Abstract

The positive effects of dialogic reading on children’s language and early literacy skills are well-known. While many researchers have designed intervention studies at home or at school, predominantly in English-speaking or in monolingual settings, there is a dearth of naturalistic studies and those that provide insights into the process of dialogic reading in multilingual contexts. The present qualitative paper investigates the case of Luxembourg where a multilingual program has been implemented over the last years that requires educators to familiarize children with Luxembourgish and French and value their home languages, for example when engaging in literacy activities. The paper draws on interviews and video recordings over a period of ten months and examines the ways in which educators in two private day care centers read dialogically with children aged two to four. The findings show that the dialogic reading practices differed in length and frequency, with the educators in one center initiating more talk in more languages, giving more feedback, and modelling language more than those in the second center. Furthermore, the educators in one center frequently translanguaged to facilitate communication and comprehension. The practices were influenced by the educators’ beliefs and understanding of the purpose of reading. These results highlight the need for professional development on the purposes of literacy and the effective use of language and strategies to engage children in dialogic reading.

Keywords

Dialogic Reading, Multilingual Education, Translanguaging, Luxembourg, ECEC, Beliefs

Share and Cite:

1. Introduction

A rising number of children in Europe and worldwide grow up bilingually or multilingually (Bergeron-Morin et al., 2023). Professionals in Early Childhood Education and Care (ECEC) have been reported to find it difficult to handle this growing language diversity (Carrim & Nkomo, 2023). In addition, they have to adapt to frequent curricular changes. For example, practitioners in many countries are required to offer children opportunities to develop early literacy skills (Seltzer et al., 2020), learn a foreign language or encounter multiple languages (Thieme et al., 2022). These expectations are combined in the multilingual program in Luxembourg, a small country in Europe with three official languages, Luxembourgish, French and German. Since 2017, educators are required to familiarize children with Luxembourgish or French (depending on the dominant language of the center) and value their home languages, for example by inviting parents to come and read in these languages (MENJE & SNJ, 2021). The question arises how professionals in Luxembourg and elsewhere can help young children develop language and literacy skills in multiple languages.

Frequent and high-quality reading has been shown to promote language and early literacy skills in one (Sénéchal et al., 2017) or two languages (Farver et al., 2013; Pico & Woods, 2023) and predict children’s later literacy skills and academic achievement (Skibbe et al., 2011; Zucker et al., 2013). Dialogic reading is one type of shared reading where adults encourage children to talk about pictured materials and give them feedback (Kirsch et al., 2023). This method, which results in the development of language and early literacy skills (Wasik et al., 2016), has been successfully used in classrooms in many countries, although mainly in English-speaking countries with monolingual speakers or with children learning English (Pillinger & Vardy, 2022; Whitehurst et al., 1988). Little is known about the ways in which professionals read with emergent multilinguals (Carrim & Nkomo, 2023) although it is known that some ECEC professionals use their entire semiotic repertoire, that is, translanguage, in literacy activities (García & Kleifgen, 2020).

The present qualitative paper explores the varied ways in which 12 educators in two private commercial centers in Luxembourg read dialogically with children aged two to four. In Luxembourg, one third of the children aged zero to four speak one language at home, another third speaks two, and the final third speaks three or more (SNJ, 2023). The residents’ language diversity is one of many reasons that the government called for multilingual education in ECEC in 2017 (MENJE & SNJ, 2021). In a recent survey, educators indicated using mainly Luxembourgish and French but also some Portuguese, German, English or Italian, both throughout the day and in early literacy activities (Kirsch & Aleksić, 2021). Translanguaging, the dynamic and fluid move between and beyond languages, is considered a legitimate practice. Of the educators surveyed, 65% reported reading daily and 43% conversing daily with children about the stories. Thus, dialogic reading does not appear to be part of the professionals’ everyday practice. To better understand dialogic reading and the language use during these moments, this article contrasts the differing practices of two ECEC centers we called Earth and Air. Whereas the educators of the former were reported to share a social practice view of language and literacy and draw on social constructivist learning theories, the latter are influenced by a skills-based approach largely informed by behaviorism (Kirsch & Bergeron-Morin, 2023; Kirsch & Hornberger, 2024). This follow-up article provides insights into educators’ reading strategies in one or more languages and shows that monolingual and translingual literacy practices are shaped by the educators’ understanding of language and literacy and result in children participating in different ways. The findings inform policymakers, researchers and professionals of the benefits of translingual dialogic reading and the reasons behind differing practices.

2. Literature Review

Early literacy encompasses children’s skills in dealing with narratives and print before they learn to read and write. Early literacy activities at home or in ECEC include, for example, simple language games, finger rhymes, storytelling, shared reading and first scribbles (Hartmann et al., 2009). The present section reviews empirical findings of dialogic reading (one type of shared reading) and translanguaging in early literacy activities and shows that dialogic reading involving more than one language contributes to the development of language and literacy skills.

2.1. Dialogic Reading

For children to develop language skills during interactive reading sessions, adults need to engage them in high-quality dialogue and be responsive to their utterances. As adults do not automatically use strategies to promote interaction and model language, Whitehurst et al. (1988) developed an intervention called dialogic reading where parents and teachers learned to prompt children, evaluate, expand and repeat utterances (PEER sequence) and ask completion prompts, recall prompts, open-ended questions, wh- questions or distancing questions (CROWD questions). Intervention studies and meta-analyses have continuously shown the positive impact of dialogic reading on children’s improved listening comprehension and increase in vocabulary (Dowdall et al., 2000; Pillinger & Vardy, 2022; Whitehurst et al., 1988, 1994). It is noteworthy that these intervention studies took place in ECEC settings and homes which may differ in the intensity of dialogic reading or the size of the groups. In addition to the promotion of language and narrative skills, dialogic reading with monolingual children in ECEC settings was found to develop early literacy skills (Wasik et al., 2016), which, in turn, facilitate the process of learning to read and write and predict later literacy skills, including alphabet knowledge and print awareness (Zucker et al., 2013).

While the above-mentioned researchers measured the effects of dialogic reading in one language, other studies explored teacher-related and child-related factors that influence reading strategies. Similarly to Mihai (2021), Adam & Barratt-Pugh (2023) reported a relationship between reading and group size in their naturalistic study in Australia. Children engaged more when the teachers read to small groups. The researchers concluded that teachers may only be able to involve children in extended conversations that go beyond the text and stimulate abstract thinking and vocabulary learning when they read to small groups and for the recommended time of 15 minutes. Furthermore, it was found that teachers used more labelling with younger children and asked older children more open-ended questions (Mol et al., 2008). Given that learning is affected by the instructional language and children’s language background, several studies examined the effects of dialogic reading on children learning a second or foreign language. For example, a meta-analysis with emergent bilinguals in the USA revealed that dialogic reading led to language development in Spanish and English when the readers provided enough input in Spanish (Pico & Woods, 2023). A study in China indicated that young pupils learning English increased their vocabulary and ability to make longer and more complex sentences through dialogic reading (Hui et al., 2020). By contrast, the use of multiple languages during dialogic reading has received little attention.

2.2. Translanguaging in Literacy Activities

Translanguaging is commonly understood as a person’s deployment of their entire semiotic repertoire to communicate and learn (García, 2023). While bilingual/multilingual children draw on all communication channels and combine them flexibly to make and express meaning outside school, their resources are frequently limited to the majority language in the classroom. This decreases their learning opportunities and results in lower achievements. Some scholars have, therefore, called for multilingual pedagogies which challenge monolingual ideologies and practices, recognize diversity, and attempt to leverage the students’ semiotic repertoire for learning (García & Kleifgen, 2020). Informed by social constructivist learning theories, these pedagogies call for high-quality interactions where teachers and learners co-construct knowledge in transglossic arrangements. In other words, translanguaging is a pillar of these pedagogies. Owing to their ideological dimension, they are not easy to implement, and many teachers appear to use translanguaging as a pedagogical strategy rather than a pedagogy (Bonacina-Pugh et al., 2021). The following sections review research findings on pedagogical translanguaging in early literacy activities.

In the United States, Gort, Sembiante and their colleagues reported that teachers used translanguaging strategically in read-alouds and show and tell activities in dual language classrooms (Pontier & Gort, 2016; Sembiante et al., 2023). The teachers moved between Spanish and English and were found to model language use, repeat, clarify, translate and ask questions to promote meaning-making (Pontier & Gort, 2016). ECEC professionals in Luxembourg translanguaged in a similar way when reading to emergent multilinguals (Kirsch & Bergeron, 2023). They used their verbal resources (e.g., features of several languages) and non-verbal resources (e.g., mime, gesture, posture) dynamically, translated, and “home languaged”, that is, briefly switched from Luxembourgish to a child’s home language to talk to this particular child, thereby remaining in a monolingual mode (Kirsch, 2021).

Several studies established the outcomes of translanguaging. Authors hold that translanguaging helps children communicate, socialize and mark their multilingual identity (Carrim & Nkomo, 2023; Seltzer et al., 2020; Sembiante et al., 2023) and deepen text comprehension (García & Kleifgen, 2020). Furthermore, Vaish (2021) and Velasco & Fialais (2018), who examined translanguaging with bilinguals in Singapore and France respectively, found that the teachers who used two languages helped the learners develop vocabulary, grammar and metalinguistic awareness by working across English and Malay and German and French respectively. The children in Singapore drew on their entire language repertoire when making sense of English texts and those in France when reading texts in German and French. The teachers in the following two studies had learned to value children’s multilingual and multicultural backgrounds, create meaningful activities and resources, and provide children with opportunities to use the school and home languages in meaningful ways. Based on an integrated literacy and play-based intervention in ECEC, Moses & Torrejon Capurro (2023) found that the children used Spanish to state social expectations and negotiate identities while also mobilizing the newly learned English vocabulary to express themselves and help others do so when playing. Similarly, Seltzer et al. (2020) reported that children used culturally relevant props as well as multilingual resources and communicated in English and Spanish when enacting and transforming familiar stories.

More recently, Carrim & Nkomo (2023) explored the feasibility of using translanguaging pedagogies to develop multilingual learners’ literacy skills in English in the Foundation Phase in South Africa. The findings of their systematic review confirmed that these pedagogies help professionals engage learners, increase their participation, develop listening and speaking skills in several languages, deepen reading comprehension and foster writing skills.

In summary, this literature review has outlined the procedure of dialogic reading and its benefits and pointed out some limitations: most findings came from interventions with monolingual children in English-speaking countries; there is a dearth of naturalistic studies, and few explore the use of multiple languages during dialogic reading. As a result, I reviewed findings on translanguaging and its positive effects in early literacy activities with young learners. However, these studies did not focus specifically on dialogic reading. More understanding of dialogic reading practices with emergent multilingual children seems crucial as many children are bilingual or encounter additional languages in literacy activities in ECEC. I therefore ask the following research questions:

1) In what ways do educators in two ECEC settings read with children aged two to four?

2) In what ways do they use languages in dialogic reading situations?

3. Methodology

This article draws its data from the larger mixed-method study COMPARE (2020-2023) which explores multiliteracy practices and partnerships between parents and ECEC professionals in Luxembourg. The research project was approved by the Ethics Review Panel of the University of Luxembourg (ERP 19-050).

3.1. Sample

The participants are 12 educators of two private commercial centers we have named Earth and Air. The two managers and two educators had participated along with 34 educators of 16 public and private day care centers in a professional development course on literacy activities and collaboration with parents, offered in May 2020 by the COMPARE team. The participants discussed dialogic reading, familiarized themselves with interaction-promoting strategies (e.g., asking questions) and language modelling strategies (e.g., elaboration, corrective feedback) and analyzed their own strategy use in self-recorded literacy activities. At the end of the course, they could volunteer to take part in our observational study starting in September 2020 and lasting over one academic year. Some private centers volunteered despite the Covid-19 pandemic.

We chose settings based on their previous experiences with either early literacy activities or collaboration with parents while also ensuring the centers in our sample came from various regions and included diverse languages. The educators in the Luxembourgish-dominant center Earth wished to further develop the range of literacy activities while those in the French-dominant center Air had established daily literacy practices but aimed to improve collaborative practices. In each center, all educators working with children aged two to four took part. They had various qualifications, but at least one member had a university degree and had worked for longer than ten years. While the educators in Earth spoke predominantly Luxembourgish to the children, they also used French, Portuguese or German with children of these home languages (Kirsch & Bergeron-Morin, 2023). By contrast, the educators in Air communicated predominantly in French apart from a few occasions where they addressed children in English or Spanish. Like in Earth, most children aged two to four were bilingual. Table 1 provides an overview of the participants.

Table 1. Overview of the participants.

3.2. Data Collection

To ensure the high quality of the data collection, the researchers used multiple methods. In this qualitative paper, I combine observations and interviews. Three researchers who closely collaborated regularly visited the settings over one academic year to observe interactions around literacy. The educators and the number of monolingual or emergent multilingual children who participated in the literacy activities differed in each observation. We often observed one or two educators read to about eight children in Earth and one adult read to 16 in Air. Other educators were nearby, as by law, each educator can be responsible for up to 10 three-year-olds. In total, we observed and video-recorded 12 reading activities lasting a total of 2 hours 33 minutes in Earth and 28 lasting 2 hours 56 minutes in Air. To ensure the trustworthiness of our data, we organized member check sessions with the educators in October 2021. We presented our understanding of their literacy practices and received feedback which we wrote up in fieldnotes. Additionally, we carried out four interviews in Autumn and Spring with the managers and the group leaders of the children. Apart from enquiring about the frequency of reading and children’s engagement, we asked the educators to comment selected video-recordings to gain understanding about their actions and the reasons behind the practices. A total of 7 hours 45 minutes was audio-recorded.

3.3. Data Analysis

The observations provided insights into the process of dialogic reading as well as the ways languages were used. Most videos were transcribed, and contextual information and non-verbal communication acts (e.g., pointing) were added. All other videos were described in detail. Information about the length of the activity, the type of books (e.g., fiction, age range), translanguaging, as well as the participants were included in a table. The subsequent analysis of the interactions during dialogic reading was influenced by Seedhouse (2005)’s conversational analysis based on a sociocultural perspective, as well as previous work on language-promoting strategies (Kirsch, 2021). To examine the speech acts of the adults and children, I identified strategies to promote understanding and meaning-making (e.g., explanations, translation, connections), interaction-promoting strategies (e.g., questions) and language-modelling strategies (e.g., expansion, corrective feedback). This analysis was discussed with the team members and the educators in the member check meetings. Once validated, I listed the reading strategies in each observation in tables and examined them in relation to the PEER sequence (prompt, evaluation, expansion, repetition) and the CROWD questions (e.g., completion, recall, opened-ended, wh- questions, distancing questions) to identify similarities and differences. Using tables, I compared the strategies across languages and centers. Table 2 and Table 3 in the following section provide insights into the range of strategies and the differences between the settings.

To triangulate the data, relevant parts of the interviews (transcribed verbatim) were subjected to a thematic analysis (Braun & Clarke, 2006) which focused on the educators’ explanations of their practices (e.g., duration), changes of practices (e.g., reading more or less), purposes of reading (e.g., participation) and beliefs (e.g., skills) as well as children’s interest and engagement (e.g., reading on their own) and language skills (e.g., lack of competence).

4. Findings

All educators indicated that they read daily in the morning and before nap time, but the frequency of reading and the length of the sessions differed, with the educators in Air reading more frequently but spending less time dialoguing with children than those in Earth. A reading could last more than 15 minutes in Earth but half of this time in Air. In what follows, I will firstly present some examples and then an analysis of the readings and show that the practices depended on the educators’ understanding of dialogic reading and affected children’s participation. All first names mentioned below are pseudonyms.

4.1. Examples of Monolingual and Translingual Dialogic Reading

The educators’ dialogic reading practices differed in relation to the reading strategies as well as the language use. The following representative example in Air presents some of the strategies deployed during mainly monolingual reading events.

“The tiny king” in Air

In January 2020, the educators read the story “Le tout petit roi” (“The tiny king”) twice a day for a week. On 13^th January, the following children (home language(s) in bracket), all familiar with the book, participated in the reading: Grégory (French), Mona (French), Sancha (Spanish, French), Dean (English), Cecilia (French, Italian) and Julia (French). The participating educator Ms. Amandine spoke French and English and knew some Italian and Spanish words.

Ms. Amandine opened the book and read (literally translated from French) “The very small king lives in a very large castle”, thereby emphasizing the word “very” and making a large movement with her arm. She then turned the book towards the children to show them the picture of the castle. The following conversation took place.

Ms. Amandine read as if the children had no prior knowledge of the book and asked similar closed and completion questions to the previous day. She frequently repeated correct answers and praised. For example, when children subsequently labelled green, orange, yellow, black and white, she repeated the answers and praised children. By contrast, when English-speaking Dean muttered something incomprehensible, Ms. Amandine did not react but responded “There are many colors in the castle”, before reading the next sentences about the small king and the big soldier.

This second excerpt shows that Ms. Amandine continued to ask closed questions and repeat answers. She added that the soldiers are knights without providing further explanations, which appeared to confuse Sancha. She only understood when the educator pointed to the soldiers and uttered “here” in Spanish. Ms. Amandine then read the next page and asked children to name the food items on the picture. The three-year-olds identified chocolate, apples, potatoes, fish, mushrooms, croissants, ham, carrots, a lobster, eggs and cheese. Ms. Amandine repeated and praised as before. When English-speaking Dean uttered something incomprehensible a second time, Ms. Amandine ignored his attempt to express a food item even though she regularly told children stories in English including some on food. Over the next five minutes, children were asked once to label food, five times to identify colors and four times to state whether something was big or small. At the end of the story, Ms. Amandine invited children to build a castle “like the one in the book”.

Table 2 details the PEER & CROWD strategies deployed in each language in each line of the excerpts. For ease of visualization, the prompts are split into non-verbal and verbal (i.e., the questions). Overall, and typically for all reading events in Air, the educator used few strategies to stimulate talk. She frequently directed children’s attention to a picture which they then described in one or a few words. She made little or no distinction between children who spoke French at home and those who did not.

Table 2. Overview of the mainly monolingual dialogic reading strategies in Air.

“The apple mouse” in Earth

Literacy practices in Earth differed on account of the wider range of strategies and the use of translanguaging. We frequently observed Ms. Dominique whose repertoire included Luxembourgish, French, German and English, and Ms. Daria whose repertoire included Portuguese in addition. The reading event of 30^th October 2020 will be used to showcase the range of meaning-based strategies, and the various languages deployed in addition to those in Table 2. The information is presented in Table 3.

The educators Ms. Dominique and Ms. Daria as well as Ygor (Luxembourgish), Etienne (French), Niklas (Icelandic, Luxembourgish, French), Gloria (Portuguese), Tim (German, Luxembourgish) and Noel (Luxembourgish) sat on mats on the ground of their roofed wooden shelter. Ms. Dominique showed them the book “The apple mouse” they had read the previous day and said “You can help me tell the story”. When she showed them a picture of the mice in the woods and asked where the little mouse lived, the following conversation took place.

Ms. Dominique drew on Ygor’s answer by repeating it and prompting him to elaborate, which he did in line 3 when he clarified where exactly he believed the little mouse lived. The educator confirmed and expanded by adding information. She then turned to two-year-old Etienne and, in a short simple utterance in French, explained where the mouse lived while still pointing to the picture. A little later, Ms. Daria translanguaged as well. She adapted the language use for Portuguese-speaking Gloria. She translated the word strawberry into Portuguese to ensure Gloria could participate in an exchange about fruits children liked. This brief conversation connected the story to children’s experiences. Having read the following sentence, Ms. Dominique asked a closed question.

Ms Daria translated Ygor’s utterances to make them understandable to Etienne. When Etienne then used an onomatopoeia to express that the mouse fell, she took up the word “bang”, reformulated his utterance and provided the appropriate word. On other occasions, the educators gave corrective feedback. An example comes from Etienne who spotted a lizard in a picture and called it “gackack”. He tried to croak, but, as observed on previous occasions, was unable to make this sound. By uttering “It is not a gackack, it is not a frog. It is a lizard”, Ms Daria, firstly, showed Etienne that she had understood that he intended to speak of a frog and, secondly, corrected by offering the right word. Apart from translations, transformations were recurrent in this conversation. Ygor elaborated by adding to his own utterance (line 9) while Niklas (line 16) changed the meaning of sentence that Ms Dominique had previously reformulated.

Table 3. Overview of translingual dialogic reading strategies in Earth.

4.2. Dialogic Reading Strategies in One or More Languages

This section presents an analysis of the 40 observations in relation to the languages and dialogic reading strategies used. A main difference between the educators in Earth and Air was their use of translanguaging to facilitate comprehension and encourage participation. The educators in Air were observed translanguaging in 7% of their readings, i.e., switching from French to Spanish or English, while those in Earth did so in 58% of the observed readings. When necessary, they would briefly turn to German, French or Portuguese before reverting to Luxembourgish to manage a situation or ensure specific children could follow and express themselves. On these occasions, they communicated the meaning of a previous sentence in a home language or translated. On a few occasions, languages themselves became the topic of conversation. In December 2020, Ms. Dominique took a book that Ms. Daria had read previously in Luxembourgish. Niklas mentioned it featured a rabbit, using the French word “lapin”. Ms. Dominque asked which language this was and named it when children were unsure. Gloria then spontaneously volunteered the Portuguese word “coelho”, Niklas tried to find the Icelandic for rabbit and Lia volunteered a Swiss German word. Ms. Dominque repeated these words and named the language. Children’s openness to languages was also mentioned by Ms. Dominique: “Children are aware of different languages and know what they are called” (interview May, 2021).

Translanguaging, a legitimate practice in Earth, engaged children, sustained their participation and helped them connect expressions and ideas across languages. Participation was explicitly encouraged, and children did not hesitate to draw on home languages when they could not find the words in Luxembourgish. “I expect that children participate and ask questions”, stated Ms. Dominique (interview, January 2021). For his purpose, she provided information on content, connected stories to children’s experiences, and “reached them through questions” when they did not engage. Children participated well and frequently told stories to each other during free play. “You hear some sentences or parts of sentences where they imitate you and this is very nice to observe”, explained Ms. Dominique in January 2021.

By contrast, educators in Air were rarely observed speaking about the content of a book or relating it to the children’s experiences. Engaging children in meaning-making was not encouraged and appeared to be based on the educators’ understanding that reading is “for fun” (interview, May 2021). They claimed that reading created enjoyable moments, stimulated the imagination, and contributed to the development of listening skills. At the same time, they disagreed that reading was about engaging children in conversations about text (fieldnotes, October 2021). This may explain why they at times discouraged uninvited participation. “Wait, wait, we will read the story, wait, wait” is a representative example where Ms. Fabienne tried to stop a few children from anticipating the content of a familiar story (observation, November 2020).

The educators’ understanding of the purpose of reading helps explain the slight differences in the CROWD questions. Figure 1 indicates the percentage of strategies observed across the centers. The first two help children make meaning of the story while the last six are based on the PEER sequence and CROWD questions. While all educators used non-verbal strategies as well as closed questions independently of the language, none deployed the more complex distancing questions. In Earth, the educators used closed questions to help children retell stories and open-ended questions to have them guess or make connections to their experiences. Only the educators in Air used completion questions. Their use of closed questions aimed to encourage children to recall content, count objects, identify colors, indicate the size of an object, as well as label and describe pictures. The repeated readings helped them teach vocabulary and revise mathematical concepts despite their statement that reading was “for fun”. Upon viewing some of the video-recordings, the manager described the approach as adult-centered and “rigid”, saying, “We have at times the wrong reflex to say it’s us who read and them who listen”. She commented that this may explain the low number of questions (interview, May 2021). At the same time, some of the educators appeared to believe that three-year-olds lacked the language skills to dialogue, despite agreeing that many of their children communicated well. This belief and the experience that some “children were unable to answer basic content questions” (fieldnotes, October 2021) explained the lack of content questions. They did not consider that the repetitive reading and the lack of cognitive stimulation through closed questions may have resulted in children’s disengagement. Towards the end of the year, they read simple books and sang more songs. They explained that the songs developed phonological awareness and familiarized children with several languages.

Given that the different types of prompts led to varying amount of talk in both settings, the educators made different use of evaluations, expansions, and repetitions. In other words, the PEER sequences differed. Evaluations in Earth took three forms independently of the language; the educators confirmed correct answers; corrected content, concepts, and language (e.g., pronunciation mistakes, grammar, lexis), and, thirdly, asked children for their opinions in open-ended questions which could result in a range of contrasting views. An example comes from the final excerpt where Niklas disagreed with Ms. Dominique. The educators understood their role as language models and expanded on children’s answers, for instance by adding details. They also repeated keywords and rephrased utterances as follow-up questions, mainly to stimulate thinking and talking. As the children in Air spoke little, the educators rarely expanded on utterances and provided little feedback, though occasionally on content. Repetitions were frequent

Figure 1. Comparison of dialogic reading strategies in the centers.

and testified to the educators’ intention to help children “memorize words and develop their vocabulary” (interview, May 2021), a statement which contradicts the “fun” approach.

5. Discussion

The present study examined the educators’ dialogic reading practices and language use. The results show, firstly, that the dialogic reading practices differed in frequency, length and reading strategies. This finding is reminiscent of studies elsewhere (Adam & Barratt-Pugh, 2023; Grøver et al., 2022; Torr, 2019). Practices vary even in intervention studies where educators have been trained to read dialogically (Pillinger & Vardy, 2022). The educators in Luxembourg did not take part in an intervention. In fact, only the manager and one educator in each center had familiarized themselves with dialogic reading and language promoting strategies in the training prior to the data collection. This explains why the educators deployed the PEER sequence and the CROWD questions (Whitehurst et al., 1988) though in different ways. Those in Air used predominantly closed questions and encouraged labelling with the three-year-old children though these prompts are typically used with two-year-olds (Mol et al., 2008). Complex questions that stimulate thinking and language-modelling strategies such as expansions, were more frequent in Earth where the educators worked with smaller groups and read for longer stretches. It has been found that extended conversations are easier in smaller groups (Adam & Barratt-Pugh, 2023; Mihai, 2021). The centers also differed in relation to evaluations, expansions and repetitions with more educators in Earth than in Air giving feedback on language and story comprehension. The former also established more connections between books and children’s experiences which is likely to promote a deeper understanding and foster literacy skills (Whitehurst et al., 1988, 1994). The educators’ scaffolding helped children make such connections themselves, thus, create meaning, when they looked at books on their own (Kemp, 2024).

A second finding relates to translanguaging during dialogic reading which was frequent at Earth and almost absent at Air although most of the children were emergent multilinguals and most staff spoke multiple languages. Like studies elsewhere, the educators modelled, repeated, clarified and recast in one or two languages and used translations to ensure comprehension (Pontier & Gort, 2016; Seltzer et al., 2020; Sembiante et al., 2023). In contrast to these studies in bilingual preschool settings, the educators in Luxembourg worked with younger children in non-formal education and used more languages. Translanguaging promoted children’s participation and meaning making (Carrim & Nkomo, 2023; Kirsch, 2021; Pontier & Gort, 2016) and fostered their language awareness (Velasco & Fialais, 2018).

The interview data help explain why some practitioners read dialogically and translanguaged: the reading practices were related to the educators’ beliefs and understanding of the purposes of reading. Those educators who believed that children were able to talk and narrate stories, and that reading could promote conversations, read dialogically. Those who believed that emergent multilinguals needed support, scaffolded conversations and translanguaged. Previous studies confirm the relationship between reading practices and beliefs about language support. This was the case in the Netherlands where teachers used less vocabulary and less syntactically complex language when reading with bilingual Turkish children than with monolingual Dutch children (Aarts et al., 2016). Similarly, a teacher who believed that Chinese learners would simply pick up English did not offer support during dialogic reading whereas a teacher who believed that the learners needed extra support, provided it (Hu et al., 2021). One wonders, however, why some educators in the present study appeared to be influenced by their beliefs in children’s language abilities rather than the language performances they could observe.

Finally, the dialogic reading and language practices reflect the centers’ overall pedagogical practices. The high-quality talk characterized by the high number of questions, the language modelling, the amount of information, the frequent connections as well as the language accommodations, reflect the social constructivist learning theories that underpin the practices in Earth (Kirsch & Bergeron, 2023; Kirsch & Hornberger, 2024). Grøver et al. (2022) found that the quality of teacher talk, in particular the length and syntactic complexity of their utterances and the number of explanations, impacted children’s language development. While this study did not examine children’s language development, the children in Earth spoke more and used a wider vocabulary during dialogic reading than those in Air. By contrast, the more classic and monolingual reading sessions in Air reflected the structured and behaviorist approach of this center (Kirsch & Hornberger, 2024).

6. Conclusion

Delivering high-quality instruction to develop language and emergent literacy skills is a challenge for many ECEC educators, particularly for those working with emergent multilinguals (Bergeron-Morin et al., 2023). The findings of the present study address an important gap in literacy practices in ECEC in both monolingual and multilingual contexts. Whether reading practices were more classic, as in Air, or dialogic, as in Earth, depended less on the language(s) of the educators or children than on the educators’ understanding of reading and their pedagogy. When educators translanguaged strategically, they leveraged children’s entire semiotic repertoire for communication, participation and meaning making (García & Kleifgen, 2020).

While the present study provides rich contextual data that deepen our understanding of dialogic reading practices, the study is not without limitations. Firstly, a small-scale study cannot be representative of all ECEC institutions in Luxembourg and, secondly, like all naturalistic studies, the participating children and educators varied, which made it more difficult to compare the data and draw conclusions. Furthermore, there is always a risk of observer bias. We mitigated these risks by observing routine situations over one year, critically discussing data in the team, sharing our interpretations with the educators, and using investigator, method and data triangulation.

Based on the findings, I recommend that practitioners regularly engage in dialogic reading with young children and create a learning environment where children can draw on their entire semiotic repertoire. This is easier if the professionals are multilingual, but it is unlikely (and not necessary) that they speak all of the children’s home languages. What is required is respect for children’s linguistic and cultural backgrounds, an understanding that the home language is the foundation for learning, an awareness of the hierarchy of languages in educational settings and in intention to promote participation and social justice. These recommendations call for professional development courses on the processes, purposes, and effects of dialogic reading. Training that includes reflective components, can lead to a change in beliefs, knowledge, and attitudes (Bergeron-Morin et al., 2023; Chung & Fisher, 2022; Pillinger & Vardy, 2022). Such courses should include a focus on pedagogical translanguaging to help professionals understand the need for openness and flexibility and design spaces that enable all learners to draw on their semiotic repertoire for meaning making (García & Kleifgen, 2020). By promoting translanguaging literacy strategies, policymakers and practitioners contribute to more equitable learning spaces.

Future studies could assess the language and early literacy skills that young multilingual learners develop during multilingual dialogic reading and add to the few studies that show positive effects of dialogic reading on two or more languages (Hui et al., 2020; Pico & Woods, 2023).

Acknowledgements

I would like to thank the participating educators, the student assistants who transcribed the data, as well as Dr Valérie Kemp, Dr Džoen Bebić-Crestany, Laura Colucci and Dr Gabrijela Aleksić who collected the data or contributed to the analysis. We are also indebted to three institutions who financed and supported the COMPARE study: the Luxembourg National Research Fund [C19/SC/13552634], the Ministry of Education, Childhood and Youth and the National Youth Service [COMPARE].

Conflicts of Interest

The author declares no conflicts of interest regarding the publication of this paper.

References