Reading, as one fundamental of early literacy, is a multi-layered process that incorporates phonological and orthographic knowledge. Since reading activities begin early in the course of formal education, children must develop their word-decoding strategies to achieve optimal reading speed, which leads to reading comprehension. Contemporary, inquiry-based approaches to teaching require a certain amount of student independence in reading and research. To meet such requirements, mastery of reading is crucial. A large body of research suggests that phonological sensitivity plays a major role in early reading development, both in monolingual and plurilingual settings. However, orthographic characteristics of different languages often determine which phonological units are more salient for successful reading: whole words, syllables, parts of syllables, or phonemes. In this case study, phonological sensitivity is observed as a comprehensive construct that includes a number of skills ranging from implicit sensitivity to larger phonological units to a more explicit awareness of smaller phonological units that can be blended, segmented, and combined in various ways. The main phonological components described in this study are words, syllables, onsets, rimes, and individual phonemes. Another comprehensive construct used in this study is plurilingualism, and it is described as a set of one’s competencies in more than one language and the ability to shift between those languages to facilitate communication and to adjust to specific situations. Past research in phonological sensitivity conducted with groups of English-speaking monolingual children resulted in a hierarchical model extending from sensitivity to larger phonological units to more advanced sensitivity to smaller phonological units. However, such sensitivity does not necessarily follow a linear path; depending on children’s abilities and the nature of literacy instruction, certain overlaps between phases are likely to occur. Another dimension of the developmental model is linguistic complexity. More precisely, the non-linear development of phonological sensitivity is further affected by the orthographic characteristics of the language in question. For instance, word-decoding in an orthographically transparent language results in early phonemic awareness, which is commonly recognized as the key to successful reading in alphabetic scripts. On the other hand, early reading in less orthographically transparent languages typically entails multiple word-decoding strategies to achieve the same outcome. Many researchers have shown interest in the development of phonological sensitivity of children attending bilingual programs in state schools or international programs, mostly in English as their second or third language. A growing body of research has demonstrated that plurilingual and monolingual children follow a similar developmental path of phonological sensitivity that has been recognized as a reliable predictor of reading in both cases. However, some temporary crosslinguistic influences have been recorded in plurilingual children whose linguistic diversity and various literacy instruction, as consequences of frequent family migrations, also interfere with quantitative research designs, and often yield inconclusive results. Nonetheless, numerous case studies have provided valuable qualitative insights into plurilingual children’s phonological sensitivity development, but their results are usually inadequate for generalizations. This thesis describes a case study designed to contribute to the existing qualitative research by presenting the Croatian context and the role of the Croatian language in children’s language repertoires. The main objective of this case study was to investigate how deliberate development of phonological sensitivity influences early reading in plurilingual first-graders attending an international primary school. Another objective was to conduct the explicit teaching of the International Phonetic Alphabet (IPA) and observe its role in developing students’ awareness of similar sounds appearing in multiple languages. The study was designed to answer three research questions and explore four hypotheses: RQ1: Which reading strategies emerge while reading in English and while reading in Croatian, and what are the differences? RQ2: Will plurilingual first-graders use the graphemes of English, the language of their education, or the graphemes of Croatian, their host-country language? RQ3: To what extent will plurilingual first-graders learn the symbols of the International Phonetic Alphabet taught within the additional phonological training? H1: There will be some differences between the students who had received explicit literacy instruction before first grade and the ones who had not received explicit literacy instruction. H2: As a consequence of explicit teaching, the plurilingual first-graders will use worddecoding strategies independently. H3: Native speakers of Croatian will predominantly use Croatian graphemes to write the designated sounds, since Croatian is also the language of their environment. H4: The explicit teaching of the International Phonetic Alphabet will facilitate the development of phonemic awareness. A class of thirteen plurilingual first-graders participated in the study, and the information was collected during one school year. Six research instruments were applied: listening tests and verbal protocols, one-minute reading aloud pre-tests and post-tests in English and Croatian accompanied by verbal protocols, a phonological sensitivity pre-test and post-test, a semistructured interview with students, a semi-structured interview with parents, and school documents (e.g., unit logs, unit planners, teacher comments, student application forms). Along with the data collection points, additional training intended for developing phonological sensitivity was designed for participants, parts of which were dedicated to teaching the IPA. Both interviews and the information retrieved from school documents indicated that each participant had quite a substantial language biography, and additionally, participants were rather diverse when mutually compared. Due to their private situations and work requirements, participants’ families often changed their places of residence, thus exposing their children to new languages, cultures, and environments. The most noticeable common denominator among all participants was English as the language of education. The initial listening tests revealed certain differences between the participants who had attended preschool in English and the others, which was in line with expectations. The two participants who had received explicit literacy training in English kindergartens demonstrated excellent orthographic knowledge, but they were unable to name the initial sounds in words. Instead, they offered letter names and were entirely focused on spelling. On the other hand, participants who had not received explicit literacy instruction prior to first grade were more open to naming initial sounds in words. However, several variations were documented in terms of phoneme pronunciation due to probable mother tongue influences. Following this phase, one participant was excluded from the study after being diagnosed with dyslexia, and one participant moved to another country. From that point on, there were eleven participants. As hypothesized, the explicit teaching of word segmentation strategies facilitated the transition from early sensitivity to spoken word structure to phonemic awareness. More specifically, initial reading in English, an orthographically opaque and inconsistent language, predominantly included whole-word recognition strategies. In due course, the majority of students engaged in word segmentation, thus combining phonological and orthographic knowledge. On the other hand, initial reading in Croatian, as an orthographically more transparent and consistent language, revealed that the rather clear letter-sound correspondence was recognized as an efficient word-decoding strategy. The end-of-year reading test showed that reading in Croatian was faster when blending syllables than when blending phonemes. Overall, the number of word-decoding strategies in English increased considerably over time, whereas the number of word-decoding strategies in Croatian remained low but sufficient for successful reading. The indication of crosslanguage transfer of phonological sensitivity abilities was found in several student reports, stating that they utilized the “Croatian way of reading” when decoding English words. The average reading speed in both English and Croatian is slightly lower in this study, but still comparable to the average reading speed of monolingual speakers of both English and Croatian, but with minor variations. The phonological sensitivity of participants notably increased during the school year and was positively associated with reading speed, but the design of this study is not suitable for generalization. Moreover, the explicit teaching of the spoken word segmentation appeared to facilitate the awareness of the spoken words of the participant who, instead of initial sounds in words, offered letter names at the beginning of the study. Her phonological sensitivity test results yielded the largest difference between the two testing points, and she climbed from the bottom of the scoreboard to the very top. Other participants exhibited slight downward or upward changes on the same scoreboard. The listening tests dedicated to specific sounds appearing in multiple languages showed that plurilingual first-graders predominantly opted for the graphemes belonging to their language of education, i.e., English, when asked to write them down using the manner they considered appropriate. In order for native Croatians to opt for Croatian graphemes in the same task, the role of the Croatian language needed to be manifold: mother tongue, family language, and host-country language. The primary purpose of the International Phonetic Alphabet in this case study was to illustrate spoken sounds regardless of the language and the spelling patterns producing them, and to improve students’ phonemic awareness as part of the additional phonological training. After completing the additional phonological training, all students took two tests measuring the extent to which they had learned the IPA symbols. The results revealed a high level of success when it comes to the recognition of the IPA symbols. However, the test of independent writing of the IPA symbols resulted in low scores. Among the IPA symbols that do not resemble letters, the least accessible ones were ə and ŋ, whereas the most accessible symbols were ʒ, ʃ, w, and θ. Furthermore, the recognition of the similar sounds appearing in words across languages (English excluded) revealed students’ ability to compare and analyze various sounds appearing in their language repertoires, which also contributes to the development of their phonemic awareness, i.e., metalinguistic knowledge, and the awareness of their plurilingualism. Although the sample size is insufficient for generalizations, this study fills the gap in contemporary research on young plurilinguals who develop their early literacy in English, a language that is not their mother tongue but is significant for their formal education. The results also indicate that existing models of phonological sensitivity development should include the third dimension containing multiple languages that could be used as sources of cross-language transfer. Along with that, the detailed descriptions of each participant demonstrate how linguistically and culturally diverse a relatively small class of first-graders could be, which can interfere with research designs due to some variables that might be difficult to control for in causal models.
