thematic centres, the common active vocabulary until position 10 belongs to level A. This situation is repeated in the rest of common types with a.i. of >0.1. in 10 centres of interest (‘Clothing’, ‘Parts of the house’, ‘Food and drink’, ‘The kitchen, School’, ‘The city’, ‘The ‘Countryside’, ‘Means of transport’, ‘Games and entertainment’ and ‘Jobs and professions’).

On the other hand, there appears most words of level B1 in ‘Objects placed on the table for the meal’ and ‘Farm and garden work’ in the first ten positions, as well as in the rest of common types with a.i. of >0.1 in ‘Body parts’, ‘House furniture’ and ‘Animals’.

As refers to the level of difficulty of the word types, exclusive to each group, also the majority belongs to level A. However, also level B types are included on both lists; still, those types of superior level are more frequently present on Year Zero lists. Thus the presence of level B word types is significant (at least two types) in 3 centres of interest (‘The animals’, ‘Games and entertainment’) when compared with 7 centres in Year Zero (‘Parts of the human body’, ‘Clothing’, ‘The kitchen’, ‘The ‘Countryside’, ‘Farm and garden work’, ‘Animals’ and ‘Jobs and professions’).

Table 9.10 summarizes the lexical structure of the lists of learners’ widely available lexicon according to the CEFR levels. Shaded are the levels with higher contribution of vocabulary.

Table 9.10 Widely available lexicon distributed according to levels
Full size table

6 Discussion

The objective of this investigation was to determine which of the modalities of the initial bilingual stage guarantees more efficient and consolidated command of the Spanish language. Objective data based on the results of surveys of lexical availability seem to indicate better lexical competence of Year Zero over Middle school. This was demonstrated by the mean values of the average number of responses given by the informants and the production types with the average advantage of +0.83 words per informant and +13.77 types per centre of interest.

When interpreting these results, it is necessary to take into account a varied range of factors (age, motivation, cognitive capacity, or world experience). All these factors are related to the selected modality at the initial bilingual stage. As for the learners’ age, the results of this study are similar to those obtained by Gallardo del Puerto and Martínez Adrián (reported in Chap. 4) in the case of English vocabulary acquisition, and support the hypothesis held in studies on age-related differences in SLA (e.g., Lightbown 2008; Muñoz 2008), in which it is claimed that older learners are better and more efficient in vocabulary knowledge, especially in school settings. It is also in agreement with late starters’ faster rate of acquisition at the initial stage (Gallardo del Puerto 2007; García Mayo and García Lecumberri 2003; Muñoz 2006). The above examples also seem to support the arguments of Tatoj et al. (2008), concerning the benefit of greater maturity of lyceum students at the moment of learning Spanish as a foreign language (SFL).

The advantage of Year Zero over Middle school is not systematic nonetheless, as slight advantages occur for Middle school in the average number of word responses given by informants for ‘Food and drink’ (+0.71), ‘The City’(+0.34) and especially ‘School’ (+1.45), as in words (types) for ‘School’ (+8), ‘Farm and garden work’ (+4), ‘Heating and lighting’ (+2), ‘House furniture’ (+4), and especially meaningfully in ‘Means of transport’ (+19).

About this particular, the influence of schooling hours and the continued practice of this school-related vocabulary seem to be evident in the case of the Middle school’s advantage in this subject. Similarly, a very plausible explanation for the four average additional words which a student in Year 0 is able to produce relating to the stimulus ‘Body parts’ could be attributed to the interest in the semantic category shown by 16 year old adolescents in comparison to 13 year old pupils age (at which the vocabulary is learnt).

The influence of the ‘sex’ factor, which interacts with the ‘age’ factor, can also be noted on the results, since the sample is mostly female. In my opinion, this explains the greater number of words and vocabulary variety about ‘Clothing’ in Year Zero: +1.68 words per students and +30 types. The interest in ‘Clothing’ and the world of fashion is also stronger in a female student of 16 than in pupils of 13–14 years old age (at which this subject is dealt with in Middle school). The greatest production and variety of the vocabulary in the class entitled ‘Clothes’, which is traditionally associated with the social female role has also been noted in the majority of investigations on lexical availability in the Spanish language (Samper Padilla y Samper Hernández 2006: 51).

In the interpretation of the quantitative data, one must take into account the anomalies shown in the distribution of tokens and types, in the analysis of the sampled schools. Significant differences could be noted in data distribution, which placed the Cervantes Lyceum in Warsaw with +2.27 words per student before the following centre, which surprisingly was a Middle school from Poznań. This tendency was also observed regarding types, in which the same schools ranked first and second: the Cervantes Lyceum in Warsaw with 90.5 types, and Poznań Middle school with 79.56 types. In the lower distribution range low results of the Lyceum in Bydgoszcz stood out: 9.52 words per student and 65.88 types on average, which situated it in the result range which was characteristic for other centres from the Middle school group. These results suggest the positive influence of additional factors, such as the teaching work itself, the used textbooks or the selection of students.

From the point of view of the objectives of teaching Spanish at the initial bilingual stage, it attempts to equip the student with lexical competence which will allow him to engage in effective communication on the level of an independent language user. This is the reason for the importance of analysing active vocabulary of a student learning Spanish as a foreign language at the end of the first bilingual stage, which in this case is identified with the available vocabulary, the indices of which exceed 0.1.

In this study, the reduced values of the magnitudes of highly available vocabulary in the Middle school −286 (14.0 %) and Year Zero −297 (13.0 %) point to what was already noted by Carcedo González (2000: 160): “much more limited magnitudes of Spanish vocabulary acquired by foreigners are accompanied by major coincidence of answers”. In fact, the percentage values in Poland, especially in the case of Year Zero, show striking similarity to the value of 13 % which was obtained by Carcedo González (2000) upon having limited the vocabulary of pre-university students in Finland to the a.i. range >0.1.

In Polish bilingual sections, the productivity differences between lexical categories within the common vocabulary with the a.i. > 0.1 are characterised by the same distribution of the words: those centres of interest which appear with increased productivity have been located above the average and those with less common vocabulary have been located below the average. If we look at the indices of cohesion obtained in the general distribution, we can see that such distribution includes both the closed-system centres (‘The human body’, ‘Clothing’ and the open ones ‘The city’, ‘The ‘Countryside’). Therefore it can be observed that the results obtained by Polish students are similar to those reported in studies in the Hispanic world; both the more and the less productive centres at the initial bilingual stage in Poland correspond to those which appeared above and below the average in the investigations under the PanHispanic Project of Lexical Availability (Samper Padilla et al. 2003: 57–60). The fact that the results are similar both in Spanish native speakers and Polish students learning Spanish as a foreign language may be attributable to less practise and less frequent use of the vocabulary related to these areas rather than to learning differences.

As for the word types produced by more than 75 % of students, we could say that it was a quite stable vocabulary, given that all words common to both modalities of the initial bilingual stage were present on the list of types produced by more than 75 % of student of the general sample. In the same manner two types from the Middle school (silla ‘chair’ and calle ‘street’) as well as three types of Year Zero (avión ‘plane’, nariz ‘nose’ and caballo ‘horse’) were present on that list.

Type repetitions on the active vocabulary lists did not go unnoticed, given the limited range of the studied sets − 286 types in the Middle school and 297 types in Year Zero – in contrast with the considerable percentage of produced repeated vocabulary, which in both cases exceeded 27.5 %. The three groups of repeated types – 27 common types, 8 types exclusive to Middle school and 15 repeated types exclusive to Year Zero – depict the phenomenon already noted by Samper Padilla et al. (2003) in the case of words such as silla ‘chair’, which was present in 6 centres on the lists of Gran Canaria, or jardín ‘garden’ present in 7 centres on the availability lists of France and Acadia presented by Mackey (1971: 417). All that shows that semantic classes proposed by Gougenheim et al. (1956) show frequent intersection areas, or even the areas of inclusion, as in the case of ‘Objects on the table for a meal’ in relation to ‘The kitchen and its utensils’.

In the analysed sample these intersections are present mainly in the centres with very broad associations, such as ‘The ‘Countryside’ (18 types) and ‘The city’ (14), as well as in directly related centres, such as ‘The ‘Countryside’ (18) and ‘Animals’ (10) or ‘The kitchen’ (13), ‘House furniture’ (12), ‘Parts of the house’ (5) and ‘Objects on the table’ (5).

Finally, the classification of highly available types according to the Reference levels of the Common European Framework of Reference for Languages (the Council of Europe 2001) shows that common active vocabulary belongs to fundamental lexis which is necessary for communication. The organization of this lexis on the lists of common available vocabulary shows that the words classified under level A in the reference inventories for teaching Spanish as a foreign language come first on those lists. Only exceptionally the majority of vocabulary of level B appears on any of the first ten positions in two centres of interest with a more specialist vocabulary (‘Objects on the table for the meal’ and ‘Farm and garden work’). The majority of word types of the rest of available vocabulary also belong to level A; however the vocabulary of the next level (B) appears more frequently in ‘The human body’, ‘House furniture’ or ‘Animals’, although it is always preceded by a significant number of level A types.

As for the exclusive word types to each group, it should be noted that Year Zero manages to acquire richer vocabulary with frequent B level types in up to 7 centres of interest when compared with 3 centres of interest in the case of the Middle school. This brings Year Zero closer to the final objective of the initial bilingual stage, which consists in reaching B1 level. In this way the acquisition of such vocabulary would comply with the criteria of communicative profitability and it will be required in order to perform communicative functions and to develop oral and written texts under the expected level at the end of the initial bilingual stage.

In any case, although Year Zero presents an advantage over the Middle school in attaining this objective, the contribution of level B vocabulary is scarce or even none in as many as 9 centres of interest (‘Parts of the house’, ‘House furniture’, ‘Objects on the table for the meal’, ‘Heating and lighting’, ‘Food and drink’, ‘School’, ‘The city’, ‘Means of transport’ and ‘Games and entertainment’). As can be inferred from the findings in this study, some of those thematic categories with lexical deficiencies are vital for the student who shall perform a role of a social agent, an inter-cultural speaker and an autonomous learner with a view to level B, which he is expected to have reached when starting the second stage of bilingual education.

7 Conclusions: Lexical Availability at the End of the Initial Bilingual Stage

In Poland there are two coexisting modalities of the first educational stage in Spanish bilingual sections which follow the same curriculum and cover the same number of teaching hours: one of them is extensive and includes 3 years in the Middle school, the other one is intensive and is conducted in the form of Year Zero at the lyceum. The objective of this study was to find out which of those modalities equips the student with better lexical competence with a view to reaching B1 lexical competence level, required for the second bilingual stage.

To this end a quantitative and qualitative comparative study of lexical availability was conducted, on the basis of the broad sample of students in both modalities (120 at the Middle school and 90 at Year Zero). The results clearly indicate that the Year Zero group has been better prepared for the second stage, since it has reached superior values in terms of the words average, word types, and the quantity and quality of highly available vocabulary.

However, regardless of Year Zero superiority in terms of overall results, the detailed study based on the centres of interest showed slight advantages on the part of the Middle school sample in some centres, which were quite significant in terms of words average in ‘School’ and in terms of types in ‘School’ and ‘Means of transport’. This could indicate that in the case of some semantic categories longer time of schooling and contact may be an advantage to their fixation in learners’ mental lexicon. Such a situation took place regarding the vocabulary related to the school field, where 3 years of continuous practice worked in favour of the Middle school group.

In any case, the study results point at the covariation of the educative modality factor, taking into account other accompanying social factors, such as age or sex. This is especially visible when considering the interest paid by a teenage boy aged 16 to the vocabulary of ‘The human body’ or a 16-year old girl to the vocabulary of ‘Clothes’. In one case the ‘age’ factor and in another the ‘sex’ factor increases the advantages of the Year Zero modality with respect to those thematic stimuli. In the same manner, significant differences in the area of lexical availability between both types of educational schools, regardless of the modality followed, also point to other additional factors, such as the teaching work or the selection of students.

Finally, the present study corroborates the impact of the lexical availability task on the obtained results. This is shown in the high coincidence of Polish results with other investigations in the Hispanic world (Samper Padilla et al. 2003) or in the area of teaching Spanish as a foreign language (SFL) (Carcedo González 1998, 2000) or teaching English as a foreign language (EFL) (Germany and Cartes 2000; Jiménez Catalán and Ojeda Alba 2009, 2010). The coincidence of results is particularly evident regarding the productivity distribution of centres of interest or in the area of repeated vocabulary, which can be explained by means of the relation and superimposition of vocabulary under some thematic stimuli.

The final part of the study has been devoted to the description of the structure and composition of highly available vocabulary (active vocabulary). It has been possible to define a quite stable vocabulary common to both modalities, mainly level A vocabulary (basic one), above all in the first ten positions. As for the vocabulary exclusive to each modality, while in the case of the Middle school the majority of types also belong to level A, in the case of Year Zero, level B types (intermediate ones) appear more frequently, although not in all centres of interest. This tendency seems to give advantage to Year Zero students over Middle school with a view to the second stage of bilingual education. However, both of them present lexical gaps in certain thematic categories, if reaching B1 level (an independent user) of lexical competence is concerned.

The study of the structure of the highly available lexicon in relation to CEFR opens new lines for future studies. One of these lines would be tracking lexical items from the inventories of levels A1, A2 and B1, absent in word lists produced by the bilingual students. It would be also advisable to check the correlation between the CEFR levels and the available lexicon produced by Spanish native speakers, since the ultimate goal of learning a language is to ensure communication with native speakers.