Expression and Perception of Endpoints during Language Acquisition: Three Studies on Czech
Abstract
7
Full text
Expression and Perception of Endpoints during Language Acquisition: Three Studies on Czech Anna Marklová — Barbara Mertins ABSTRACT: The description of goal-oriented motion events differs cross-linguistically. Speakers of languages with aphasal perspective (such as Russian) describe these events as aprocess, while speakers of languages with aholistic perspective (such as Czech) emphasize the final state of the event, i.e., an endpoint. For instance, when someone walks toward ahouse, Czech speakers are more likely to mention the house as an endpoint in their description than Russian speakers (Schmiedtová, 2013a). This paper investigates the prominence of endpoints during language acquisition of Czech preschoolers in three experiments. Firstly, we investigate the frequency of endpoint expression in Czech children (N=40) in an elicitation task compared to an interactive setting. Secondly, we compare the endpoint production of Czech children (n=25) with that of Russian children (n=25). Lastly, we examine whether visual fixations on endpoints increase in Czech children (n=75) with age. The results show that Czech children produced more endpoints in the interactive setting than in the elicitation task. We did not find asignificant difference between the overall number of endpoints produced by Russian and Czech children. However, there were significant differences in the types of utterances in which the children were expressing them. The eye-tracking data did not show agradual increase in attention to endpoints with age. Our findings suggest that parent–child interaction is crucial in directing children’s attention toward endpoints. However, Czech children appear to be considerably independent in their production of endpoints. This study provides insights into how Czech children describe and perceive endpoints, atopic that has not been previously analyzed. KEY WORDS: first language acquisition, language-specific patterns, holistic and phasal perspective, cross-linguistic comparison 1 INTRODUCTION The grammar of languages is selective, emphasizing some aspects of real-world situations more than others (Boroditsky, 2006). It has been argued that features that are grammaticalized and obligatory to express (like the category of aspect or grammatical gender in some languages) cause language speakers to pay closer attention to these features while using language or preparing language content (Sapir, 1921; Slobin, 1996). The current paper focuses on language-specific patterns in encoding and perceiving goal-oriented motion events. Agoal-oriented motion event is asituation where afigure moves toward an endpoint. Studies have shown that while describing agoal-oriented motion event, speakers of different languages focus on adifferent part of the scene (Carroll et al., 2011; OPEN ACCESS
8 STUDIE ZAPLIKOVANÉ LINGVISTIKY 2/2023 Mertins, 2018; Schmiedtová et al., 2011; Slobin, 1996; Talmy, 1985). While speakers of languages with the so-called ‘phasal’ perspective describe the scene as aprocess, speakers of the language with the so-called ‘holistic’ perspective emphasize the final state of the event in their description. For example, when someone is heading towards ahouse, ‘holistic’ speakers likely mention the house (endpoint) in their description, while ‘phasal’ speakers more probably leave it out. It was shown that the number of endpoints is consistently higher in descriptions from speakers of languages with aholistic perspective than from speakers of languages with aphasal perspective ( Sahonenko & Schmiedtová, 2008; Schmiedtová, 2008, 2013b; Schmiedtová et al., 2011; von Stutterheim et al., 2012). Such cross-linguistic differences are connected to the grammatical systems, particularly to the grammaticalized category of aspect. While aspectual languages tend to follow the phasal perspective, languages without aspectual categories conceptualize goal-oriented motion events holistically. Still, the conventional use of grammatical categories plays asignificant role. As was shown, an aspectual language usually adopts the phasal perspective; however, Czech is an outlier. While the perfective and the imperfective are grammaticalized (more detail below), their usage differs from other Slavic languages, and Czech speakers conceptualize the goal-oriented motion events holistically (Schmiedtová, 2013a). We assume that this preference is acquired during language acquisition. It requires asubstantial amount of input to follow the conventions and preferences of aparticular language, as studies have shown (e.g., Bowerman & Choi, 2001 on classification; Lucy & Gaskins, 2001 on categorization). The research on the acquisition of language-specific patterns has been inconsistent in determining when the language-specific patterns become prevalent in achild’s production (see Section 2). In our previous research (Marklová et al., 2023a), we focused on the input itself; we discovered that in child-directed speech, Czech speakers produce the highest number of endpoints compared to Russian speakers and Russian-German speakers. This implies that the direct parental input is already rich in preferences concerning endpoints. However, the results of the experiment did not tell us whether preschoolers already follow the preferences. The current paper addresses this issue. 1.1 GOAL-ORIENTED MOTION EVENTS IN CZECH AND RUSSIAN During the description of goal-oriented motion events, speakers follow the holistic or phasal perspective. Von Stutterheim and Nüse (2003) introduced these terms to delineate two contrasting approaches when discussing events: aspeaker can either focus on the goal, and therefore describe the event holistically, or focus on the process of the event, and therefore describe the event in phasal perspective. Each language possesses tools for describing events from both perspectives. Nevertheless, research has revealed that speakers of different languages tend to consistently favor one perspective over the other. This inclination has been attributed to differences embedded within language grammar. The pivotal factor influencing the selection of aspecific perspective has been identified as the grammaticalized aspect category (von Stutterheim & Nüse, 2003). This category marks the ongoingness or completion of an OPEN ACCESS
ANNA MARKLOVá — BARBARA MERTINS 9 event on the verb. Czech and Russian, which are the focus of the experiments in this paper, have formally similar aspectual systems: ongoingness is expressed by the simplex imperfective or secondary imperfective, and completion by the perfective form. Since the secondary imperfective form is incompatible with motion verbs, which are at the core of the motion events discussed in this paper, we will not elaborate on this form any further (for further information, see Schmiedtová et al., 2011). For an example of the aspectual categories in Czech (1) and Russian (2), see the following examples: (1) a. Zajíc běží do lesa. Hare run-IMPFV.3SG into the forest ‘Ahare is running into the forest.’ b. Zajíc poběží do lesa. Hare run-PFV.3SG into the forest ‘Ahare will run into the forest.’ (2) a. Zajac bežit vles. Hare run-IMPFV.3SG into the forest ‘Ahare is running into the forest.’ b. Zajac pobežit vles. Hare run-PFV.3SG into the forest ‘Ahare will run into the forest.’ In (1a.) and (2a.), the simplex imperfective marks an ongoing event in the present tense. There are some differences in the usage of this form in Czech and Russian: while it has been shown that it is common in Russian to express such an event in abare-verb phrase (i.e., Zajac bežit. ‘Ahare is running.’), such bare verb phrases tend to be supplemented by additional arguments, such as the information about the path (i.e., Zajíc běží po poli. ‘Ahare is running on through the field.’) or endpoint (Zajíc běží do lesa. ‘Ahare is running to the forest.’) (Schmiedtová et al., 2011). The expression of completion is presented in example (1b.) for Czech and (2b.) for Russian. The perfective form of motion verbs is created by adding aprefix to the simplex imperfective. This leads to achange in the tense since it shifts the verb’s meaning into the future. However, it has been observed that Czech speakers use the perfective form also when describing an ongoing event. Therefore, the perfective, while formally expressing future tense, can also be used in the so-called here-and-now reading (Schmiedtová, 2008).1 This shift in the usage of the aspectual forms was explained by Czech’s longterm contact with German, anon-aspectual language with atendency to emphasize endpoints (Mertins, 2018). These specifics of the Czech aspect led to its belonging to the languages with aholistic perspective, as will be discussed further. Overall, speakers of non-aspectual languages show the holistic perspective, highlighting the endpoint while encoding goal-oriented motion events. Speakers of aspectual languages show the phasal perspective, which emphasizes the processual 1 Schmiedtová (2008) identified this phenomenon in the elicitation data of Czech native speakers. OPEN ACCESS
10 STUDIE ZAPLIKOVANÉ LINGVISTIKY 2/2023 characteristic of the event. Carroll and von Stutterheim (2006) explain why the absence of the aspectual category is linked to the expression of endpoints: speakers of aspectual languages connect the time of speaking with the time of the topic through aspectual markers, enabling them to directly establish ongoing events. That eliminates the requirement for one event to be portrayed as finished or limited before introducing another. Conversely, speakers without the aspectual category need to explicitly indicate the completion points or results of previous events to connect the topic time to the preceding time of the situation. Czech has been employed in the cross-linguistic comparisons of languages with different perspectives and has been consistently categorized as alanguage with the holistic perspective. This means that the expression of endpoints by Czech speakers was significantly more frequent compared to ‘phasal languages’ such as English, Spanish, Dutch, Norwegian, or Russian (Carroll et al., 2011; Mertins, 2018; Schmiedtová et al., 2011). On the other hand, it was comparable to languages such as German or Dutch, which were identified as languages with the holistic perspective ( Schmiedtová, 2013a). Russian and Czech were directly compared in Schmiedtová (2013a). Speech data from Czech, Russian, and German adult speakers (N=83) who described video clips depicting goal-oriented motion events were examined. The results showed that Russian speakers expressed significantly fewer endpoints (22 in total) compared to Czech (52 in total) and German speakers (44 in total). The comparison between Russian and Czech speakers yielded asignificant difference. Additionally, Russian speakers used bare-verb phrases significantly more often than Czech and German speakers (20 times compared to 4 and 5 times, respectively). Furthermore, it was observed that these perspectives also emerged outside of verbalization; the endpoints were emphasized by ‘holistic’ speakers also in memory tasks following verbalization and in visual attention preceding verbalization. It is worth mentioning that the prominence of the perspective is strong enough to also be transferred into L2, as studies with highly proficient L2 speakers uncovered (Carroll & Lambert, 2003; Carroll & von Stutterheim, 2006; Sahonenko & Schmiedtová, 2008; Schmiedtová et al., 2011; van Ierland, 2009; von Stutterheim, 2003; von Stutterheim & Lambert, 2005). In astudy conducted by Schmiedtová and Sahonenko (2008), the researchers examined how Russian and Czech speakers of L2 German verbalized goal-oriented motion events. The results showed that both groups used the encoding patterns of their first language when conceptualizing goal-oriented motion events in the target language. 1.2 ACQUISITION OF LANGUAGE-SPECIFIC PATTERNS So far, there has been alack of studies focusing on the acquisition of these perspectives (for alarge cross-linguistic comparative work on event descriptions in language development, see Berman & Slobin, 1994). We follow an approach towards language acquisition which argues that it is input-driven (see Clark, 2016 for an overview). Language is seen as aproduct of social interaction, in which children receive cues about the conventional use of particular words, expressions, and constructions. OPEN ACCESS
ANNA MARKLOVá — BARBARA MERTINS 11 Achild can only successfully acquire language with sufficient input (Arnon et al., 2014; Clark, 2016). Moreover, the language-specific patterns and idiosyncratic patterns found in caregivers’ language use are reflected in their children’s language use (Bowerman & Choi, 2001, 2003; Gopnik, 2001). Studies have shown that children’s speech is more similar to their parents’ speech than to other children’s (Bowerman & Choi, 2003). However, from what age do children follow the language-specific patterns, such as the expression of endpoints? We ask this question in our three experiments focusing on the acquisition of holistic and phasal perspectives since the research on other language-specific preferences shows mixed results. Some studies recognized an influence even on the production of first words (Bowerman, 1996; Choi & Bowerman, 1991; Hickmann et al., 2009; Slobin et al., 2010), while others posited the grasp of language-specific conceptualization in the period of three to four years of age (Allen et al., 2007). For example, Choi (2006) identified that language-specific categorization of containment in Korean plays arole from an early age. Lucy’s research on classifiers in Yucatec tracked that they are acquired gradually until approximately nine years of age (Lucy, 2004; Lucy & Gaskins, 2001). Several studies have investigated the acquisition of conceptual perspectives for locomotion events, with aparticular focus on the difference between satellite-framed and verb-framed languages and the emergence of differences in the description of such events. This research follows Talmy’s (1985) lexicalization typology. Harr (2012) provides an overview of this research. It was uncovered that children are able to encode different elements of motion, i.e., path and manner, since the first language production (Mandler, 2007; Pulverman et al., 2008) and around three years of age, emphasizing the elements of the motion events according to the lexicalization patterns of their language (Choi & Bowerman, 1991; Oh, 2003; Ozcaliskan & Slobin, 1999; Özyürek et al., 2008; Papafragou et al., 2002). Hickmann and Hendriks (2010) observed various language-specific patterns during the description of motion events in children as young as two and ahalf years. 2 PRESENT STUDY The current paper focuses on children’s speech and visual attention regarding endpoints. Firstly, we analyze data from Czech children (aged 3–5) while describing picture representations of goal-oriented motion events in anon-interactive elicitation task. We compare them to Czech children who discuss the same stimuli in an interactive setting. We hypothesize that the interactive setting will affect children’s production of endpoints since it encourages and leads their attention to them. We base this assumption on our previous research (Marklová et al., 2023a) and the rich literature concerning the role of interaction in language acquisition. Secondly, we compare the expression of endpoints in the Czech children from the ‘interactive’ group from the first experiment with data from Russian children recorded in the same setting and over the same stimuli. We expect the number of endpoints produced by Czech children to be higher than that produced by Russian children. We aim to conduct athorough analysis of the child–parent conversations. Lastly, we OPEN ACCESS
12 STUDIE ZAPLIKOVANÉ LINGVISTIKY 2/2023 recorded the eye movements of Czech children who silently watched the same stimuli as in the first two experiments on the computer screen. We analyze if the fixation on endpoints increases with age. Online research methods such as eye-tracking can uncover unconscious processes that are not apparent through language production alone. That is particularly relevant for young children, as their language production abilities may not reflect their overall language abilities. By measuring eye movements, it is possible to gain valuable insights into whether language perspectives influence children’s visual attention. This method has been employed in previous research on phasal and holistic perspectives and uncovered differences between adult speakers’ visual perception (von Stutterheim et al., 2012). It has not been used with preschoolers yet. We are interested in the following questions: Do Czech children follow the holistic preference while encoding motion events on their own, or do they need encouragement from their parents to do so? Do Czech children express more endpoints than Russian children when describing motion events in an interactive setting? Does the visual fixation of endpoints in Czech children increase with age? This paper describes the main findings observed in how Czech and Russian children discuss and perceive endpoints. 2.1 EXPERIMENT 1: CZECH CHILDREN’S PRODUCTION WITH AND WITHOUT INTERACTION In the first experiment, we compared the expression of endpoints in the speech of Czech children in two settings. We presented picture stimuli depicting goal-oriented motion events in an elicitation task and active interaction with acaregiver. We were interested if the total amount of endpoints expressed by the children would be higher in the interactive setting since our previous study revealed ahigh number of endpoints in the child-directed speech of Czech parents and the tendency of parents to lead attention to them actively (that is, expressing them in the form of leading questions such as Kam běží ten pes? ‘Where is the dog running to?’). 2.1.1 PARTICIPANTS Our analysis includes data from 45 children divided into two groups: an ‘elicitation group’ (n=20) and an ‘interactive group’ (n=25). The data for the elicitation group was collected between August 2021 and November 2022, while the data for the interactive group was obtained from apublicly available dataset (Marklová et al., 2023b). The parents of children from the elicitation group were paid 200 Czech crowns (approximately 8 euros) and they were recruited individually from the social network of the authors or by flyers in kindergartens. The parents of the children in the interactive group were recruited from the pool of participants in experiments at the Institute of Psychology, Academy of Science in Prague. Their participation was voluntary and without financial reward. The demographic information is displayed in Table 1. OPEN ACCESS
ANNA MARKLOVá — BARBARA MERTINS 13 Elicitation group Interactive group gender 13 female, 7 male 15 female, 10 male age Ø 4;5 (range 2;9–5;11) Ø 3;9 (range 2;6–5;8) Table 1:Participants’ overview. Figure 1:Examples of the stimuli. 2.1.2 DESIGN AND MATERIALS Ten pictures were used as stimuli in the elicitation task; six depicted goal-oriented motion events where an animal moves towards an endpoint of various prominence (see Figure 1). Four items depicted an animal/animals in astatic position, and they served as fillers. The fillers were added to ensure that the children would not iterate the same sentence structure describing movement, which happened during the piloting phase (some children started to use the same verb for all animals, such as koník běží ‘the horse is running’, čáp běží ‘the stork is running’, etc.). The conversations over the fillers were not analyzed. The items were presented in linear randomized order. The collection was conducted by the parents and children in their natural environment (at home). The instructions and the stimuli were presented online via the LimeSurvey tool (LimeSurvey Project team, 2012). Parents were instructed to run the experiment with their children while they recorded them on avoice recorder. The experiment was self-paced. Each item was presented on aslide, and parents had to click on abutton to skip to another picture. Their task was to ask the child Co se děje na obrázku? ‘What is happening in the picture?’ and let the child describe the picture without interventions. They were explicitly asked not to elaborate the child’s answers, but they could encourage them with vague questions such as Adál? ‘And then?’ or Aco se ještě děje? ‘And what else is happening?’. There were individual situations when the parents did not follow the instructions completely (for example, when the child was very silent, the parents asked afew more targeted questions). Since these Filler Critical OPEN ACCESS
14 STUDIE ZAPLIKOVANÉ LINGVISTIKY 2/2023 situations were not frequent and in the majority of situations, the questions did not target the motion event (but for example, the properties of the animal), we did not exclude these transcripts from the analysis. The interactive group data was taken from the larger dataset (Marklová et al., 2023b). The data was recorded, as in the elicitation task, in the natural environment by caregivers of the children. Each parent obtained 12 pictures depicting goal-oriented motion, which were part of alarger battery of 30 pictures.2 The six critical stimuli in the elicitation task were from the same set. The items were printed in color on individual sheets of paper of A5 size. There was no prescript order of viewing that the parents had to follow. The pictures were rearranged in arandom order for each parent–child pair. The parents were instructed to talk about the pictures with the children ‘as naturally as possible’, i.e., in an interactive way. However, it was highlighted that the goal is not to ‘show off’ the child’s skills but rather to record anatural conversation between children and their parents. Parents were instructed to start with the question Co se děje na obrázku? ‘What is happening in the picture?’ and develop aconversation about the events depicted in the pictures. 2.1.3 TRANSCRIPTION, SEGMENTATION AND CODING CRITERIA The audio data was transcribed at full length. We partitioned every transcript into discrete segments, which we refer to as “conversations”. Each conversation targets one stimulus. Following this initial segmentation, we further divided these segments into individual utterances, with afocus on identifying those referring to locomotion events. Amotion verb and optional verbal adjuncts determined these utterances. We followed the basic definition of motion verbs from Miller and Johnson-Laird (1976, p.527): “verbs that describe how people and things change their places and their orientations in space”. For specific or unclear instances, we consulted the literature focused on Czech motion verbs and their typology (Daneš & Hlavsa, 1981; Saicová Římalová, 2010). We considered the ‘simple’ usage of motion verbs in any tense (although the present tense was prevalent), and we also counted situations with amodal verb used with amotion verb, such as koník chce běžet do stáje ‘the horse wants to run to the stable’. Most of the verbs produced by children were the basic verbs of motion; more than 70% of the analyzed verbs consisted of the lemmas fly, run, go/walk, and jump in the indicative present tense. See the following transcript of asegment from aconversation: (3) A: hm. tak aco se tady děje? em. so and what is happening here? CH: čáp aten nevim co dělá stork and Idon’t know what he is doing A: tak vidíš co dělá ne? you see what he is doing don’t you? CH: jo! letí na komín yes! (he is) flying onto the chimney! The utterance which we identified as amotion event is underlined. We coded the endpoints when they occurred in the utterance (as in the transcript above). Endpoints 2 The set included aleft-directional and right-directional variant of 15 original pictures. OPEN ACCESS
ANNA MARKLOVá — BARBARA MERTINS 15 are typically expressed by prepositional phrases (PPs). The typical prepositions in Russian are к ‘to’, в ‘in’, на ‘to/on’, по направлению к ‘towards’, in Czech do ‘into’, k‘to’, na ‘to/on’: (4) Sobaka bežit vsvoju budku. ‘The dog is running into his dog house.’ Ten pes běží do svojí boudy. ‘The dog is running into his dog house.’ Questions targeting endpoints do so by the interrogative pronoun куда ‘where to’ in Russian, and kam ‘where to’ in Czech: (5) Ikuda skačet lošad’? ‘And where is the horse running to?’ Akam ten kůň běží? ‘And where is the horse running to?’ Each utterance containing an endpoint was also coded according to its type. This examination criterion comprised four categories: statement (example 4), answer, question (example 5) and repetition. Only answers to wh-questions directly targeting endpoints were categorized as ‘answers’, as in the following segment: (6) A: líbí se mu to? no to možná by neutíkal. kam asi utíká? podívej he likes it? he would maybe not run then. where is he running? look CH: do boudy. into the dog house An example of repetition is displayed in the following: (7) A: aha. takže koník běží domů isee. so the horse is running home CH: jo, domů yes, home These types of utterances were chosen because question–answer sequences are among the most prominent conversation sequences (Sacks et al., 1974), and repetitions and statements are important components of parent–child interaction (Sinclair & Fernández, 2021). The analysis of the type of utterance enabled us to analyze the specific features of child–parent communication in connection with how endpoints are expressed. The conversations were transcribed by the authors of the publicly available dataset. To ensure the accuracy of the coding process, sixteen transcripts consisting of 717 utterances were independently coded by two proficient coders. Both coders possessed astrong command of the languages under analysis. Subsequently, intercoder reliability was assessed using Cohen’s kappa index. The calculation involved dividing the number of actual matches by the total number of utterances. The resulting Cohen’s kappa index value between the two coders was found to be 0.97. According to Landis and Koch’s (1977) benchmarks for evaluating the strength OPEN ACCESS
22 STUDIE ZAPLIKOVANÉ LINGVISTIKY 2/2023 2.3.3 RESULTS AND DISCUSSION The endpoint areas were predetermined for each stimulus prior to the recordings. These regions conformed to the shapes of the endpoints, with consistent overlapping along the sides. The areas were not marked for the participants during the stimuli presentation. Regression models were used to test the influence of age on fixation time and entry time in agroup of participants. The analysis was performed using RStatistical Software 4.1.3 (R Core Team, 2022). The results showed no significant influence of age on fixation time (p = 0.5917) or entry time (p = 0.3083) with alevel of significance of p = 0.05. The details can be seen in Table 2 for the fixation time and Table 3 for the entry time. Estimate Std. Error t value Pr(>|t|) (Intercept) 294,93 92,56 3,186 0.00182 ** Age –10,35 19,25 –0,538 0,59166 Table 2:Regression analysis results for Fixation Time (ms) on Age. Residual standard error: 192.8 on 126 degrees of freedom Multiple R-squared: 0.00229, Adjusted R-squared: -0.005628 F-statistic: 0.2892 on 1 and 126 DF, p-value: 0.5917 Estimate Std. Error t value Pr(>|t|) (Intercept) 1992 726,3 2,743 0.00698 ** Age 154,5 151,1 1,023 0,30834 Table 3:Regression analysis results for Entry Time (ms) on Age. Residual standard error: 1513 on 126 degrees of freedom Multiple R-squared: 0.008235, Adjusted R-squared: 0.0003638 F-statistic: 1.046 on 1 and 126 DF, p-value: 0.3083 No correlation was found between the age of the children and the attention spent on the endpoint area. Therefore, we did not observe aprogression of this tendency with increasing age. As depicted in Figures 4 and 5, there is asubstantial degree of inter-individual variability, and no discernible pattern of adherence can be identified across the various age groups. There are several possible interpretations of our findings. We used the analysis of visual perception without the prompt for verbalization. There is mixed evidence about language influence in such asetting. It is possible that in this setting, the influence of the perspectives is not strong enough to be reflected in children’s visual attention. Additionally, conducting eye-tracking experiments with children has several limitations. The high variability in eye movements might not allow us to see ageneral pattern. There is also apossibility that the visual attention into the area of the endpoint simply does not increase any more after approximately 2;6 years of age. This could either suggest that the influence of preferences is already established or that OPEN ACCESS
ANNA MARKLOVá — BARBARA MERTINS 23 the influence on visual perception would be visible later in development. Overall, these potential explanations highlight the complex nature of visual perception in young children and suggest the need for further research to fully clarify the underlying mechanisms involved. Figure 4:Entry time into the endpoint area. Each point represents one participant. The x-axis depicts the age of the children in years, and the y-axis the entry time in ms. Figure 5:Fixation time in the endpoint area. Each point represents one participant. The x-axis depicts the age of the children in years, and the y-axis the fixation time in ms. Entry TimeFixation Time 0 1000 3000 5000 Age 3 4 5 6 7 200 400 600 800 1200 Age 3 4 5 6 7 OPEN ACCESS
24 STUDIE ZAPLIKOVANÉ LINGVISTIKY 2/2023 3. GENERAL DISCUSSION In this paper, we presented three experiments with acommon focus on the prominence of endpoints in Czech preschoolers. Firstly, we compared the speech production of Czech children during an elicitation task and active interaction over stimuli depicting goal-oriented motion events. We found asignificant difference between these two groups: The active interaction led to the expression of more endpoints than the elicitation. When analyzing the types of utterances in the interactive group, we discovered that in 30 out of 81 cases, children expressed the endpoint in adirect answer to aparent’s wh-question targeting it. We hypothesized that the interactive setting is essential in leading children’s attention to the endpoints. To assess the prominence of the holistic perspective in Czech children in general, we conducted an experiment comparing the Czech children with agroup of Russian children. We discovered that while the expression of endpoints was comparable in both groups, Russian children expressed the endpoint in answers much more often than Czech children. This might indicate that while the number of endpoints in the elicitation of Czech children is lower than when they talk interactively with their parents, they are still very independent in their production. That would align with studies showing the language-specific preferences in 3-year-old children describing motion events (e.g., Choi & Bowerman, 1991; Harr, 2012; Hickmann & Hendriks, 2010; Oh, 2003). We assume that in Czech preschoolers, the preference for expressing endpoints already plays arole in spontaneous speech. This needs to be supported by elicitation data from Russian children. The ambition to collect comparative data from Russian children has been interrupted by external circumstances, which led to our inability to collect data in Russia as was initially planned. In our third experiment, we offered afirst insight into the eye movements of Czech children while they observe pictures depicting goal-oriented motion events. We were interested if the fixation on endpoints increases with the age of the children. In the free-viewing task, we did not find such apattern. Our data showed significant variability in children’s eye-gaze, but it did not follow ageneral pattern. While there is evidence that the effect of the holistic perspective also influences other processes than verbalization, such as memory tasks and visual fixation (von Stutterheim et al., 2012), there is only scattered evidence of alanguage-specific effect on processes that do not precede or follow language use. For example, Lucy and Gaskins (2001) observed that language-specific patterns affected the grouping of various objects. While speakers of Yucatec preferred grouping based on material, English speakers preferred grouping based on shape. This follows the Yucatec requirement of unitizers for nouns based on materials. McDonough et al. (2000) found differences in the fixation time of Korean and English speakers in observing pictures depicting categories of containment. For Korean speakers, it was easier to recognize tight-fit and loose-fit distinctions in the presented pictures because this distinction is grammaticalized in their language. Levinson’s (1996) “turn-table” experiments on spatial relations uncovered that speakers of languages with an absolute frame of reference, such as Tzeltal, adhere to this principle even in non-linguistic tasks, such as in arranging an array of objects. The rare studies on visual perception of goal-oriented motion events with OPEN ACCESS
ANNA MARKLOVá — BARBARA MERTINS 25 no verbalization following the stimuli presentation brought mixed evidence: while Flecken et al. (2014) claimed an effect on visual fixation, Marklová et al. (2023) did not find any. Additionally, Athanasopoulos et al. (2015) brought evidence of the effect of aspectual systems in similarity judgment tasks. We chose the method of presenting the pictures depicting goal-oriented motion events without aprompt for description because we expected that with verbalization, the results would likely be similar to the results of the elicitation task since the children would follow the desired verbalization (see Carroll et al., 2011 for the seeing for speaking hypothesis). However, the methodology we used might not be able to uncover the unconscious processes in full scope. For further research, we aim to conduct acomparative eye-tracking study to see if differences appear when comparing Czech and Russian children. There are several implications we can draw from our results. Firstly, while we did not find asignificant difference between the Russian and Czech children while comparing the total number of endpoints, an asymmetry in the statements about the endpoint in the two children’s groups was found. The Czech children were more independent in their expression of endpoints than Russian children. This might suggest that the holistic and phasal perspectives play arole in children’s production from early on. In the context of research on the acquisition of various language-specific features, we assume that, similarly to the acquisition of lexicalization patterns ( Ozcaliskan & Slobin, 1999; Papafragou et al., 2002), children follow the perspective of their language from early language production. Additional evidence is required to support this assumption, such as comparing Czech children in an elicitation setting with Russian children. Secondly, the interaction between achild and acaregiver drives how endpoints are expressed in children’s speech. It affects the number of expressed endpoints, as was shown in Experiment 1, and the type of utterance in which the endpoints are expressed (see Experiment 2). The discrepancy in the type of utterances in Czech and Russian children suggests that the interaction around the endpoints was profoundly different in these two groups. Follow-up studies focusing on the expression of endpoints in children acquiring phasal and holistic languages, both in interaction and in independent elicitation, are needed to understand which patterns are language-specific and which are rooted in cultural or social conventions. Thirdly, we offered an insight into the role of holistic preference in the visual perception of Czech preschoolers. We wanted to see the development of fixation on the endpoint in the eye movements, but we did not find evidence for that. Studies with adult participants uncovered that the preference for the holistic vs. phasal perspective in native speakers of Czech and Russian is systematic and remains in highly proficient speakers of L2 (von Stutterheim, 2003). Therefore, it is deeply anchored in the habits of the speakers. This might also explain the insignificant results of the eye-tracking Experiment 3, since the preference for eye movements might already be strong as early as at three years of age. However, the experiment has too many limitations to make such aconclusion, namely the large intra-individual variability. Based on our results, we conclude that there is no evidence that the visual attention on endpoints increases with age. In conclusion, the present study sheds light on the role of language-specific perspective in shaping Czech-speaking children’s speech production and visual OPEN ACCESS
26 STUDIE ZAPLIKOVANÉ LINGVISTIKY 2/2023 perception. However, to gain amore comprehensive understanding of this phenomenon, follow-up studies comparing Czech children with those speaking alanguage with aphasal perspective are crucial. The present paper is the first step towards exploring the effects of holistic and phasal perspectives on children’s speech production and visual perception. REFERENCES: Allen, S., Özyürek, A., Kita, S., Brown, A., Furman, R., Ishizuka, T., & Fujii, M. (2007). Language-specific and universal influences in children’s syntactic packaging of Manner and Path: Acomparison of English, Japanese, and Turkish. Cognition, 102(1), 16–48. https://doi. org/10.1016/j.cognition.2005.12.006 Arnon, I., Casillas, M., Kurumada, C., Estigarribia, B., & Clark, E.V. (Eds.). (2014). Language in Interaction: Studies in Honor of Eve V.Clark. John Benjamins. Athanasopoulos, P., Bylund, E., Montero-Melis, G., Damjanovic, L., Schartner, A., Kibbe, A., Riches, N., & Thierry, G. (2015). Two languages, two minds: Flexible cognitive processing driven by language of operation. Psychological Science, 26(4), 518–526. https:// doi.org/10.1177/0956797614567509 Berman, R.A., & Slobin, D.I. (Eds.). (1994). Relating Events in Narrative. L.Erlbaum Associates. Boroditsky, L. (2006). Linguistic relativity. In L.Nadel (Ed.), Encyclopedia of Cognitive Science (pp. 917–921). John Wiley & Sons, Ltd. https:// doi.org/10.1002/0470018860.s00567 Bowerman, M. (1996). Learning how to structure space for language: Acrosslinguistic perspective. In P.Bloom, M.F.Garret, L.Nadel & M.A.Peterson (Eds.), Language and Space (pp. 385–436). The MIT Press. https://doi. org/10.1146/annurev.anthro.25.1.353 Bowerman, M., & Choi, S. (2001). Shaping meanings for language: Universal and language-specific in the acquisition of semantic categories. In M.Bowerman & S.Levinson (Eds.), Language Acquisition and Conceptual Development (pp. 475–511). Cambridge University Press. https://doi. org/10.1017/CBO9780511620669.018 Bowerman, M., & Choi, S. (2003). Space under construction: Language-specific spatial categorization in first language acquisition. In D.Gentner & S.Goldin-Meadow (Eds.), Language in Mind: Advances in the Study of Language and Thought (pp. 387–427). Boston Review. https://doi.org/10.7551/ mitpress/4117.001.0001 Carroll, M., & Lambert, M. (2003). Information structure in narratives and the role of grammaticised knowledge: Astudy of adult French and German learners of English. In C.Dimroth & M.Starren (Eds.), Information Structure and the Dynamics of Language Acquisition (pp. 267–287). John Benjamins. Available at https://www.jbe-platform.com/ content/books/9789027296436-sibil.26.13car Carroll, M., & von Stutterheim, C. (2006). The impact of grammaticalised temporal categories on ultimate attainment in advanced L2-acquisition. In H.Byrnes (Ed.), Educating for Advanced Foreign Language Capacities. Constructs, Curriculum, Instruction, Assessment (pp. 40–53). Georgetown University Press. Carroll, M., von Stutterheim, C., & Nüse, R. (2011). The language and thought debate: Apsycholinguistic approach. In T.Pechmann & C.Habel (Eds.), Multidisciplinary Approaches to Language Production (pp. 183–218). De Gruyter Mouton. https://doi. org/10.1515/9783110894028.183 Choi, S. (2006). Influence of language-specific input on spatial cognition: Categories of containment. First Language, 26(2), 207–232. https://doi.org/10.1177/0142723706060748 Choi, S., & Bowerman, M. (1991). Learning to express motion events in English and Korean: The influence of language-specific lexicalization patterns. Cognition, 41(1–3), OPEN ACCESS
ANNA MARKLOVá — BARBARA MERTINS 27 83–121. https://doi.org/10.1016/00100277(91)90033-Z Clark, E.V. (2016). First Language Acquisition (3rd ed.). Cambridge University Press. https:// doi.org/10.1017/CBO9781316534175 Daneš, F., & Hlavsa, Z. (1981). Větné vzorce včeštině. Academia. Flecken, M., von Stutterheim, C., & Carroll, M. (2014). Grammatical aspect influences motion event perception: Findings from acrosslinguistic non-verbal recognition task. Language and Cognition, 6(1), 45–78. https:// doi.org/10.1017/langcog.2013.2 Gopnik, A. (2001). Theories, language, and culture: Whorf without wincing. In M.Bowerman & S.Levinson (Eds.), Language Acquisition and Conceptual Development (1st ed., pp.45–69). Cambridge University Press. https://doi.org/10.1017/ CBO9780511620669.004 Harr, A.-K. (2012). Language-Specific Factors in First Language Acquisition: The Expression of Motion Events in French and German. De Gruyter Mouton. https://doi. org/10.1515/9781614511748 Hickmann, M., & Hendriks, H. (2010). Typological constraints on the acquisition of spatial language in French and English. Cognitive Linguistics, 21(2), 189–215. https://doi. org/10.1515/COGL.2010.007 Hickmann, M., Taranne, P., & Bonnet, P. (2009). Motion in first language acquisition: Manner and Path in French and English child language. Journal of Child Language, 36(4), 705–741. https://doi.org/10.1017/ S0305000908009215 Holmqvist, K., & Andersson, R. (2017). Eye Tracking: AComprehensive Guide to Methods, Paradigms, and Measures (2nd ed.). Lund EyeTracking Research Institute. Landis, J.R., & Koch, G.G. (1977). The measurement of observer agreement for categorical data. Biometrics, 33(1), 159–74. Levinson, S.C. (1996). Frames of reference and Molyneux’s question: Crosslinguistic evidence. In P.Bloom, M.F.Garret, L.Nadel & M.A.Peterson (Eds.), Language and Space (pp. 109–169). The MIT Press. https://doi. org/10.1146/annurev.anthro.25.1.353 LimeSurvey Project Team. (2012). LimeSurvey: An Open Source survey tool. LimeSurvey Project. Available at http://www.limesurvey. org Lucy, J.A. (2004). Language, culture, and mind in comparative perspective. In M.Achard & S.Kemmer (Eds.), Language, Culture and Mind (pp. 1–22). CSLI Publications. Lucy, J.A., & Gaskins, S. (2001). Grammatical categories and the development of classification preferences: Acomparative approach. In M.Bowerman & S.Levinson (Eds.), Language Acquisition and Conceptual Development (pp. 257–283). Cambridge University Press. https://doi.org/10.1017/ CBO9780511620669.011 Mandler, J.M. (2007). Foundations of Mind. Oxford University Press. https://doi.org/10.1093/acprof:o so/9780195311839.001.0001 Marklová, A., Panfilova, E., & Mertins, A. (2023a). Description of Locomotion Events: Language-specific Patterns in Child-directed Speech. WORD. https://doi.org/10.1080/0043 7956.2023.2269702 Marklová, A., Panfilova, E., & Mertins, A. (2023b). Parent-child Conversations about Motion Events (Russian, Russian-German, Czech). DataverseNO, V1. https://doi.org/10.18710/ BJGGZU McDonough, L., Choi, S., & Mandler, J.M. (2000). Development of language-specific categorization of spatial relations from prelinguistic to linguistic stage: Apreliminary study. [Conference presentation]. Finding the Words, Stanford, CA. Mertins, B. (2018). Sprache und Kognition: Ereigniskonzeptualisierung im Deutschen und Tschechischen. De Gruyter. https://doi. org/10.1515/9783110615746 Miller, G.A., & Johnson-Laird, P.N. (1976). Language and Perception. Belknap Press of Harvard University Press. Oh, K.J. (2003). Manner and Path in motion event descriptions in English and Korean. In B.Beachley, A.Brown & F.Conlin (Eds.), OPEN ACCESS
28 STUDIE ZAPLIKOVANÉ LINGVISTIKY 2/2023 Proceedings of the 27th Annual Boston University Conference on Language Development I-II (pp. 580–590). Cascadilla Press. Ozcaliskan, S., & Slobin, D. (1999). Learning how to search for the frog: Expression of manner of motion in English, Spanish, and Turkish. In A.Greenhill, H.Littlefield & C.Tano (Eds.), Proceedings of the Annual Boston University Conference on Language Development (Vol. 23, pp.541–552). Cascadilla Press. Özyürek, A., Kita, S., Allen, S., Brown, A., Furman, R., & Ishizuka, T. (2008). Development of cross-linguistic variation in speech and gesture: Motion events in English and Turkish. Developmental Psychology, 44(4), 1040–1054. https://doi.org/10.1037/00121649.44.4.1040 Papafragou, A., Massey, C., & Gleitman, L. (2002). Shake, rattle, ‘n’ roll: The representation of motion in language and cognition. Cognition, 84(2), 189–219. https:// doi.org/10.1016/S0010-0277(02)00046-X Pulverman, R., Golinkoff, R.M., Hirsh-Pasek, K., & Buresh, J.S. (2008). Infants discriminate manners and paths in non-linguistic dynamic events. Cognition, 108(3), 825–830. https://doi. org/10.1016/j.cognition.2008.04.009 R Core Team. (2022). R: Alanguage and environment for statistical computing. R Foundation for Statistical Computing. Available at https://www.R-project.org/ Riege, A., Gourdon-Kanhukamwe, A., & ValléeTourangeau, G. (2021) Covert eye-tracking: an innovative method to investigate compliance with instructions. Psychological Research, 85, 3084–3093. https://doi.org/10.1007/s00426020-01451-9 Sacks, H., Schegloff, E., & Jefferson, G. (1974). Asimple systematic for the organisation of turn taking in conversation. Language, 50, 696–735. https://doi.org/10.2307/412243 Sahonenko, N., & Schmiedtová, B. (2008). Die Rolle des grammatischen Aspekts in der Ereignis-Enkodierung: Ein Vergleich zwischen tschechischen und russischen Lernern des Deutschen. In M.Walter & P.Grommes (Eds.), Fortgeschrittene Lernervarietäten: Korpuslinguistik und Zweitsprachenerwerbsforschung (pp. 45–72). Max Niemeyer Verlag. https://doi. org/10.1515/9783484970342.1.45 Saicová Římalová, L. (2010). Vybraná slovesa pohybu včeštině: studie zkognitivní lingvistiky. Karolinum. Sapir, E. (1921). Language. Harcourt, Brace & World. Schmiedtová, B. (2008). At the Same Time… The Expression of Simultaneity in Learner Varieties. PhD Thesis, Radboud University Nijmegen, Nijmegen. https://doi. org/10.17617/2.59569 Schmiedtová, B. (2013a). Zum Einfluss des Deutschen auf das Tschechische: Die Effekte des Zeitdrucks auf die Sprachproduktion. In M.Nekula, K.Šichová & J.Valdrová (Eds.), Bilingualer Sprachvergleich und Typologie: Deutsch—Tschechisch (pp. 177–206). Julius Groos. Schmiedtová, B. (2013b). Traces of L1 patterns in the event construal of Czech advanced speakers of L2 English and L2 German. International Review of Applied Linguistics in Language Teaching, 51(2), 87–116. https://doi. org/10.1515/iral-2013-0005 Schmiedtová, B., von Stutterheim, C., & Carroll, M. (2011). Language-specific patterns in event construal of advanced second language speakers. In A.Pavlenko (Ed.), Thinking and Speaking in Two Languages (pp. 66–107). Multilingual Matters. https://doi. org/10.21832/9781847693389-005 Sinclair, A., J., & Fernández, R. (2021). Construction coordination in first and second language acquisition. In E.Breitholtz, K.Georgila & D.Schlangen (Eds.), Proceedings of the 25th Workshop on the Semantics and Pragmatics of Dialogue, Potsdam. University of Potsdam. Slobin, D.I. (1996). From “thought and language” to “thinking for speaking.” In J.Gumperz & S.Levinson (Eds.), Rethinking Linguistic Relativity (pp. 70–96). Cambridge University Press. https://philarchive.org/rec/SLOFTA Slobin, D.I., Bowerman, M., Brown, P., Eisenbeiß, S., & Narasimhan, B. (2010). Putting things in places: Developmental OPEN ACCESS
ANNA MARKLOVá — BARBARA MERTINS 29 consequences of linguistic typology. In J.Bohnemeyer & E.Pederson (Eds.), Event Representation in Language and Cognition (1st ed., pp.134–165). Cambridge University Press. https://doi.org/10.1017/ CBO9780511782039.007 Talmy, L. (1985). Lexicalization patterns: Semantic structure in lexical forms. In T.Shopen (Ed.), Language Typology and Syntactic Description (pp. 36–149). Cambridge University Press. van Ierland, S. (2009). Grammatical Features Influencing Information Structure: The Case of L1 and L2 Dutch and English. LOT. von Stutterheim, C. (2003). Linguistic structure and information organisation: The case of very advanced learners. In S.H.FosterCohen & S.Pekarek Doehler (Eds.), EUROSLA Yearbook, 3 (pp. 183–206). John Benjamins. https://doi.org/10.1075/eurosla.3.11stu von Stutterheim, C., & Lambert, M. (2005). Cross-linguistic analysis of temporal perspectives in text production. In H.Hendriks (Ed.), The Structure of Learner Varieties (pp. 203–230). De Gruyter. https:// doi.org/10.1515/9783110909593.203 von Stutterheim, C. & Nüse, R. (2003). Processes of conceptualization in language production: Language-specific perspectives and event construal. Linguistics, 41(5). https://doi. org/10.1515/ling.2003.028 von Stutterheim, C., Andermann, M., Carroll, M., Flecken, M., & Schmiedtová, B. (2012). How grammaticized concepts shape event conceptualization in language production: Insights from linguistic analysis, eye tracking data, and memory performance. Linguistics, 50(4), 833–867. https://doi.org/10.1515/ling2012-0026 Anna Marklová | Department of Slavic and Hungarian Studies, Faculty of Language, Literature and Humanities, Humboldt Universität zu Berlin <[email protected]> Barbara Mertins | Institute for Diversity Studies, Fakultät 15 Kulturwissenschaften, TU Dortmund <[email protected]> OPEN ACCESS