ŽODŽIŲ SUVOKIMO IR EKSPRESYVIOSIOS KALBOS YPATUMAI NEIŠPLĖTOTOS KALBOS ATVEJU

The article discusses the features of the vocabulary of children with developmental language disorder. The quantitative analysis of the children’s language corpus and the results of experimental tasks helped to reveal the peculiarities of word comprehension and production. The results of children with developmental language disorder (n=80) and children with typical language development (n=80) have been compared. The study revealed that children with developmental language disorder performed statistically significantly worse on tasks of comprehension and production of nouns and verbs. Difficulties in comprehension and production of words are determined by the rare use of the word in the child’s environment, the length, the structure of the word, and the abundance of words of similar meaning.


Introduction
Disorders of language expression and comprehension negatively affect the child's learning, socialization, and emotional well-being. Specific language impairment (or developmental language disorder) is an underdevelopment of the whole language system, including the pronunciation of sounds, the discrimination of similar sounds, the development of vocabulary, grammatical structure and cohesive language. In the Lithuanian Classification of Speech and Language Disorders (MES, 2011), the term language underdevelopment is employed, however, recently, the terms specific language impairment or delayed language development are often used to describe children's language disorders. The concept of underdeveloped language has begun to be used because of the change of the medical approach to speech and language disorders to pedagogical and the focus not on the causes of the disorder but on its structure. According to the Document for Identifying Groups of Pupils with Special Educational Needs and Levels of Their Special Educational Needs (MES 2011), three levels of underdeveloped language are distinguished: mild, moderate, and severe.
In the scientific literature of foreign countries, the term specific language impairment is used to describe children's language disorders. Dabašinskienė and Ruzaitė (2010) indicate that a specific language disorder is diagnosed in children of typical development who do not have brain damage, hearing or intellectual impairment. Specific language impairment is characterized by disorders of vocabulary, grammar and phonology, delayed development of the entire language system. One more term for language disorders of children is a developmental language disorder. ASHA's (2016) scope of professional practice states that language development disorders are characterized by impairment of language comprehension and/or expressive language, writing or other character system. Features of language disorder include several areas: language form (phonology, morphology, syntax), language content (semantics), and language use (pragmatics). According to ICD-10 AM (2015), disorders of language expression (F 80.1) and language comprehension (F 80.2) are distinguished. Expressive language disorder is a specific developmental disorder in which the child's ability to use expressive language (language expression) is well below the appropriate mental age level, together with typical language comprehension. Speech sound disorders may or may not exist.
Language comprehension disorder is defined as a specific developmental disorder in which the child's ability to understand language is lower than is typical for a given mental age. There is almost always impaired expression, frequent mispronunciation of words and sounds. In both Lithuanian and foreign literature, language disorders are classified into primary and secondary. Primary language disorders are those that occur in the absence of other disorders (Paul, Norbury, & Gosse, 2018;Bishop, Snowling et al., 2016). Secondary language disorders are associated with a primary diagnosis, such as intellectual or hearing impairment.
According to the data of the Ministry of Education and Science (MES, 2017), the number of preschool children with special educational needs is increasing every year in Lithuania, the majority of them are children with speech and language disorders. About 15% of pre-school and 7% of pre-primary school age children have language disorders that determine the child's further development, behavior and mental health (Petrulytė, Perminaitė, & Mikulėnaitė, 2020). It is stated that in Lithuania in 2016 about 84.3% of preschool children with special educational needs had speech and language disorders (MES, 2017).
Language disorders can be related to the variety of causes. The development of language comprehension and expressive language can be impaired due to cerebral cortex, damage to language areas, childbirth, or the early period. Depending on the location of the impairment, the child may be diagnosed with a disorder of language expression, comprehension, or language expression and comprehension. Paul, Norbury and Gosse (2018) indicate that language disorder results from a complex interaction of genetic and environmental factors. Signs of underdeveloped language are difficulty in comprehending and memorizing new words, longer sentences, narrow vocabulary, often misused grammatical forms, inaccurate word order in a sentence. Simple syntactic structures, short sentences, difficulties in comprehending and pronouncing words with more complex phonetic structures are typical (Kunari et al., 2011;Leonard, 2014;Paul et al., 2018;Garšvienė, Dapšienė, & Gudavičienė, 2001;Ivoškuvienė, Mamonienė et al., 2002). Although the signs of the disorder are heterogeneous, a linguistic profile is distinguished in preschool children with language disorders.
Children's language disorders are a well-known problem that has been researched and solved for many decades in Lithuania. With the active participation of Lithuanian researchers in international research projects, the language development of Lithuanian children is increasingly included in experimental comparative crosslinguistic research (Dabašinskienė, 2014;Dabašinskienė & Kamandulytė-Merfeldienė, 2014;Gagarina, Klop et al., 2012;2015;Kamandulytė-Merfeldienė, Fürst et al., 2010;Armon Lotem, Haman et al., 2016).
Research on children's language disorders in Lithuania is traditionally classified as social sciences (educology), however, based on the experience of other countries (Reilly, Bishop et al., 2014), it is clear that a systematic approach to language and language disorders requires interdisciplinary methodological approaches and different cooperation between specialists in the fields of linguistics, psychology, psycholinguistics, special pedagogy and speech language therapy. In Lithuania, research on language disorders is more focused on one or another group of speech or language disorders, and there is a lack of integration of research and its results based on an interdisciplinary approach. A review of research shows that the aspects of overcoming phonetic and phonological disorders (Garšvienė & Ivoškuvienė, 2003;Kairienė, 2015), fluency disorders (Ivoškuvienė & Makauskienė, 2019;Ivoškuvienė, Gerulaitis, Makauskienė et al., 2010), written language disorders (Gedutienė, 2017;Unčiurys, 2001), undeveloped children's language (Ivoškuvienė, Mamonienė et al., 2002) have been widely studied. Language disorders in preschool children have been less studied, although they usually affect the entire language system and often lead to written language disorders. In Lithuania, too little attention is still paid to children's language research, various issues of speech language therapy are usually analyzed, and there is still a lack of research that can objectively reveal the variety, frequency and peculiarities of children's language disorders.
It is important to mention that the lack of research on children's language disorders in Lithuania is partly due to the lack of appropriate methodological approach and research sources. Although a number of studies on first language acquisition in early childhood have been conducted (Savickienė, 2007;Dabašinskienė, 2010;Dabašinskienė & Kamandulytė-Merfeldienė, 2017;Kamandulytė-Merfeldienė, 2015;2020;Balčiūnienė & Kamandulytė-Merfeldienė, 2018a, etc.), data on the peculiarities of language disorders are still lacking. 2018-2020 Vytautas Magnus University project 'The symptoms of developmental language disorder at preschool and pre-primary school age: spontaneous analysis and experimental studies' (funded by the Lithuanian Research Council (LMTLT), No LIP-18-36; supervisor Dr. Laura Kamandulytė-Merfeldienė) developed a language corpus of children with developmental language disorder and a database of experimental test results. This project provided conditions for researchers of Vytautas Magnus University to conduct experimental research and to compile a corpus of narratives and dialogues of 80 children with delayed language development and 80 children with typical language development.
Statistical analysis of these specialized sources allowed conclusions to be drawn about the characteristics of children's typical and underdeveloped language. The research helped to achieve the main goal -to reveal the peculiarities of language disorders based on the results of quantitative and qualitative analysis, including phonological, lexical, grammatical and pragmatic features. Based on the results of the project, as well as new researches and good practice from other countries, the methodology 'Urtė and Motiejus' for the assessment of the language of preschool and pre-primary school children was restandardized.
This article focuses on one aspect of language disorders -vocabulary, however, the results of this study correlate with morphosyntactic, pragmatic, and phonological studies.
The aim of the research described in this article is to investigate the lexical development of children (4;6-6;5 1 ) with language disorders. To achieve the goal, the following objectives have been set: 1) to perform a quantitative analysis of the corpus of children's language and to discuss the indicators of the lexical diversity; 2) to perform a qualitative analysis of the corpus and to discuss semantic errors; 3) to perform the analysis of the results of the experimental comprehension task and to describe the peculiarities of comprehension of nouns and verbs; 4) to perform the analysis of the results of the experimental word production task and to describe the peculiarities of production of nouns and verbs. The results of children with language disorders and children with typical language development are compared.
Several methods were used in the study. The natural observation method was used to collect data on children's spontaneous language (dialogues). Data were processed and analyzed using the method of corpus linguistics: dialogues of all children were transcribed, annotated and analysed using CHILDES program (MacWhinney, 2000). The study of word comprehension and production was conducted using an experimental research method. The children were tested in their educational institution, and the tasks of selecting and naming pictures were performed. The statistical package "SPSS / w 24.0" and MS Excel 2016 were used to perform the quantitative analysis of the experimental research data and to evaluate the obtained research results.

The features of word comprehension and production in children with language disorders
Studies of children with language disorders are important for several reasons. Extensive knowledge of the disorder allows for more accurate methods of assessment and speech language therapy; successful prevention and early intervention programs are based on research. The earlier signs of a language disorder are identified, the more effective is early intervention. Research on children with language disorders is important in developing theories of first language acquisition, developing assessment methodologies and support strategies. Leonard (2014) points out that grammatical regularity of language that does not correspond to age is a feature of children's language disorder characteristic of children who speak different languages. All over the world, children learn the grammar of their mother tongue with the same consistency, which is explained by a usage-based model (Tomasello, 2003). First, children begin to use the present tense, the most common inflections of nouns, diminutives. The rarer inflections and prepositions, the plural forms, the agreement and other rarer elements are acquired later. Studies show that children with language disorders omit the morphosyntactic markers of future, past, and have difficulties to use prepositions and affixes (Rice, Hoffman & Wexler, 2009). It is stated that it is difficult for children with language disorders to understand passive constructions (e.g., The boy is kissed by the girl, Marshall, Marinis & van der Lely, 2007;Armon-Lotem, Haman et al., 2016), WH questions (Friedmann & Novogrodsky, 2011), relative clauses, especially object clauses (Stavrakaki, 2001;Novogrodsky & Friedmann, 2009).
Studies of various typologically different languages show that children with language disorders have particular difficulty in mastering the adjective lexicon, as well as generalizing, overextending and underextending of the meanings.
McGregor, Oleson, Bansen and Duff (2013), Sheng, Bedore, Peña, and Fiestas (2013) and other studies reveal that children with language disorders learn new words slowly, have difficulty remembering the meanings of words, make mistakes when using already known words. These children need more time to name pictures (Lahey & Edwards, 1996), and the acquisition of new words takes two to three times longer than that of peers with typical language development (Gray, 2003;Rice, Oetting, & Marquis 1994). In addition, children with language disorders have more difficulties remembering the words they learned a few days ago. Language disorder is characterized not only by a smaller vocabulary but also by weaker associations between the words and their semantics, as learned words are associated with fewer semantic features than in the case of typical language development (Rice et al., 1994).
One of the main signs of language disorder is insufficient social communication skills. Children have difficulty understanding anecdotes, figurative language, metaphors (Alt, Plante & Creusere, 2004;Norbury, 2004;Davies, Andres-Roqueta, & Norbury, 2016). The language pragmatic and social communication skills of children with language disorders develop late (Gibson et al., 2013). Compared to peers whose language development meets age norms, they have been found to have a lower social understanding and application of the situations, context, and pragmatic rules (Farrar et al., 2009;Nilsson & de Lopez, 2016;Taylor, Maybery, Grayndler & Whitehouse, 2015). Children have difficulty integrating language and context, making appropriate conclusions, and creating coherent narratives (Reed, Patchell et al., 2007;Karasinski & Weismer, 2010;Norbury, Gemmell, & Paul, 2014). Most researchers associate such difficulties with long-term and short-term memory abilities, making it difficult for children with language disorders to repeat new or less commonly used words. Language impaired children also have difficulties repeating meaningful sentences that include uncommon grammatical categories or rarer words, therefore the repetition tasks are often included in standardized language assessment tests such as GAPS ( Van der Lely et al., 2007).

Research source and methods
In order to identify differences between the impaired and typical development of language, a representative source -a database created at Vytautas Magnus University -was used. This database includes: 1) The corpus of spontaneous language (dialogues) of 80 children with developmental language disorder (DLD) (4;6-6;5).
The research source was collected during the LMT-funded project 'The symptoms of developmental language disorder at preschool and pre-primary school age: spontaneous analysis and experimental studies' (S-LIP-18-36). The methodology of data collection and research was developed by the project promoters, based on the experience of longitudinal and experimental research on language development. Longitudinal and experimental studies of Lithuanian as first language acquisition were started by researchers of Vytautas Magnus University (led by Prof. I. Dabašinskienė), participating in international research and cooperation projects 2 .
The corpus of the spontaneous speech was collected using the method of natural observation and the task 'what's wrong?'. The aim of this task is to discuss with the child a picture of unrealistic objects (e.g., piglets are sitting in a tree, children are skating in summer, fishes are in the sky).
During the task, the researcher discusses the drawing with the child (Why can't it be like this? Maybe the piglets like to live in a tree because everything can be seen better from there?), encourages involvement in the dispute, structures the conversation so that the child focuses not on their own language but on the content of the conversation. This is especially important when studying children with DLD. Children realize quite early on that they have language difficulties and learn to hide them: they use gestures, try to speak in short sentences and use words that they know how to pronounce correctly. By focusing on linguistic expression, children with DLD are able to avoid lexical and/or grammatical errors. Meanwhile, directing the child's attention to the purpose of the conversation (to prove his/her opinion to the partner of conversation) reveals the actual level of communication skills. In this method, the language data of each child were collected individually (duration -approximately 15-20 minutes), and the conversations were recorded on a dictaphone.
The experimental task used in the research was developed in the COST Action IS0804 (2010-2013) 'Language Impairment in a Multilingual Society: Linguistic Patterns and the Road to Assessment' (www.bi-sli.org). A battery of tools under the umbrella name of LITMUS (Language Impairment Testing in Multilingual Settings) was developed (Armon- Lotem, de Jong & Meir, 2015) during this project. One of these tools is the Cross-linguistic Lexical Tasks (LITMUS-CLT), which assesses both receptive and expressive knowledge of nouns and verbs through a picture-based test.
The experimental task consists of word comprehension and word production parts. During the word comprehension task for children the series of 64 pictures presented in a slide format on a computer screen were shown: 32 for the assessment of the comprehension of nouns, 32 for the assessment of the comprehension of verbs. Each series consisted of 4 pictures, from which the child had to choose one noun or verb indicated by the researcher.
The word production test consists of two parts: the first is for naming nouns and the second is for naming verbs. Each part has 32 pictures (64 pictures in total) presented in a slide format. Each test slide displays a picture and the child was asked "Who is here?" or "What is he/she doing here?" with the purpose to encourage the child to name the object or action.
All tasks were carried out with test subjects by speech therapists from the institutions they attend. Before the testing procedure, informed parental consents, consisting of an information sheet and a certificate of consent, were collected. Children were tested individually in the speech language therapist's office. The entire test was recorded on a dictaphone, then the audio records were transcribed and the results of the experimental study were compiled into a database.
With the purpose to identify children with DLD (diagnosed by a speech language therapist and/or approved by the Pedagogical Psychological Service) without other developmental disorders the assistance of a psychologist was used. All children planned to be included in the study performed the Raven's Coloured Progressive Matrices Test, which is the most common in psycholinguistic research and is standardized in Lithuania. This test researches children's non-verbal, productive, inductive thinking skills. As well all children were tested with Bender-Gestalt II test (standardized in Lithuania in 2018) to assess visual and motor abilities and visual memory. Children with belowaverage abilities in these tests were not included in the study group, assuming that the identified language disorder or typical language development may be accompanied by other undiagnosed developmental disorders.

Lexical diversity and the difficulties in expressive language: an analysis of spontaneous speech
A study of spontaneous speech of children with DLD and TD children allowed to compare the variety of lexemes used, as well as the types and frequency of semantic errors. The article is limited to the analysis of lexical features of children's language, however, it is important to mention that the results of our researches highlighted that delayed language is more characterized by morphosyntactic difficulties. Vocabulary errors are often related to grammatical difficulties, e.g. children choose the wrong prefix, ignore the rules of word derivation. The errors in word semantics are less frequent.
The analysed corpus consists of 78583 words (36255 words were used by TD children, 42328 by children with DLD). The analysis of the results showed that the lexical diversity did not differ in the two groups of children: type and token ratio in the group of children with DLD averaged 0.37 (i.e. 37 different words in a 100-word conversation) and in the group of TD children -0.38 (i.e. 38 different words in a 100-word conversation). These data show that children in both groups use a poor vocabulary, often repeat the same words, therefore we can state that the low lexical diversity cannot be associated with language impairment. Although there are no differences between the two groups in assessing the lexical diversity, it has been found that the distribution of the parts of speech differs within these groups.
The data in the first table show that fewer nouns were used in the dialogues of children with DLD than of TD children (18.6% and 23.4% of all the parts of speech). This difference can be explained by the frequent use of pronouns in the case of DLD (12.1% cf. 8.5% in the case of typical development). An analysis of the dialogues shows that this trend can be attributed to the tendency of children with language disorders to replace nouns with the pronouns that and this. Another difference that has emerged when comparing parts of speech is the frequency of adjectives. In the speech of children with DLD, adjectives make up only 0.8%, while in the speech of TD children they make up 2.4% (Table  1). This confirms the already known fact that this part of speech is difficult for children with DLD. The frequency of onomatopoeia also varies. Children with DLD used onomatopoeia several times more often than TD children (1.1% and 0.15%, respectively), for example, the verb to fall was replaced by boom, to jump -opa. In response to the speech therapist's question, What is the bird doing in the picture?, children often used the onomatopoeia ku-ku, etc. Research shows that in the case of typical development, the use of onomatopoeia decreases in about the second or third year and since then its use is mostly stylistically motivated (Balčiūnienė & Kamandulytė-Merfeldienė, 2018b). In this study, children with language disorders used onomatopoeias as equivalents of nouns or verbs, which may mean that the early disclaim of onomatopoeias is natural for a typical language development and is not so characteristic of children with DLD.
An analysis of semantic errors in children's spontaneous speech showed that this type of error was several times more common in the cases of language disorders than in the typical language development (306 and 44 semantic errors, respectively). In TD children's group, the most common mistakes were related to replacing an unknown word with another word that identifies a visually similar object or action, such as: (1) Vaikas: Varlės nešokinėja per lanką.
Child: Frogs do not jump through the bow.
Researcher: Frogs do not jump over the jumping rope.

Tyrėjas: Ugnį užgesinti, taip?
Researcher: And it seems to me that he is a firefighter. What should he do?
Researcher: Put out the fire, yes? Child: And this whole event is not needed.

Researcher: No? Why?
Child: Because summer is very hot.

Comment:
The child uses the word event instead of clothes because of the similar sounding of the lexemes renginys 'event' and rengti 'to dress'.
Children with DLD make more semantic mistakes. Most of them involve the replacing of an unknown word with a pronoun (12,13,14) or with a phrase that attempts to describe an object or action (15,16). This is especially true for children with moderate to severe language disorders (underdeveloped language). In contrast to typical language development, these children often find it difficult to choose a different lexeme for an unknown word based on the similarity of objects or other semantic word relationships:

Researcher: Doing what?
Child: It must be washed so that it does not burn.
Summarizing the results of the analysis of semantic errors, it can be stated that underdeveloped language is characterized not only by a higher number of semantic errors, but also by a lack of communication competence and compensation strategies. TD children usually replace an unknown word with another semantically related word, and children with DLD often do not find a similar word, replace it with pronouns (those, this, that, such), adverbs (there, here), incoherent word combinations. Insufficient ability to compensate for lack of vocabulary limits the ability to communicate, does not allow to express thoughts and develop narratives.

Word comprehension and production: analysis of experimental test results
Experimental tests were applied to the study of vocabulary comprehension and production. The analysis of the results allowed to determine the differences of the two children's groups according to the children's language level (typical language development vs language disorder) and age (4;6-5;5 vs 5;6-6;5), as well as to compare the comprehension and production skills of nouns and verbs.
The results of the noun and verb comprehension test are presented in Table  2, which shows the number of errors in the groups. As expected, for children in all groups a task of understanding nouns was the simplest -even children with DLD made a few mistakes: in age group 4;6-5;5 errors accounted for 4.5% of all 1280 words, 5;6-6;5 -3.1%. TD children very rarely made mistakes: there were only 1.4% of errors in the group of 4;6-5;5 years, in the group of 5;6-6;5 -only 1.2%. Although there were a few errors in the comprehending of nouns in all groups, the statistical analysis revealed significant differences. In the group of children with DLD, age was found to be an important factor in determining the ability to comprehend nouns: comparing the groups of 4;6-5;5 and 5;6-6;5, a statistically significant difference was obtained p = 0.020 (*). There is no statistically significant difference between groups of TD children of different ages. It was also found that children with DLD performed statistically significantly lower on the task of comprehending nouns than TD children (4;6-5;5 children with DLD vs 4;6-5;5 TD children p = 0.007 (**); 5;6-6;5 children with DLD vs 5;6-6;5 TD children p = 0.000 (***)).
Children in all groups had more difficulties in completing the verb comprehension task. In the group of 4;6-5;5 years old children with DLD, the cases of misunderstanding accounted for 18.1%, and older children in this group were slightly less likely to make mistakes -11.7% (Table 2). This task was performed better by TD children -errors made up 8.3% and 5.3% in different age groups. A statistical analysis was performed to determine whether these quantitative differences could be attributed to language impairment (e.g., to suggest that the verb comprehension task can help identify language impairment). Its results showed that the difference between children of the same age with DLD and TD children is statistically significant: in the younger group (4;6-5;5) p = 0.006 (**), in the older group (5;6-6;5) p = 0.000 (***). Consequently, difficulties in comprehending verbs can be attributed to impaired language development.
Statistical analysis of the results of verb comprehension revealed interesting data related to the age factor. There was a significant difference between younger and older TD children p = 0.017 (*), however, age is not a significant factor in the group of children with DLD. It can be stated that TD children significantly improve their ability to comprehend verbs during the analyzed period. In the case of language disorders, on the other hand, no significant improvement in results was observed, therefore, it can be said that the comprehension skills of verbs change a little during this period.
Word production results are much lower than comprehension in all groups of children. As with comprehension, children had more difficulty completing the task with verbs. Children aged 4;6-5;5 with DLD named incorrectly or did not name 24.2% of nouns and 37.5% of verbs ( Table 2). The group of older children with DLD (5;6-6;5) made mistakes in 16.2% and 32.8% of cases, respectively. TD children aged 4;6-5;5 incorrectly named 14% of nouns and 26% of verbs. TD children aged 5;6-6;5 were less likely to make mistakes, however, even here incorrect naming accounted for 10% of nouns and 21.9% of verbs.
Although the word production results presented in Table 2 show differences between age groups, statistical analysis revealed that there was no statistically significant difference between younger (4;6-5;5) and older (5;6-6;5) children with DLD in the production of either nouns or verbs. The same tendency has been observed among TD children of different ages. It can be stated that none of these groups made significant progress during the period analyzed.
Quantitative analysis revealed significant differences related to language development: both when comparing younger children (4;6-5;5) with DLD and TD children, and when comparing the corresponding older groups (5;6-6;5); statistically significant differences were obtained (p = 0.004 (**), p = 0.006 (*), respectively. This indicates that children with DLD have significantly more difficulty in naming objects and actions.
Comparing the results of comprehending nouns and verbs, it was found that the comprehending of verbs in all groups of children was statistically significantly lower than that of nouns (in all groups the difference in age and language development was p = 0.000 (***)). Statistical analysis of the results of word production showed that there is a significant difference between the production of nouns and verbs in the group of 5;6-6;5 years old TD children (p = 0.024) (*). All the other groups found it similarly more difficult to name both objects and actions.
In summarizing the results of word comprehension and production, it should be mentioned that not all tasks of word comprehension and production are equally complex or easy for children. It was found that there are statistically significant differences between the understanding of some lexemes and the naming of some concepts. This needs to be taken into account in the development of language research tools and methodologies. The analysis of the individual lexemes used in the test allowed to single out words that are difficult for children with DLD to comprehend, but are easily understood by TD children (e.g., objects -envelope, rocket, actions -tears, stings). There is also a list of words, the difficulty in naming which can help identify a language disorder (e.g., objects -iron, turtle; actions -whistles, rides). Based on these results, the recommendations for the developers of the assessment tool 'Urtė ir Motiejus' were presented, and a more detailed analysis of individual lexemes could be the subject of further research.

Conclusions
The study revealed some features of the lexical development of children with DLD. Language disorder is characterized by the frequent use of pronouns and adverbs to replace unknown nouns, as well as by the frequency of onomatopoeia and the rare use of adjectives. Children with DLD make many semantic errors. Most of these errors involve the replacing of an unknown word with a pronoun or phrase that attempts to describe an object or action. In contrast to typical language development, it is often difficult for these children to choose a different lexeme for an unknown word based on the similarity of objects or other semantic word relationships. Thus, language disorder is characterized not only by a higher number of semantic errors, but also by a lack of compensation strategies: TD children usually replace an unknown word with another semantically related word, and children with DLD often do not find a similar word, replace it with a pronoun or adverb. Insufficient ability to compensate for lack of vocabulary limits the ability to communicate, does not allow to express thoughts and develop narratives.
The study revealed that children with DLD performed statistically significantly lower on both noun and verb comprehension tasks as well as noun and verb production tasks. Although TD children (especially the younger age groups) also have difficulties in word comprehension and production tasks, their errors are much less common and are usually associated with the infrequent use of some words in children's linguistic environment. Difficulties of children with DLD include frequently used words and concepts.
The obtained results showed that during the study period (4;6-6;5) children with DLD only improved their ability to comprehend nouns. Both the ability to comprehend verbs and the production of nouns and verbs do not change statistically significantly during this period. It can be argued that in the case of language disorders, difficulties in comprehension and production persist until the 6;5 years and later. To find out when children with language disorders overcome these difficulties, it is necessary to analyze the language abilities of older children.
A comparative analysis of the comprehension and production of nouns and verbs has shown that it is more difficult for children to comprehend verbs than nouns, and it is more difficult to name an action than an object. It has been found that not all words are equally complex or equally easy for children to understand and produce. Difficulties in comprehending and producing words are due to the rare use of the word in the child's environment, the length of the word, the structure of the word, the synonymy, and the abundance of words with similar meanings. the kindergarten 'Vėrinėlis' Laima Rickevičienė, speech language therapist of the kindergarten 'Žara' Alė V. Šoblinskienė for assistance in collecting the data of the experimental research, as well as for children and parents who agreed to participate in the research. We also thank to Dr. Ingrida Balčiūnienė for assistance in statistical data analysis. and to describe the peculiarities of comprehension of nouns and verbs; 4) to perform the analysis of the results of the experimental word production task and to describe the peculiarities of production of nouns and verbs. The results of children with language disorders (n = 80) and children with typical language development (n = 80) are compared.
Several methods were used in the study. The method of natural observation was used to collect data on children's spontaneous language (dialogues). Data were processed and analyzed using the method of corpus linguistics: dialogues of all children were transcribed, annotated and analysed using CHILDES program (MacWhinney, 2000). The study of word comprehension and production was conducted using an experimental research method. The children were tested in their educational institution, and the tasks of selecting and naming pictures were performed.
The study revealed some features of the lexical development of children with DLD. Language disorder is characterized by the frequent use of pronouns and adverbs to replace unknown nouns, as well as by the frequency of onomatopeia and the rare use of adjectives. Children with DLD make many semantic errors. Most of these errors involve the replacing of an unknown word with a pronoun or phrase that attempts to describe an object or action. In contrast to typical language development, it is often difficult for these children to choose a different lexeme for an unknown word based on similarity of objects or other semantic word relationships. Thus, language disorder is characterized not only by a higher number of semantic errors, but also by a lack of compensation strategies: TD children usually replace an unknown word with another semantically related word, and children with DLD often do not find a similar word, replace it with pronoun or adverb. Insufficient ability to compensate for lack of vocabulary limits the ability to communicate, does not allow to express thought and develop narrative.
The study revealed that children with DLD performed statistically significantly lower on both noun and verb comprehension tasks as well as noun and verb production tasks. Although TD children (especially the younger age groups) also have difficulties in words comprehension and production tasks, their errors are much less common and are usually associated with the infrequent use of some words in children's communicative environment. Difficulties of children with DLD include frequently used words and concepts.
The obtained results showed that during the study period (4;6-6;5) children with DLD only improved their ability to comprehend nouns. Both the ability to comprehend verbs and the production of nouns and verbs do not change statistically significantly during this period. It can be argued that in the case of language disorders, difficulties in comprehension and production persist until the 6;5 years and later. In order to find out when children with language disorders overcome these difficulties, it is necessary to analyze the language abilities of older children.