VYTAUTAS MAGNUS UNIVERSITY

Eglė KRIVICKAITĖ

PHONOTACTICS DEVELOPMENT OF CHILDREN’S

Summary of Doctoral Dissertation Humanities, Philology (04 H)

Kaunas, 2016 The dissertation has been prepared at Vytautas Magnus University in 2011– 2015. The right for the joint doctoral studies was accorded to Vytautas Magnus University, Klaipėda University, Šiauliai University and the Institute of the on July 8, 2011 according to the decree of the Minister of Education and Science of Lithuania No. V-1019. Scientific supervisor: prof. dr. Ineta Dabašinskienė (Vytautas Magnus University, Humanities, Philology – 04 H)

The dissertation is defended in Philology Science Board of Vytautas Magnus University together with Klaipėda University, Šiauliai University and Lithuanian Language Institute: Chairperson: Prof. Dr. Asta Kazlauskienė (Vytautoas Magnus University, Humanities, Philology – 04H) Members: Prof. Habil. Dr. Bonifacas Stundžia ( University, Humanities, Philology – 04H) Prof. Dr. Dalia Pakalniškienė (Klaipėda University, Humanities, Philology – 04H) Prof. Dr. Genovaitė Kačiuškienė (Šiauliai University, Humanities, Philology – 04H) Assoc. Prof. Dr. Laura Kamandulytė-Merfeldienė (Vytautas Magnus University, Humanities, Philology – 04H)

The dissertation shall be defended in an open session of Philology Science Board on May 20, 2016 at 12 o‘clock at Vytautas Magnus University, room 211. Address: K. Donelaičio str. 52, 44244 , Lithaunia

The summary of the doctoral dissertation was sent out on April 19, 2016. The dissertation is available in the libraries of Vytautas Magnus University, Klaipėda University, Šiauliai University and the Institute of the Lithuanian Language.

2 VYTAUTO DIDŽIOJO UNIVERSITETAS

Eglė KRIVICKAITĖ

VAIKŲ KALBOS FONOTAKTIKOS RAIDA

Daktaro disertacijos santrauka Humanitariniai mokslai, filologija (04 H)

Kaunas, 2016

3 Mokslo daktaro disertacija rengta 2011–2015 metais Vytauto Didžiojo universitete pagal suteiktą Vytauto Didžiojo universitetui su Klaipėdos universitetu, Šiaulių universitetu ir Lietuvių kalbos institutu (2011 m. birželio 8 d. Švietimo ir mokslo ministro įsakymas Nr. V-1019) doktorantūros teisę. Mokslinė vadovė: prof. dr. Ineta Dabašinskienė (Vytauto Didžiojo universitetas, humanitariniai mokslai, filologija – 04 H)

Mokslo disertacija ginama Vytauto Didžiojo universiteto su Klaipėdos universitetu, Šiaulių universitetu ir Lietuvių kalbos institutu Filologijos mokslo krypties taryboje: Pirmininkė: prof. dr. Asta Kazlauskienė (Vytauto Didžiojo universitetas, humanitariniai mokslai, filologija – 04H) Nariai: prof. habil. dr. Bonifacas Stundžia (Vilniaus universitetas, humanitariniai mokslai, filologija – 04H) prof. dr. Dalia Pakalniškienė (Klaipėdos universitetas, humanitariniai mokslai, filologija – 04H) prof. dr. Genovaitė Kačiuškienė (Šiaulių universitetas, humanitariniai mokslai, filologija – 04H) doc. dr. Laura Kamandulytė-Merfeldienė (Vytauto Didžiojo universitetas, humanitariniai mokslai, filologija – 04H)

Daktaro disertacija bus ginama viešame filologijos krypties tarybos posėdyje 2016 m. gegužės 20 d., 12 val. Vytauto Didžiojo universiteto 211 auditorijoje. Adresas: K. Donelaičio g. 52, 44244 Kaunas, Lietuva

Disertacijos santrauka išsiųsta 2016 m. balandžio 19 d. Disertaciją galima peržiūrėti Vytauto Didžiojo universiteto, Klaipėdos universiteto, Šiaulių universiteto ir Lietuvių kalbos instituto bibliotekose.

4 1. INTRODUCTION

Relevance of the Topic and Novelty of the Thesis The present thesis analyses the phonotactics acquisition of Lithuanian children. The acquisition of Lithuanian sounds, their combinations, and words with different structure has been described only fragmentary and focusing on different periods (Garšvienė, Ivoškuvienė 1993; Garšvienė, Juškienė 2008; Auksoriūtė 2010; Kamandulytė 2006a, b; Kamandulytė-Merfeldienė 2015). It has been observed that an increasing number of children need speech therapists’ or special pedagogues’ help every year because of impairments in language development: 1 children have difficulty in remembering new words, pronouncing forms with more complicated structure, concentrating, etc.2 Thus, consistent research, using innovative methods, is necessary, the results of which would reveal the acquisition of Lithuanian and phonotactics in greater detail. Applying the experimental method, also referred to as a special task, the present thesis explores the language data of five children age groups: 4;00–4;113, 5;00– 5;11, 6;00–6;11, 7;00–7;11, 8;00–8;11; i.e. preschool age (4;00–5;11), preprimary school age (6;00–6;11), and junior primary school age (7;00–8;11). Earlier experimental research on Lithuanian children’s language was mainly carried out with one or two age groups. It is important to mention that the present dissertation research has a large sample as each age group contains 95-97 participants. In total, the data of 480 typically developing children aged from four to eight were collected at the kindergartens and schools in Kaunas city, Kaunas district, and Kaunas county. It is suggested that such large sample size could help to determine the features of Lithuanian phonotactics acquisition more objectively. In order to collect the data, the Lithuanian non-word repetition test was used for the first time. The Objective of the Study The aim of the dissertation is to analyse the features of phonotactics of preschool age (4;00-5;11), pre-primary school age (6;00-6;11), and junior primary

1 According to the data of Statistics Lithuania, the number of children in special education groups in preschool educational institutions has been increasing: 694 children in 2012, 712 children in 2013, and 836 children in 2014. 2 This has been noted during individual consultations with speech therapists at kindergartens and schools. 3 Age is marked the following: 4;00–4;11 from four years 0 months to four years 11 months.

5 school age (7;00-8;11) children, applying the non-word repetition test. In order to reach the aim, the following objectives have been formulated: 1. to present the methodology of the Lithuanian non-word repetition test; 2. to collect representative data of children’s language, using non-word repetition test; 3. to perform a quantitative and qualitative analysis of the data: 3.1. on the basis of the statistical analysis, using SPSS 17.0 programme, to determine the following: 3.1.1.1. the influence of age and sex on phonotactics acquisition; 3.1.1.2. the acquisition tendencies of non-words with different structure among children aged from four to eight; 3.2. to determine the strategies of pronunciation simplification of non-words employed by different age groups; 3.3. to describe and compare the pronunciation features of sounds and their clusters of each age group; 4. to determine the most typical features of phonotactics acquisition of Lithuanian children’s language based on the results of the qualitative and quantitative results. Research Methods The present research uses the combination of several methods: a cross- sectional method was used in order to collect the sample; an experimental method (non- word repetition test) was used to collect the research data; in order to analyse the data, error analysis, comparative, and statistical methods were employed. The Lithuanian non-word repetition test was designed for this research; therefore, the application of this test by foreign researchers and the structure of the Lithuanian non-word repetition test are described in greater detail. The non-words repeated inaccurately were analysed using a qualitative error analysis method. When comparing different age groups, as well as the results of boys and girls, a comparative method was applied. Statistical methods were used in order to discuss the influence of sex and age on phonotactics acquisition. The use of the combination of these methods allowed identifying the main acquisition features of Lithuanian phonotactics in different age groups.

6 The Practical Value of the Research The results of the present dissertation might be useful and topical in preparing practical speech development tasks for typically developing children. After determining the sounds, their clusters, and the structure of words which are difficult to pronounce for children of different age groups, kindergarten teachers, secondary school teachers, speech therapists, and parents could devote more attention for their development. The results obtained during the present research have revealed the acquisition features of sounds and their clusters, as well as of words with different structure in the speech of typically developing children aged from four to eight. Therefore, these results might be useful for speech therapists and special pedagogues who work with children having language impairment: they could help in diagnosing phonological language impairments and creating the methodology of their adjustment. The methods and the results of the present dissertation might also become the basis for future scholarly research, analyzing the features of phonotactics acquisition of children with language impairment and/or bilingual/multilingual children. Structure of the Study The dissertation is comprised of the Introduction (Chapter 1) and four main parts. Chapter 2 ‘Phonology Acquisition’ focuses on the acquisition features of phonology and phonotactics. Chapter 3 ‘The Research on Children’s Language Using the Method of Non-word Repetition’ discusses the application of the non-word repetition test in foreign research and presents the methodology of the Lithuanian non- word repetition test and research data. Chapter 4 ‘Phonotactics: A Quantitative Data Analysis’ describes the repetition results of the non-words with different structure produced by the boys and girls aged 4;00-8;11. Chapter 5 ‘Qualitative Analysis. The Tendencies of Non-word Pronunciation’ presents the qualitative analysis of non-words which were repeated inaccurately: the pronunciation features of sounds and their clusters typical of a certain age group were identified. At the end of the dissertation, Conclusions (Chapter 6), the list of references, and appendices are provided. Statements to be Defended 1. Sex does not have a statistically significant influence on the acquisition of phonotactics.

7 2. The acquisition of the principles of phonotactics is related to age: older children can pronounce the words with different structure, their sounds, and clusters more accurately. 3. The pronunciation accuracy of sounds and their clusters depends on the word structure: it is more difficult to pronounce a sound in longer words and in the words with a cluster. 4. The ability to repeat the words with different structure is related to the Lithuanian phonotactics: when repeating the word that they hear for the first time, children pronounce the sound clusters typical of Lithuanian more accurately, and, on the contrary, make more errors in pronouncing the clusters untypical of Lithuanian. 5. The universal strategies applied by children of different age were used for simplifying the pronunciation and reflect the acquisition path of sounds and their clusters.

2. PHONOLOGY ACQUISITION

In the present research, the analysis of word forms pronounced inaccurately was based on the ideas of the Naturalness Theory (Dressler 1993, 2005; Dressler et al. 2001; Donegan, Stampe 2009; Dziubalska-Kołaczyk 2007, 2014; Marecka, Dziubalska- Kołaczyk 2014). Naturalness in a language refers to a form which is cognitively simple, easy to pronounce, and to perceive; therefore, it is used frequently. Unmarked forms are viewed as natural elements in a language (i.e. less marked=more natural); they are frequent in a language and they are acquired early. Naturally, in language acquisition, children make more errors in using marked forms, i.e. the elements which are not very natural are frequently substituted by more natural and more usual elements and/or the elements that are easier to pronounce (Dressler 1993; Dressler et al. 2001; Marecka, Dziubalska-Kołaczyk 2014; Dziubalska-Kołaczyk 2015). Universal principles of phonology acquisition. Comparative language acquisition research has revealed that until six months the children of various nationalities pronounce the same sounds: (p, t, k, b, d, g), nasals (m, n), and . Constantly hearing the language of adults, the children naturally start acquiring the sounds in their environment, and more often pronounce the more typical of their native language.

8 One-year-old children can understand the sounds and their clusters in their environment better than they can pronounce themselves; therefore, during the second, the third year, and later, the vocal tract still develops until children can pronounce the sounds and clusters of their mother tongue (Brooks, Kempe 2014). Consonant clusters appear in children’s language when they are approximately two years old. The ability to pronounce a consonant cluster is related to the maturity and development of the vocal tract. Children can pronounce various ; however, it is substantially more complicated to pronounce a consonant cluster. Children learn to pronounce clusters gradually: first, CCV structure syllables are substituted by CV structure syllables; later, one of the constituent parts is substituted by the sound which is easier to pronounce; finally, the cluster is pronounced accurately (McLeod 2001: 100– 102). Comparative research on shows that consonant cluster acquisition reflects the rules of phonotactics of a language: children acquire those aspects more quickly which are typical of their environment. In such a case, the clusters which often appear in words are practiced more often and, consequently, pronounced accurately earlier (Marecka, Dziubalska-Kołaczyk 2014: 45). The words which have a more complicated structure are usually simplified. Scholars have identified the following universal strategies of word simplification typical of all languages: 1) the sounds which are more difficult to pronounce are substituted by the sounds which are easier to pronounce; 2) the sounds are interchanged (metathesis); 3) a sound or a syllable is omitted; 4) a sound is moved from one syllable to another (McLeod 2001; Velleman, Vihman 2007; Harley 2010; Zanobini et al. 2012). These pronunciation strategies of sound simplification are also referred to as natural language development processes (see Brooks, Kempe 2014). Lithuanian children aged 2;0-2;6 already pronounce all Lithuanian consonants (Garšvienė, Ivoškuvienė 1993). Until two years of age, they can pronounce certain consonant clusters. The research on the Lithuanian language acquisition has demonstrated that children can pronounce the consonant clusters across boundaries rather than within more accurately (Kamandulytė 2006a, b; Kamandulytė-Merfeldienė 2015) because, according to Dressler (2006), children tend to acquire the language features which denote a function. The clusters between different

9 morphemes mark morphological boundaries. At approximately five years of age, children can already pronounce the clusters of two-three consonants; however, it is more complicated to pronounce the words with a more complicated syllable structure: the sounds can be mixed, changed, or omitted, as well as regressive and progressive assimilation might take place. The development of phonology lasts for a long time; therefore, even the children of pre-primary or junior primary school age might have difficulty in pronouncing more complicated sounds, consonant clusters, or multi- syllable words.

3. THE RESEARCH ON CHILDREN’S LANGUAGE USING THE METHOD OF NON-WORD REPETITION TEST

The present dissertation employs the Non-word Repetition Test, which is considered to be an experimental method when the respondent is asked to repeat non- words4. In order to be able to repeat the word which is heard for the first time and does not have any meaning, linguistic-cognitive abilities are necessary: phonological processing, short-term memory, articulation abilities, etc. (Rispens, Parigger 2010). Even though non-words are created using the rules of a particular language, they do not have any similarity to real words in a language; therefore, the already obtained lexical knowledge is not used when performing the test (Radeborg et al. 2006: 187). Rather, the capacity of short-term memory is used, i.e. the ability to code and maintain the phonological sequence of sounds (Gathercole 1995: 83). The non-word repetition test is important in monitoring the child’s language development. At a young age, the child’s ability to repeat a new polysyllabic word that s/he hears for the first time shows his/her ability to learn new words and broaden his/her lexicon (Gathercole 2006a). Each word that the child heard for the first time some time ago sounded unusual and strange similarly to the words in this test (Chiat, Roy 2007). Longitudinal research results have demonstrated that the children who performed better in the test had a broader lexicon after a year compared to the children who performed worse (Gathercole 1995; Baddeley et al. 1998).

4 Non-words are a phonological sequence of sounds which corresponds to phonotactic rules of a specific language and do not have any meaning and function in a sentence.

10 The conclusions of researchers from various countries (English, Italian, Spanish, Dutch, and other languages) are the same: children with SLI repeat non-words significantly worse than typically developing children (Dollaghan, Campbell 1998; Gahtercole 2006a, 2006b; Girbau, Schwartz 2008; Dispaldro et al. 2013). Comparative research using non-word test designed for different languages allows distinguishing three main factors which influence the accuracy of non-words repetition: 1) Complexity of non-words (CV vs. CCV). Non-words with consonant clusters are repeated less accurately than non-words without consonant clusters (Kavitskaya et al. 2011). Word-medial and word-final clusters are repeated less accurately then word-initial consonant clusters (Marshall, van der Lely, 2009: 50); 2) Non-words length (number of syllables in the word). Non-words repetition accuracy declines with the increasing number of syllables, i.e. one- or two- syllable words are repeated more accurately than three- or four-syllable non-words (Chiat, Roy 2007). It is related to the ability to keep phonological information in short-term memory. 3) Age of participants: the older the age group, the more accurately both shorter (one-two syllable) and longer (three-four syllable) non-words are repeated (Santos et al. 2006; Park, Scarz 2012). Older children have a larger lexicon and develop it more; they also find out different sound clusters and can pronounce various consonant clusters more accurately (Munson et al., 2005: 1034). Structure of the Lithuanian Non-word Repetition Test. 5 The Lithuanian Non-word Repetition Test was designed with regard to the structural characteristics of Lithuanian words (word length, and consonant frequency, and syllable structure) (more about Lithuanian word structure see Kazlauskienė 2007; Kazlauskienė, Raškinis 2008a, b; Kazlauskienė 2010; Girdenis, Karosienė 2010). The test consists of 24 non- words of different structure (8 non-words have two syllables (4-6 phonemes); 8 non- words have three syllables (6-7) phonemes; 8 non-words have four syllables (7-8) phonemes). There are two non-words without consonant clusters and 6 non-words with

5 The Lithuanian non-word repetition test (Dabašinskienė, Krivickaitė 2013) was developed participating in the project COST IS0804 Language Impairment in a Multilingual Society: Linguistics Patterns and the Road to Assessment (2009–2013).

11 consonant clusters in each group (see Table 1). Each non-word is associated with a Lithuanian equivalent, taking into account word length and syllable structure.

Table 1. Lithuanian Non-words 2 syllables non-words 3 syllables non-words 3 syllables non-words Ke.mu Ge.lo.ša Su.le.ri.tė Do.ja Ši.ru.ta Ža.de.vi.na Ski.mo Šku.li.nė Sna.li.di.na Šve.la Ple.mu.ta Spi.ra.tu.ša Ga.prė Ma.spu.lė Ni.spa.ri.ma Gi.tva La.smu.vi Ma.gvu.no.lė Smin.to Spa.de.ki Sta.li.go.sa Kle.sta Pa.sva.pi Go.sa.klu.ni Note: syllables are separated by dots and a stressed syllable is underlined

The non-word repetition test in Lithuanian is presented as a game using Ms PowerPoint (see Figure 1). The child is introduced to the main game character, the monkey, who wants to get some bananas and has to complete the tasks. The child is suggested to help the monkey and do the tasks. In each step, the child hears a recorded non-word, which s/he has to repeat. All conversation is recorded; this way, imprecise answers are not marked in the child’s presence. Children like this task because it is playful, fun, and short (it lasts only for three-four minutes).

Figure 1. Visual of the Non-word repetition test (design by Kunnari, Tolonen, and Chiat, 2011)

12 Participants. The dissertation research consists of 480 typically developing children of five age groups: 4;00–4;11 age group; 96 children: 37 boys and 59 girls; 5;00–5;11 age group; 95 children: 48 boys and 47 girls; 6;11–6;11 age group; 95 children: 46 boys and 49 girls; 7;00–7;11 age group; 97 children: 49 boyd and 48 girls; 8;00–8;11 age group; 97children: 47 boys ir 50 girls.

4. PHONOTACTICS: A QUANTITATIVE DATA ANALYSIS

This chapter discusses the data from a social (sex and age) and linguistic (length of a word and word structure) perspectives. For the qualitative analysis, the data were coded manually and analysed using SPSS 17.0 (Statistical Package for the Social Sciences) programme. In order to check statistically significant differences, the Analysis of Variance, ANOVA, and Post Hoc criterion were used. In order to check the significance and strength of statistical relationship (correlation), Pearson’s (r) correlation was used. For statistical calculations, a standard 0.05 level of statistical significance was chosen.

4.1. The Factor of Sex

The general 6 test results indicate that both boys and girls repeated non-words with the accuracy of 75%. The results among different age groups are slightly different: in the groups of four-year-old and eight-year-old children, the results of boys are slightly better, while in the groups of five-year-old, six-years-old, and seven-year-old children, the results of girls are better. The words without a consonant cluster were repeated with the accuracy of 81% by the boys aged 4;00-4;11 and with the accuracy of 74% by the girls of the same age. From the age 5;00-5;11 to 7;00-7;11, girls repeated the non-words slightly more accurately. The results are the same for both boys and girls in the age group 8;00-8;11: the words were repeated with the accuracy of 91%.

6 In this section, a general analysis means that all age groups are analysed together, distinguishing only the results of boys and girls.

13 The repetition analysis of words with a consonant cluster has revealed that girls repeated the words slightly more accurately in the age groups 5;00-7;11, while boys were more accurate in the age group 4;00-4;11. In the age group 8;00-8;11, both sexes repeated the words with the same accuracy. Initial and medial position of a cluster. Four-year-old boys repeated the clusters in the initial position more accurately than girls (respectively, with the accuracy of 86% and 80%). In the group of five-year-old children, girls repeated the clusters in the initial and medial position more accurately (girls repeated the clusters in the initial position with the accuracy of 91%, while boys with the accuracy of 87%; girls repeated the clusters in the medial position with the accuracy of 79% and boys of 76%). Boys and girls in the age group from six to eight repeated the clusters in the initial and in the medial position with a similar accuracy. The differences in the results of non-word repetition are not statistically significant with respect to sex; therefore, in the analysis of age, word length, and word structure, boys and girls are discussed in the same age group.

4.2. The Factor of Age

This section analyses the general7 results of the non-word repetition test. The features of word structure (word length and structure) of different age groups are not discussed in this section because they are described in another section (see 4.3. The Word Structure). The results of the non-word repetition test are statistically significant for children aged 4-8 (p=0.000). Older children repeated non-words more accurately (see Table 2). One way ANOVA Post Hoc analysis has indicated that the children aged 4;00- 4;11 repeated non-words statistically significantly worse than other groups (p=0.000), while the results of the age group 5;00-5;11 are significantly different from the age groups 4;00-4;11 and 8;00-8;11. Statistically significant differences have not been noted in the results of other age groups.

7 In this section, a general analysis means that the results of boys and girls are analysed together, distinguishing only different age groups.

14 Table 2. The results of the non-word repetition test (age groups 4;00-8;11) 4;00–4;11 5;00–5;11 6;00–6;11 7;00–7;11 8;00–8;11 Results (%) 65 % 74 % 78 % 77 % 80 %

As indicated in Table 2, the results of the age group 7;00-8;11 are very similar to the results of five-year-old and six-year-old children. On the basis of these results, one can maintain that phonological abilities develop the most intensively until six years of age. Boxplot (see Figure 2) shows the result distribution of non-word repetition and allows a consistent comparison of the results of each age group. The ability of four- year-old children to repeat non-words accurately is very different as the interval of responses is very wide: the lowest result in the group is 25% (six words were repeated accurately out of 24), while the highest result is 92% (22 words out of 24). The interval of non-word repetition for five-year-old children is shorter: the lowest result is 50%, and the highest result is 100% of the correct answers. In addition, different abilities of five-year-old children may be observed: some children can repeat all words of the test accurately, while others – only a part of the words.

Figure 2. Mean of non-word repetition for each age group

15 The results of six-year-old- and seven-year-old children are similar and close to the results of five-year-old children. The highest test result in the group of six-year-old children is slightly lower than in the group of five-year-old children: in the group, there was no child who repeated all words accurately. The lowest result in the group of seven- year-old children is slightly higher than in the group of five-year-old- and six-year-old children. The results of eight-year-old children are higher than younger groups and they are distributed in the shortest interval: the lowest result is 58% of the correct answers, and the highest result is 100%. Thus, the boxplot has revealed that the interval of test result distribution is shorter with each age group; with each group, the lowest index value is higher. This means that children can repeat non-words more accurately; on the other hand, it can be observed that the general test results of school-age children are very similar to the results of younger children. Therefore, several hypotheses were raised. First, it was hypothesized that the test results were influenced by poor results of one or several schools; however, this hypothesis was not confirmed. Second, it was hypothesized that poor test results could have been influenced by particularly high results of six-year-old children of a kindergarten; nevertheless, this hypothesis was also rejected. Third, it was thought that poorer test results of school-age children were influenced by the so-called ‘schooling effect.’ It is suggested that psychological factors of adaptation could have some influence on the test results (i.e. new environment, new methods, etc.; in addition, the non-word repetition test was performed during the first half of the new school year, in November). This hypothesis was not tested by any special tools. It is important to mention that some differences can be observed inside each group, i.e. it has been investigated how the results change in one year (see Table 3). For instance, in the group of 4;00-4;05, the mean of non-word repetition is 61.5%, while in the group of 4;06-4;11, it is 68.2%.

16 Table 3. The results of the non-word repetition test inside each group 4;00–4;05 61,5 proc. Four years old 4;06–4;11 68,2 proc. 5;00–5;05 73,2 proc. Five yeards old 5;06–5;11 74,6 proc. 6;00–6;05 76,6 proc. Six years old 6;06–6;11 79,5 proc. 7;00–7;05 75,6 proc. Seven years old 7;06–7;11 79,0 proc. 8;00–8;05 77,3 proc. Eight years old 8;06–8;11 82,5 proc.

As indicated in Table 3, the results of the age group 7;00–7;05 are slightly worse than those of younger children: in the group of 7;00-7;05, the mean of non-word repetition is 75.6%, while in the group of 6;00-6;05, it is 76,6%, and in the group of 6;06-6;11, it is 79,5%; the results of the age group 8;00-8;05 are slightly worse than results of 7;06-7;06 group (see Table 3). These results are not statistically significant.

4.3. The Word Structure

4.3.1. The Word Length. The analysis of word length is composed of two parts: first, it is analysed how children can retain the number of syllables in a word; second, it is investigated how children can repeat two-to-four syllable words. For the children aged 4;00-8;11, it is not difficult to retain the number of syllables in a word:8 the length of two- and three- syllable words was retained with the accuracy of 99%, and the length of four-syllable words was retained with the accuracy of 97%. The analysis of different age groups has revealed that it was difficult for the youngest research participants to pronounce three or four syllables in a word, for five- year-old children – only four syllables, and for six-year-old and eight-year-old children, it was not difficult to pronounce all four syllables in a word.

8 In the analysis of retaining syllables in a word, the answers when the sounds were changed, omitted, added, or the words were mispronounced if the number of syllables does not change in a word, were not considered to be wrong. The answers were considered to be wrong if the word becomes one syllable shorter because of an omitted sound; for instance, škul.nė (= šku.li.nė); in addition, the answers were considered to be wrong if all syllable was omitted (e.g. spa.ri.ma (= ni.spa.ri.ma)), or an additional syllable was added (e.g. do.lu.ja (= do.ja)).

17 The analysis results of accurate repetition9 of two-to-four syllable words have demonstrated that longer words are repeated statistically significantly worse than the shorter ones (p=0.000): two-syllable words are repeated with the accuracy of 89%, three-syllable words with the accuracy of 72%, and four-syllable words with the accuracy of 64%. A statistically significant correlation (p=0.000) has been determined between age groups and the repetition results of words with different length. The accuracy in repetition of two- and three-syllable words increases until six because seven- and eight- year-old children repeated the two- and three-syllable words slightly worse than six- year-old children. The accuracy in repetition of four-syllable words increases to seven years of age, and the results of eight-year-old children are very similar to the ones of seven-year-old children. It has also been determined that school-age children (7;00- 8;11) repeated four-syllable words more accurately than three-syllable words; they also repeated three-syllable words slightly worse than five-year-old and six-year-old children (see Figure 3).

Figure 3. The results of non-word repetition in the age groups of 4;00-8;11

9 In this analysis, a correct answer was considered to be only the word repeated absolutely correctly.

18 ANOVA Post Hoc Tamhane’s T2 test analysis has demonstrated that the results of two-syllable word repetition test are significantly different only between the age groups of 4;00-4;11 and 5;00-8;11, while the results of other age groups (5;00-8;11) are not statistically significant.

4.3.2. Syllable structure

The repetition results of words without a consonant cluster can be divided into three groups: the age group of 4;00-4;11, which repeated the words with the accuracy of 77%, the age group of 5;00-6;11, which repeated the words with the accuracy of 86%, and the age group of 7;00-8;11, which repeated the words with the accuracy of 90-91%. There are not many words without a consonant cluster in the test (six words: two two-syllable, two three-syllable, and two four-syllable words); therefore, the pronunciation accuracy of these words in each age group can be analysed in great detail. In the age groups 4;00-6;11, four or five words were accurately repeated the most frequently, and the children aged 7;00-8;11 accurately repeated five or six words the most frequently. Four-year-old children repeated the words with a consonant cluster significantly worse than other research participants (p=0.000), while five-year-old children repeated them significantly worse than eight-year-old children (p=0.006). In other age groups, (5;00–7;11 and 6;00–8;11), the results are different (see Figure 4); however, they are not statistically significant. The analysis on word length has shown a statistically significant (p=0.000) medium strength negative correlation (r= -0.407), which suggests that the word length predetermines its repetition, i.e. the longer the word is, the worse it is repeated. From the data provided in Figure 4, one can note that two-syllable words with a consonant cluster were repeated the most accurately (79-92% of the correct answers in the interval). Three- and four-syllable words were repeated substantially worse: three- syllable words in the interval of 55-72%, and four-syllable words in the interval of 50- 71%. It has been found out that the repetition results of two- and three-syllable words differed statistically significantly (p=0.000). There is no significant difference between the repetition results of three- and four-syllable words (p=0.365).

19

Figure 4. Repetition results of non-words of different length with a consonant cluster in the age groups 4;00-8;11

The results of school-age children are distinct because seven-eight-year-old children repeated four-syllable words with a consonant cluster more accurately than the shorter ones, i.e. three-syllable words (see Figure 4). Pronunciation of consonant clusters. The research participants repeated consonant clusters with the accuracy of 83%. The pronunciation results of consonant clusters in the age groups 4;00-8;11 are distributed in the interval 74-87% (see Table 4). The children aged 4;00-4;11 repeated the clusters statistically significantly (p=0.000) worse than other research groups, and the results of the children aged 5;00-8;11 are similar. It can be generalized that school-age children repeated the clusters similarly or slightly worse than younger children.

Table 4. The pronunciation accuracy of consonant clusters in the age groups 4;00-8;11

Age group 4;00–4;11 5;00–5;11 6;00–6;11 7;00–7;11 8;00–8;11

Results 74 % 83% 87% 85% 87%

There is a statistically significant (p=0.000) low negative correlation (r= - 0.271) between the word length and cluster repetition. This means that the longer the

20 word is, the worse the cluster is repeated (see Figure 5): in two-syllable words it reaches the accuracy of 93%, in three-syllable words 81%, and in four-syllable words 76%. A statistically significant difference among the repetition results of clusters in two-, three-, and four-syllable words has also been indicated (p=0.000).

Figure 5. Accurate repetition of clusters in the words of different length in the age groups 4;00-8;11

The children aged 4-6 repeated the clusters in longer words worse, depending on their length; for instance, the children aged 4;00-4;11 repeated the clusters in two- syllable words with the accuracy of 87%, in three-syllable words with the accuracy of 72%, and in four-syllable words with the accuracy of 64%. Seven- and eight-year-old children repeated the clusters in two-syllable words accurately, and pronounced them in three- and four-syllable words similarly (see Figure 5). Position of the cluster: initial vs medial. The clusters in the initial position were repeated significantly more accurately than the clusters in the medial position (p=0.000): 91% and 76%, respectively. The results of different age groups have revealed that younger children have difficulty in repeating the clusters in the medial position (the clusters in the initial position were repeated by four-year-old children with the accuracy of 82% and five- year-old children – 89%, while the clusters in the medial position were repeated by

21 four-year-old children with the accuracy of 68% and by five-year-old children – 78%). The children in the age groups from six to eight repeated the clusters in the initial position with the accuracy of 92-95%, and in the medial position with the accuracy of 78-79%.

Table 5. The repetition accuracy of clusters in the initial and medial position of words with different length

2 syllables 3 syllables 4 syllables Initial position 94 % 92 % 86 % Medial position 91 % 71 % 66 %

There is a statistically significant (p=0.000) negative very low correlation (r=- 0.175) between the word length and the repetition results of the cluster in the initial position, as well as a statistically significant (p=0.000) negative weak correlation (r=- 0.384) between the word length and the repetition results of the cluster in the medial position. This suggests that the longer the word is, the more difficult it is to repeat the clusters accurately, especially in the medial position (see Table 5).

Figure 6. Repetition results of clusters in the initial and medial position of two- to-four-syllable words in different age groups

22 A more detailed repetition analysis of clusters in the initial and medial position of the children aged four-eight has demonstrated that the children aged 4;00-6;11repeat the initial clusters similarly in two- and three-syllable words, while the children aged 7;00-8;11 repeated the clusters similarly in three- and four-syllable words. The repetition results of the clusters in the medial position are slightly different: the children aged 4;00-4;11 and 7;00-7;11 repeated the clusters similarly in three- and four-syllable words, while the children aged 5;00-6;11 and 8;00-8;11 repeated the clusters in longer words increasingly worse (see Figure 6).

Generalizing the quantitative analysis results of the non-word repetition test, the following most important aspects might be distinguished: x the results of boys and girls of the general test and of different word structure features (i.e. word length and syllable structure) are either the same or very similar; x older children repeated non-words more accurately. The groups of four- year-old and seven-eight-year-old children are distinct. Some differences have been noted in each group as well: older group members repeated the words more accurately; x it is more difficult for the children to repeat the words longer than three syllables without a consonant cluster and the words longer than two syllables with a consonant cluster; x the position of the cluster in a word is important: the clusters in the initial position are repeated significantly more accurately than the clusters in the medial position.

5. QUALITATIVE ANALYSIS. THE TENDENCIES OF NON-WORD PRONUNCIATION

In order to pronounce the words with a more complicated structure, children look for ways of facilitating their pronunciation; for instance, they omit the consonants with more complicated pronunciation or substitute them by other consonants, which are easier to pronounce. Inaccurate pronunciation of sounds or their substitution by other sounds are a natural development of children’s language, which demonstrates the

23 cognitive processes when learning new words (Dodd et al. 2003: 623; Santos et al. 2006: 372).

5.1. General observations

The Lithuanian non-word repetition test has shown that children apply universal pronunciation simplification strategies: substitution, omission, consonant assimilation, metathesis, sound migration to another syllable, and sound addition. The results of the present research suggest that children could have used associations with the Lithuanian words which are already known, or their forms, in repeating the word that they heard for the first time.

5.2. Substitution

The substitution of vowels with other vowels is especially rare, especially in the age groups older than six; therefore, it is difficult to note certain tendencies of vowel substitution. The following substitution instances have been identified: u was substituted by o, for example, kemo (= kemu), spiratoša (= spiratuša); e was substituted by a, for example, spadaki (= spadeki), žadavina (= žadevina), kamu (= kemu). All age groups often substituted the vowel ė in the final position into e; for instance, gapre (= gaprė). The vowel e is not spread; therefore, it is, possibly, easier to pronounce. It is important to note that the substitution of ė to e in the final position is also typical of adults, and children learn the language from adults, listening to their pronunciation of words and word combinations. A tendency to change a vowel into a was noted in the pronunciation of all research participants; for instance, kiemuo (= kemu), geluoša (= geloša), šviela (= švela). This might be explained by the fact that children do not distinguish phonetic properties because of their phonemic hearing, which is not developed well enough. The phonetic properties of these vowels and are similar. The consonants were substituted depending on airflow both in the syllables with the structure CV and CCV. consonants. All age groups (especially 4;00- 6;11) tend to substitute s by š and vice versa; for instance, šuleritė (= suleritė), škimo (= skimo), gošakluni (= gosakluni), lašmuvi (= lasmuvi). It is suggested that children do not find an appropriate tongue position when pronouncing s and š.

24 consonants. It is typical of all age groups to substitute voiceless consonants by the voiced ones. It is assumed that the consonants which sound similarly, such as g and k or t and d, are more complicated to distinguish aurally. In addition, the place of articulation of voiced and voiceless plosives is the same: p and b are bilabials, t and d are dental consonants, k and g are glottal. In consonant clusters, voiceless plosives were substituted by voiced consonants when the cluster contained a liquid consonant: blemuta (= plemuta), gosagluni (= gosakluni), glesta (= klesta). The plosives with a different place of articulation were also substituted: g was substituted by d, for example, daprė (= gaprė), k was substituted by t, for example, spadeti (= spadeki), d was substituted by b, d, l, or v, for example, žabedina (= žadevina), snaligina (= snalidina), spaleki, spaveki (= spadeki). In consonant clusters, p was often substituted by t or k, for example, mastule, maskulė (= maspule). It can be observed that children tend to use the sounds pronounced at the front of the mouth: velar consonants (g and k) were substituted by dental (t and d) or bilabial (b) because of an easier pronunciation. Liquid consonants. The consonant r, the pronunciation of which is difficult, was substituted only in the age group 4;00-6;11. It was often substituted by l and more rarely by j; for instance, šiluta, šijuta (= širuta). The consonant l was substituted in all age groups by j, r, or v; for instance, maskuje (= maspulė), škurinė, škujinė (= škulinė), snajidina (= snalidina), suvejitė (= suleritė), žadevina (= žadelina). All children substituted the consonant v by j, m, n, and r; for instance, lesmoji, lasmori, lasmuni (= lasmuvi), žademina, žadelina (= žadevina). Children from five years of age substituted the consonant v in the last syllable of a word by plosives d, b, or g; for instance, lasmudi, lasmubi, lasmugi (= lasmuvi). In consonant clusters, v was also substituted by p and t; for instance, štela, špela (= švela). When children substituted one of the consonant cluster elements, the cluster was changed into the one which is more typical of Lithuanian words or easier to pronounce. For example, the cluster sp was often substituted by sk: maskule (= maspulė), niskarima (= nisparima). The cluster sk is more frequent in Lithuanian words than sp (Kazlauskienė 2007: 38). In addition, the cluster sk also appears between morphemes; for instance, in the word skris-k, it marks the future tense. Therefore, children might recognize it better. Another example is the word gaprė, which was

25 pronounced as gatre. The clusters pr and tr tend to appear in Lithuanian words with a similar frequency (11% and 9%, respectively). It is assumed that it is easier to pronounce tr than pr because when pronouncing the former cluster the active articulator remains at the front position of the tongue, while in the pronunciation of pr, the active articulator has to change from lips to the front of the tongue. The substitution of clusters cannot be always explained by their frequency. There are some examples when children pronounced the clusters which appear in Lithuanian words less frequently than the ones in the non-word. It has been determined that the clusters containing a fricative were more often substituted in the medial rather than in the initial position, while the clusters containing liquid or plosive consonants were substituted at the same frequency in the word initial and medial position.

5.3. Sound omission

When repeating the non-words, the respondents almost did not omit vowels. Four-year-old- and eight-year-old children omitted the vowel several times in the last open syllable; for instance, kem (= kemu10), lasmuv (= lasmuvi). The children might not have heard the last sound because of some reasons (absent-mindedness, hurry, etc.). Consonants in CV syllables were also omitted rarely. In all age groups, the omission of the first sound has been observed; for instance, osmuvi (= lasmuvi), asvapi (= pasvapi). It seems likely that children might not have heard the first sound because of absent-mindedness or other reasons. Consonants in CCV syllables were omitted substantially more frequently. It was noticed that if a cluster appears in the initial position, the first element is omitted the most frequently; for example, kimo (= skimo), lemuta (= plemuta) (only the children aged 4;00-4;11 and 7;00-7;11 omitted the first and the second element in the cluster). If the cluster appeared in the medial position, the second cluster element was omitted the most frequently; for example, gosakuni (= gosakluni), gita (= gitva) (only the children aged 7;00-7;11 omitted the first element as well).

10 The omitted sound or syllable is marked in bold in the examples given.

26 The data of the non-word repetition test support the sonority model, which claims that children omit a more sonorous11 consonant in a cluster; for instance, kimo (= skimo), šela (= švela), magunolė (= magvunolė). Syllables are omitted rarely. Most examples were produced by younger children (4;00-5;11). The youngest research participants omitted syllables in the word initial position: deki (= spadeki), in the medial position, and this tendency is more frequent (suletė (= suleritė)), and in the final position: lasmu (= lasmuvi). The children aged 6;00-6;11 omitted a syllable in the word medial position: snalida (= snalidina), suletė (= suleritė) and in the final position: lasmu (= lasmuvi). The children aged 7;00- 8;11 omitted the word final syllable only in a few instances: lasmu (= lasmuvi). The syllable following a stressed syllable was often omitted; for instance, lasmu (= lasmuvi)12, nispari (= nisparima); this is the final syllable the most often. In addition, a stressed syllable was omitted; for example, deki (= spadeki), suletė (= suleritė). The syllables preceding the stressed syllable were omitted the least frequently, for example, stagosa (= staligosa).

5.4. Sound addition

Consonants were added frequently. All age groups added liquid consonants; from the age of five, plosives, and from the age of seven, were added. The sounds pronounced at the front of the mouth dominate (i.e. s, n, l, t). The research has indicated that liquid consonants were the most frequently added in the word medial position (e.g. skurlinė (= škulinė)), and plosives in the word final position; for instance, pasvapti (= pasvapi), spadekėk, spadekik (= spadeki). The children who added an additional consonant, pronounced the clusters which are frequent in Lithuanian words; for instance in syllables nt, nk, ng, st, št, šk, gr or between syllables: pt, rm, lv, ml, and rv (skirmo, lasmulvi, sumleritė). A less frequent tendency was to add a consonant to an already existing consonant and to produce a three-consonant cluster; for instance, mangvunolė (= magvunolė).

11 Sonority hierarchy: the most sonorous sounds are considered to be vowels, and the least sonorous ones – plosive consonants. Thus, the sonority model is based on the following classification: the least sonorous sounds are plosives (p, b, d, t, k, g,); they are followed by (dz, dž, c, č), fricatives (s, š, h, f), liquid nasals (n, m), other liquid consonants (l, r), semivowels (v, j); the most sonorous sounds are vowels (a, e, a, o, u, i) (Marecka, Dziubalska-Kołaczyk 2014: 38; Pater, Barlow 2003: 488). 12 In the examples provided, the underlined syllable is stressed, and the syllable in bold is omitted.

27 The research data have revealed that children pronounced the clusters which appear across morpheme boundaries: št (spiraštuša, spiratušta), šk (=spirastuška), st (spirastuša), nk (spadenki), nt (sulerintė), and within morphemes: ln (švelna), dr (droja), rm (skirmo), lv (žadelvina), ml (kemlo). A syllable was rarely added. The largest number of such instances were observed in the age group 4;00-4;11; for example, damoja (= doja), pispadeki (= spadeki). Only CV structure syllables were added.

5.5. Consonant assimilation

Total assimilation refers to the process when one sound becomes exactly identical to another consonant. Consonant assimilation is more typical of younger children: the most instances were noticed in the age group of 4;00-4;11; in the age groups of older children, this process was rarer. In all age groups, the assimilation of the place of articulation was the most typical; for instance, a labiodental sound becomes bilabial or vice versa: lasmumi (= lasmuvi). In the group of eight-year-old children, assimilation of voicing was noted, i.e. when voiceless consonants become voiced and vice versa: gosagluni (= gosakluni), gabrė (= gaprė). Assimilation affects liquid consonants the most frequently (e.g. lasmumi (= lasmuvi)); liquid consonants assimilate plosives less frequently (e.g. leloša (= geloša)), and plosives affect other plosives the least frequently (e.g. spapeki (= spadeki)). The nearby consonants are assimilated the most frequently; for instance, snanidina (= snalidina), lasmumi (= lasmuvi). Less frequently, the consonants in other syllables are affected; for instance, nisparina (= nisparima), risparima (= nisparima).

5.6. Metathesis and consonant migration

The results of the non-word repetition test have revealed that metathesis, i.e. switching two contiguous sounds, is not a frequent but common phenomenon typical of all age groups. It can be observed the most frequently in the groups of younger children. For all age groups, especially the children aged 4;00-4;11, the metathesis of the consonants l and r was typical in pronouncing a specific word, i.e. surelitė (= suleritė). The articulation of these two consonants is similar: they are liquid front alveolar. In the age groups of 5;00-8;11, some other and more varied instances of metathesis have been

28 observed; for instance: magnuvolė, magvulonė maglunovė (= magvunolė), snalinida (= snalidina), suralitė (= suleritė). It has been found out that the children who switched the sounds pronounced the consonant clusters which are rarer in Lithuanian. It is assumed that repeating the non-word children switched the consonants l and n subconsciously. Sound migration from one syllable to another is the least frequent strategy of pronunciation simplification, even though it was observed in all age groups. The same model of switching the sound was used: one element was shifted from the word-medial consonant cluster to the beginning of the word, and a new consonant cluster was coined: glosakuni (= gosakluni), gvita (= gitva). Despite consonant migration, the cluster of the same type is maintained; for instance, a plosive+ a liquid: pr->gr, kl->gl, tv->gv; a fricative+a plosive: st->sg, sp->sd. It can be assumed that the sounds migrated because they were remembered like this and they seemed to be easier to pronounce at that point.

In general, the present research has revealed that children acquiring Lithuanian apply the pronunciation simplification rules typical of other languages: the dominant strategies are the substitution of sounds, the articulation of which is more complicated or which are rarer, with the sounds with a more common articulation or more common sounds. In addition, sound omission is frequent, while addition, assimilation, metathesis, or sound migration appears more rarely. Despite the application of the same pronunciation simplification strategies, the comparison of the present research with other experimental language acquisition research shows that the results do not always coincide. For example, the results of the non-word repetition test have shown that the children from four to eight years of age substituted fricatives only by fricatives, i.e. the consonant s is substituted by š and vice versa. The results of another experimental research have revealed that fricative consonants are also substituted by plosives (s was substituted by t or z was substituted by d) (Garšvienė, Juškienė 2008; Ruzaitė, Dabašinskienė 2010). In addition, on the basis of the present research, it can be maintained that most six- and seven-year-old children pronounce the consonant r accurately, while other research data have demonstrated that even seven-year-old children do not pronounce the consonant r accurately and substitute it by other sounds (Garšvienė, Juškienė 2008).

29 The differences of the present and other research show various results and possible variants as in early children’s language acquisition processes, a substantial number of individual differences might occur. The qualitative analysis has revealed the strategies of pronunciation simplification of sounds, their combinations, and the words with different structure; however, each child takes individual steps to reach the model of adult language, and the time span might be different in order to reach this aim. Thus, different research data on Lithuanian can only broaden our knowledge about children’s phonotactics acquisition of different age groups.

30 CONCLUSIONS

In order to determine the acquisition features of phonotactics of preschool (4;00-5;11), pre-primary school (6;00-6;11), and junior primary school (7;00-8;11) children, an experimental research of non-word repetition test has been conducted. Summing up the results of the research, the following conclusions can be made: 1. Sex does not have influence on the acquisition of phonotactics: there is no statistically significant difference between test results of boys and girls. Slight differences have been observed only in different age groups: 1.1. General test results are slightly better in the age groups of four- and eight-year-old boys and five- and seven-year-old girls. 1.2. The repetition results of different word structure parts (different word length, words without a consonant cluster, and words with a consonant cluster) were better for boys only in the youngest age group (i.e. four years of age). As for girls, this was the case in the age group of five-year-old and seven-year-old girls. 2. The age of the research participants influences the acquisition of sounds, their clusters, and words with different structure. Older children repeated non-words slightly more accurately. Four-year-old children form a distinct group, the results of which are statistically significantly worse than of other age groups (5;00-8;11). 2.1. Some differences were observed inside each group: older children in the group repeated the words more accurately. The most significant differences were observed in the groups of four-year-old, seven-year-old, and eight-year-old children. 2.2. The comparison of repetition results of different word structural parts, two groups can be distinguished: 1) pre-primary school age and preschool age (4;00-6;11), the results of which were improving with age; 2) school age (7;00-8;11) group, the results of which are very similar or slightly worse than those of younger children. 3. The factor of word structure is important (the word length and syllable structure): 3.1. The children repeated longer words statistically significantly worse than the shorter ones, even though the rhythmical structure (the number of syllables in the word) was retained accurately.

31 3.2. The words without a consonant cluster were repeated substantially more accurately (with the accuracy of approximately 86%) than the words with a consonant cluster (with the accuracy of approximately 72%). 3.3. The results of the test allowed determining the correlation between the word length and a consonant cluster: it is difficult for children to repeat longer than three-syllable words without a consonant cluster and longer than two-syllable words with a consonant cluster. 3.4. The position of a cluster in a word is important: the clusters in the word-initial position were repeated statistically significantly more accurately (with the accuracy of 82-95%) than the clusters in the word-medial position (the accuracy of 68- 79%). 4. The repetition analysis of the words with different structure has revealed the acquisition of Lithuanian phonotactics of children aged from four to eight: 4.1. The children aged from four to eight can pronounce consonant clusters: simplifying the pronunciation of a word, children substituted one element in a consonant cluster with another sound; omission of sounds was not that frequent. The consonant clusters which are rarer were substituted by more typical and more frequent consonant clusters in Lithuanian. 4.2. Children are more used to pronouncing the words with two- element consonant clusters. Repeating the words, the participants tended to add a consonant to CV structure syllables and pronounced the consonant clusters typical of Lithuanian. Two-element consonant clusters typical of Lithuanian were also pronounced shifting a consonant from one syllable to another. 4.3. A closed syllable was pronounced at the word-final position. The children either did not pronounce the final vowel of an open syllable or added a consonant to the final open syllable. 5. Phonotactics acquisition depends on the child’s development of phonological perception and the development of the vocal tract; therefore, the pronunciation simplification strategies have revealed the main pronunciation features of sounds and their combinations typical of children of different age:

32 5.1. Younger children tend to pronounce the sounds articulated at the front of the mouth: s, l, n, t, and d. These consonants were often added, or other consonants were substituted by these sounds. 5.2. Preschool and pre-primary school children do not identify liquid consonants. When repeating non-words, children substituted these consonants by one another unsystematically. For school-age children, it is difficult to pronounce only the consonants l and v accurately. 5.3. It is difficult to distinguish fricative consonants s and š, especially for four-year-old children. The youngest research participants substituted s by š and vice versa; from five years of age, children substituted s by š substantially more often. 5.4. Children tend to pronounce voiced consonants in CV and CVV structure syllables; if they substituted a plosive consonant, they change a voiceless consonant into a voiced one.

33 Publications: 1. Krivickaitė E. 2014: The Research of Phonotactics Acquisition of Typically Developing Children and Children with Specific Language Impairment by Non-word Repetition Test. Special Education 2 (31), 83–105. 2. Krivickaitė E. 2014: Netikrų žodžių kartojimo testas: lietuvių vienakalbių ir dvikalbių vaikų kalbos tyrimas. Taikomoji kalbotyra 4. 3. Krivickaitė E., Dabašinskienė I. 2013: Lietuvių kalbos fonologijos įsisavinimas: netikrų žodžių kartojimo eksperimentinis tyrimas. Kalbos kultūr, 86, 237–255.

The dissertation has been presented at 5 international scientific conferences: 1. Krivickaitė E., Dabašinskienė I., Balčiūnienė I., Kalninytė A. 2014: Reliability of Screening Tests for Lithuanian: Measuring Grammar, Phonological and Narrative awareness. 13th International Congress for the Study of Child Language. Amsterdam 14–18 07 2014 2. Krivickaitė E. 2013: Išgalvotų žodžių testas: Lietuvių vienakalbių ir dvikalbių vaikų kalbos tyrimas. International conference Kalbos ir žmonės: erdvė, laikas, tapatybė / Languages and People: Space, Time, Identity. Vilnius 03–04 10 2013. 3. Dabašinskienė I., Krivickaitė E. 2013: poster Language Delay, Incomplete Learning or SLI: A Non-word Repetition Test for Lithuanian. 9th International Symposium on Bilingualism (ISB9). Singapore. 10–13 06 2013. 4. Krivickaitė E., Dabašinskienė I. 2013: Phonological Awareness of Lithuanian Children: Case Study of Non-word Repetition Test. International conference EAAL 12 th Annual Conference of Applied Linguistics WORDS, WORDS, WORDS: Words in the Language System in Use and in Applications. Tallinn, 18–19 04 2013. 5. Krivickaitė E. 2012: Development of the Non-words Task in Lithuanian. International conference Empowering Children with Non-Typical Speech and Language Development. Kaunas 14–15 09 2012.

The dissertation has been presented at 2 international scientific events: 1. Krivickaitė E., Dabašinskienė I. 2013: Lithuanian Non-word Test. COST members meeting. Lisbon 15 02 2013. 2. Dabašinskienė I., Krivickaitė E. 2012: Lithuanian Non-word Repetition Test: Data from Monolingual and Lithuanian-English Bilingual Children. COST members meeting. Berlin, 14 05 2012.

34 INFORMATION ABOUT THE PHD STUDENT

Eglė Krivickaitė graduated from Vytautas Magnus University in 2005 and was awarded Bachelor’s Degree in Lithuanian Philology. In the same year, she started Master’s Studies in Applied Linguistics at Vytautas Magnus University. After successfully completing Master’s studies, in 2011, Eglė Krivickaitė began her Doctoral studies in Philology Science Field at Vytautas Magnus University with Klaipėda University, Šiauliai University and the Institute of the Lithuanian Language. Since 2011, Eglė Krivickaitė has been working as a Lithuanian language lecturer at Vytautas Magnus University. She teaches Occupational language standards and academic writing and the Lithuanian language for foreigners. Research experience: during her PhD study years, Eglė participated in 3 international projects (in 2011–2013, in COST IS0804 project Language Impairment in a Multilingual Society: Linguistics Patterns and the Road to Assessment; in 2011–2013, in G&G Enhancement2 (no. LLP-GRU-MP-2001-LT-00100) project; since 2015, Eglė joined COST IS0406 project Enhancing Children's Oral Language Skills across Europe and beyond – A Collaboration Focusing on Interventions for Children with Difficulties Learning their First Language) and one national project (2012–2013 Lietuvių vaikų kalbos ilgalaikio stebėjimo tekstyno formavimas (Formation of Longitudinal Corpus of Lithuanian Children Language)). In 2014, Eglė won Erasmus practice grant to City University London for three months. During the period of 2011–2015, Eglė Krivickaitė read papers at international scienfic conferences in Lithuania, Estonia, Netherlands, and Singapore.

35 REZIUMĖ

Disertacijos tikslas – išanalizuoti ikimokyklinio (4;00–5;11), priešmokyklinio (6;00–6;11) ir pradinio mokyklinio (7;00–8;11) amžiaus vaikų fonotaktikos ypatybes taikant išgalvotų žodžių pakartojimo testą. Siekiant tikslo numatyti tokie tyrimo uždaviniai: 1. Pristatyti lietuvių kalbos išgalvotų žodžių pakartojimo testo metodiką. 2. Sukaupti reprezentatyvius vaikų kalbos duomenis išgalvotų žodžių pakartojimo testo metodu. 3. Atlikti surinktų duomenų kiekybinę ir kokybinę analizę: 3.1. Atlikus statistinę duomenų analizę SPSS 17.0 programa, nustatyti: 3.1.1. amžiaus ir lyties įtaką fonotaktikos įsisavinimui; 3.1.2. ketverių–aštuonerių metų vaikų skirtingos struktūros žodžių įsisavinimo tendencijas; 3.2. Nustatyti skirtingų amžiaus grupių taikytas išgalvotų žodžių tarimo paprastinimo strategijas; 3.3. Aptarti ir palyginti kiekvienos amžiaus grupės garsų ir jų junginių tarimo ypatybes. 4. Remiantis kokybinės ir kiekybės analizės rezultatais, nustatyti tipiškiausias 4;00– 8;11 amžiaus lietuvių vaikų kalbos fonotaktikos įsisavinimo ypatybes. Šis darbas yra naujas, nes iki šiol lietuvių vaikų kalbos eksperimentiniai tyrimai dažniausiai buvo atliekami su viena–dviem amžiaus grupėmis, šį tyrimą sudaro 5 amžiaus grupės: 4;00–4;11, 5;00–5;11, 6;00–6;11, 7;00–7;11, 8;00–8;11 amžiaus, t. y. ikimokyklinio (4;00–5;11), priešmokyklinio (6;00–6;11) ir jaunesnio pradinio mokyklinio amžiaus (7;00–8;11) vaikai. Tyrimas pasižymi didele medžiagos apimtimi: kiekvieną amžiaus grupę sudaro po 95–97 tiriamuosius, o ankstesni lietuvių kalbos eksperimentinių darbų medžiagą sudarydavo iki 90 tiriamųjų. Kalbos įsisavinimo duomenis rinkti pirmą kartą taikytas lietuvių kalbos išgalvotų žodžių pakartojimo testas. Tiriamąją medžiagą sudaro 480 ketverių–aštuonerių metų tipinės kalbos raidos vaikų testo duomenys, rinkti Kauno miesto, Kauno rajono ir Kauno apskrities miestų darželiuose ir mokyklose. Kiekvienoje amžiaus grupėje stengtasi išlaikyti panašų tiriamųjų skaičių bei lyties proporciją.

36 Disertacijos tyrimas atliktas taikant ir derinant tarpusavyje keletą metodų: skersinio pjūvio metodu sudaryta tiriamųjų imtis, eksperimentiniu metodu (išgalvotų žodžių pakartojimo testu) rinkta tyrimo medžiaga, kuriai analizuoti taikyti klaidų analizės, gretinamasis ir statistiniai metodai. Lietuvių kalbos išgalvotų žodžių pakartojimo testas kurtas specialiai šiam tyrimui, todėl darbe jis apibūdintas išsamiau: aptartas testo taikymas užsienio kalbų tyrimuose, taip pat aprašyta lietuvių kalbos išgalvotų žodžių testo struktūra. Netiksliai pakartoti žodžiai analizuoti taikant kokybinį klaidų analizės metodą. Lyginant skirtingų amžiaus grupių, taip pat berniukų ir mergaičių testo rezultatus, taikytas gretinamasis metodas. Statistiniai metodai taikyti aptariant lyties ir amžiaus veiksnių įtaką fonotaktikos įsisavinimui. Tarpusavyje derinant visus šiuos metodus tikimasi nustatyti pagrindines lietuvių kalbos fonotaktikos įsisavinimo ypatybes skirtingose amžiaus grupėse. Darbo rezultatai gali būti aktualūs ir naudingi rengiant praktines kalbos lavinimo užduotis tipinės kalbos raidos vaikams. Nustačius, kokius kalbos garsus ir jų junginius bei kokios struktūros žodžius skirtingo amžiaus vaikams sudėtinga ištarti, darželių auklėtojai, mokytojai, logopedai bei tėvai galėtų daugiau dėmesio skirti jų tarimo lavinimui. Gauti rezultatai atskleidžia tipinės kalbos raidos garsų ir jų junginių bei skirtingos struktūros žodžių įsisavinimo ypatybes ketverių–aštuonerių metų vaikų kalboje, todėl logopedams ir spec. pedagogams galėtų būti naudingi dirbant ir su kalbos sutrikimų turinčiais vaikais: padėtų diagnozuoti fonologinius kalbos sutrikimus, kurti jų koregavimo metodiką. Šio tyrimo metodologija ir tyrimo rezultatai turi perspektyvą tapti tolimesnio mokslinio darbo pagrindu tiriant vaikų, turinčių kalbos sutrikimą, ir / ar dvikalbių / daugiakalbių vaikų lietuvių kalbos fonotaktikos dėsnių įsisavinimo ypatybes. Didelė skirtingo amžiaus vaikų grupių imtis, gautų duomenų apdorojimas statistine programa bei išsami surinktų duomenų kokybinė analizė – tai svarbus pagrindas ateičiai, kad išgalvotų žodžių pakartojimo testas taptų standartizuotu lietuvių kalbos fonotaktikos įsisavinimo vertinimo įrankiu. Darbo struktūra: disertaciją sudaro įvadas (pirmoji dalis) ir keturios pagrindinės dalys. Antra dalis Fonologijos įsisavinimas skirta fonologijos ir fonotaktikos įsisavinimo ypatybėms aptarti. Trečioje dalyje Vaikų kalbos tyrimai išgalvotų žodžių pakartojimo metodu aptartas išgalvotų žodžių pakartojimo testo taikymas užsienio kalbų tyrimuose, pristatyta lietuvių kalbos išgalvotų žodžių testo

37 metodologija ir tiriamoji medžiaga. Ketvirtoje dalyje Fonotaktika: kiekybinė duomenų analizė aprašyti 4;00–8;11 amžiaus berniukų ir mergaičių skirtingos struktūros žodžių pakartojimo rezultatai, apdoroti statistine programa. Penktoje dalyje Kokybinė analizė. Išgalvotų žodžių tarimo tendencijos nustatyti kiekvienai amžiaus grupei būdingi garsų ir jų junginių tarimo ypatumai. Šiame skyriuje, remiantis nauja kalbos įsisavinimo tyrimų rezultatų aprašymo metodika, pirmiausia pristatomi susisteminti tyrimo rezultatai, be išsamesnio jų aiškinimo; paskutiniame penktojo skyriaus skyrelyje Apibendrinimas ir diskusija rezultatai yra analizuojami, aptariamos galimos jų interpretacijos, gauti duomenys lyginami su kitais lietuvių bei kitų kalbų įsisavinimo tyrimų rezultatais. Ketvirta ir penkta darbo dalys užbaigiamos trumpu viso skyriaus apibendrinimu. Darbo pabaigoje pateikiamos išvados (šeštoji dalis), cituotų literatūros šaltinių aprašai ir priedai, papildantys darbo medžiagą. Atlikus tyrimą ir apibendrinus gautus rezultatus, pateikiami ginamieji teiginiai: 1. Lytis neturi statistiškai reikšmingos įtakos fonotaktikos dėsnių įsisavinimui. 2. Fonotaktikos dėsnių įsisavinimas susijęs su tiriamųjų amžiumi: vyresni vaikai geba tiksliau ištarti skirtingos struktūros žodžius, taip pat juos sudarančius garsus ir jų junginius. 3. Garsų ir jų junginių ištarimo tikslumas priklauso nuo žodžio struktūros: tiksliai garsą ištarti sunkiau ilgesniuose žodžiuose ir žodžiuose, turinčiuose priebalsių samplaiką. 4. Gebėjimas pakartoti skirtingos struktūros žodžius susijęs su lietuvių kalbos fonotaktika: kartodami pirmą kartą išgirstą žodį, vaikai tiksliau ištaria tuos garsų junginius, kurie būdingi lietuvių kalbos žodžiams ir dažniau klysta tardami retai kalboje pasitaikančius junginius. 5. Garsų ir jų junginių įsisavinimo raidą atspindi skirtingo amžiaus vaikų taikomos universalios žodžio tarimo paprastinimo strategijos.

38 REFERENCES:

1. Auksoriūtė V. 2010: Ketverių metų vaikų gebėjimas tarti kalbos garsus. Bakalauro darbas. Šiauliai: ŠU. 2. Baddeley A., Gathercole S. E., Papagno C. 1998: The Phonological Loop as a Language Learning Device. Psychological Review 1, 158–173. 3. Chiat Sh., Roy P. 2007: The Preschool Repetition Test: Am Evaluation of Performance in Typically Developing and Clinically Referred Children. Journal of Speech, Language, and Hearing Research 50, 429–443. 4. Dabašinskienė I., Krivickaitė E. 2013: poster Language Delay, Incomplete Learning or SLI: A Non-word Repetition Test for Lithuanian. 9th International Symposium on Bilingualism (ISB9). Singapore. 10–13 06 2013. 5. Dispaldro M., Leonard L. B., Deevy P. 2013: Real-Word and Nonword Repetition in Italian-Speaking Children with Specific Language Impairment: A Study of Diagnostic Accuracy. Journal of Speech, Language, and Hearing Research 56, 323–336. 6. Dodd B., Holm A., Hua Z., Crosbie Sh. 2003: Phonological Development: A Normative Study of British English-speaking Children. Clinical Linguistics & Phonetics 17 (8), 617–643. 7. Dressler, W. U. 1993: What is Natural in Natural Morphology (NM)? Prague Linguistic Circle Papers. Prague: Charles University, 135–145. 8. Dressler W. U., Dziubalska-Kołaczyk K., Rossella S. 2001: Sources of Markedness in Language Structures. Folia Linguistica Historica 22 (1-2). 103– 135. 9. Dressler W. U. 2005: Word-formation in Natural Morphology. Handbook of Word-Formation. Springer Netherlands, 267—284. 10. Dollaghan Ch., Campbell T. F. 1998: Nonword Repetition and Child Language Impairment. Journal of Speech, Language, and Hearing 41, 1136–1146. 11. Donegan P., Stampe S. 2009: Hypothesis of Natural Phonology. Poznan Studies in Contemporary Linguistics 45 (1), 1–31. 12. Dziubalska-Kołaczyk K. 2007: Natural Phonology: Universal Principles for the Study of Language (Insiders Meet Outsiders). Saarbrücken 6 (10), 71–75. 13. Dziubalska-Kołaczyk K. 2014: Explaining Phonotactics Using NAD. Language Sciences, 6–12. 14. Dziubalska-Kołaczyk K. 2015: Are Frequent, Early and Easy Clusters also Unmarked? Italian Journal of Linguistics 27 (1), 29–44. 15. Encyclopedia of Language Development 2014, (red) Brooks P. J., Kempe V. (Eds.) Thousand Oaks, CA: Sage Publications. 16. Garšvienė A., Ivoškuvienė R. 1993: Logopedija: Vadovėlis specialiosios pedagogikos fakulteto studentams. Kaunas: Šviesa. 17. Garšvienė A., Juškienė G. 2008: Priešmokyklinio amžiaus vaikų tarimo ypatumai. Jaunųjų mokslininkų darbai 2 (18), 78–84.

39 18. Gathercole S. E. 1995: Is Nonword Repetition a Test of Phonological Memory or Long-term Knowledge? It all Depends on the Nonwords. Memory & Cognition 23 (1), 83–94. 19. Gathercole S. E. 2006a: Nonword Repetition and Word Learning: The Nature of the Relationship. Applied Psycholinguistics 27, 513–543. 20. Gathercole S. E. 2006b: Complexities and Constraints in Nonword Repetition and Word Learning. Applied Linguistics 27, 599–613. 21. Girbau D., Schwartz R. G. 2007: Non-word Repetition in Spanish-speaking Children with Specific Language Impairment (SLI). International Journal of Language & Communication Disorders 42, 59–75. 22. Girdenis A., Karosienė V. 2010: Bendrinės lietuvių kalbos statistinė struktūra: fonologijios dalykai. Vilnius: Vilniaus universiteto leidykla. 23. Harley T. A. 2010: The Psychology of Language. Psychology press. 24. Kamandulytė L. 2006a: Morfonotaktikos įsisavinimas. Didaktika 1, 84–88. 25. Kamandulytė L. 2006b: The Acquisition of Mophonotactics in Lithuania. Winer Linguistische Gazette 73, 88–96. 26. Kamandulytė-Merfeldienė L. 2015: Morphonotactics in L1 Acquisition of Lithuanian: TD vs. SLI. Eesti Rakenduslingvistika Ühingu aastaraamat 11, 95– 109. 27. Kavitskaya D., Babyonushev M., Walls T., Grigorenko E. 2011: Investigating the Effects of Syllable Complexity in Russian-speaking Children with SLI. Journal of Child Language 38, 979–998. 28. Kazlauskienė A. 2007: Intervokaliniai priebalsiai: vienanarės ir dvinarės grupės. Kalbų studijos 11, 36–42. 29. Kazlauskienė A. 2010: Lietuvių kalbos žodžių foneminės struktūros dėsningumai. Žmogus ir žodis 1, 35–41. 30. Kazlauskienė A., Raškinis G. 2008a: Lietuvių kalbos fonologinio skiemens struktūrinių modelių dažnumas. Žmogus ir Žodis 1, 24–31. 31. Kazlauskienė A., Raškinis G. 2008b: Intervokaliniai priebalsiai: trinarės, keturnarės ir penkianarės grupės. Kalbų studijos 13, 51–59. 32. Kunnari S., Tolonen A-K., Chiat Sh. 2011: Language Impairment in a Multilingual Society: Linguistic Patterns and the Road to Assessment. http://www.bi-sli.org 33. Marecka M., Dziubalska-Kołaczyk K. 2014: Evaluating Models of Phonotactic Constraints on the Basis of sC Cluster Acquisition Data. Language Science, 37– 47. 34. Marshall C. R., van der Lely H. K. J. 2009: Effects of Word Position and on Onset Cluster Production: Evidence from Typical Development, Specific Language Impairment, and Dyslexia. Language 85 (1), 39–57. 35. McLeod Sh., van Doorn J., Reed V. A. 2001: Normal Acquisition of Consonant Clusters. American Journal of Speech – Language Pathology 10 (2), 99–110.

40 36. Munson B., Kurtz B. A., Windor J. 2005: The Influence of Vocabulary Size, Phonotactic Probability, and Wordlikeness on Nonword Repetitions of Children With and Without Specific Language Impairment. Journal of Speech, Language, and Hearing Research 48, 1033–1047. 37. Park H., Schwarz I. 2012: Nonword Repetition Performance of Early School-Age Children in Korea: A Preliminary Study. Asia Pacific Journal of Speech, Language, and Hearing 15 (3), 178–196. 38. Pater J. Barlow J. A. 2003: Constraint Conflict in Cluster Reduction. Journal of Child Language 30, 487–526. 39. Radeborg K., Barthelom E., Sjoberg M., Sahlen B. 2006: A Swedish Non-word Repetition Test for Preschool Children. Scandinavian Journal of Psychology 47, 187–192. 40. Rispens J., Parigger E. 2010: Non-word Repetition in Dutch-speaking Children with Specific Language Impairment with and without Problems. British Journal of Developmental Psychology 2010 (28), 177–188. 41. Ruzaitė J., Dabašinskienė I. 2010: Specific Language Impairment: Adaptation of a Screening Test for Lithuanian. Darbai ir Dienos 54, 277–300. 42. Santos F. H., Bueno O. F. A., Gathercole S. E. 2006: Errors in Nonword Repetition: Bridging Short- and long-term Memory. Brazilian Journal of Medical and Biological Research, 39, 371–385. 43. Velleman S. L., Vihman M. V. 2007: Phonology Development in Infancy and Early Childhood: Implications for Theories of Language Learning. Phonology in Context. Luton: Palgrave Macmillan, 25–50. 44. Zanobini M., Viterbori P., Saraceno F. 2012: Phonology and Language Development in Italian Children: An Analysis of Production and Accuracy. Journal of Speech, Language, and Hearing Research 55, 16–31.

41

Eglė KRIVICKAITĖ

PHONOTACTICS DEVELOPMENT OF CHILDREN’S LANGUAGE

Summary of Doctoral Dissertation

Spausdino – Vytauto Didžiojo universitetas (S. Daukanto g. 27, LT-44249 Kaunas) Užsakymo Nr. K16-020. Tiražas 35 egz. 2016 04 14. Nemokamai.