When context matters more than language: Verb or noun in French and Turkish caregiver speech

Abstract

The main goal of this article is to study the respective role of language typology and context on the noun to verb asymmetry in caregiver speech. The speech of 20 French- and 20 Turkish-speaking mothers addressed to their children in two different situations (book-reading and toy-play) were analysed in terms of noun to verb ratio as well as in terms of object-oriented to action-oriented utterances ratio. Only a tendency to noun orientation was shown in French mothers. Both groups of mothers behave in the same ways in both contexts with more nouns and object-oriented utterances in the book-reading context vs more verbs and action-oriented utterances in toy-play. These results confirm a more important role of context on mothers’ linguistic behaviours than language typology.

Keywords

Caregiver context French nouns Turkish verbs

Introduction

Two different perspectives regarding the early trajectories of children’s first language acquisition have been highly influential. These are, in short, the Universal Noun Advantage View and the Input Dependent View. According to the Universal Noun Advantage View, children’s early lexicon is made up of nouns, representing concrete objects (Gentner, 1981, 1982; Maratsos, 1991; Markman, 1987). Verbs, and verb-like items, are acquired later as they involve more cognitively complex tasks for children to accomplish. This account has been supported by a great number of studies in various languages (in English by Bornstein et al., 2004; Fenson et al., 1993; Goldfield, 2000, in Italian by Bornstein et al., 2004; Caselli et al., 1995; Salerni, Assanelli, D’Odorico, & Rossi, 2007; Tardif, Shatz, & Naigles, 1997, in Hebrew by Bornstein et al., 2004; Dromi, 1987; Maital, Dromi, Sagi, & Bornstein, 2000, in Spanish by Jackson-Maldonado, Thal, Marchman, Bates, & Guitierrez-Clellen, 1993 and in Dutch, Bornstein et al., 2004; De Houwer & Gillis, 1998; Verlinden & Gillis, 1988). However, researchers challenging this view claim that the characteristics of any language being acquired are as important as cognitive factors. The linguistic properties of a language, such as word-order and morphology, may make some word categories more salient than others (Choi, 2006; Choi & Gopnik, 1995; Gathercole & Min, 1997; Gopnik & Choi, 1990; Tardif, 1996).

Pragmatic aspect in input

In addition to cognition and language-based perspectives in the aforementioned literature about children’s early vocabulary composition, several studies have shown that pragmatic aspects of caregiver speech should also be taken into consideration to explain noun vs verb dominance in children’s early vocabularies (Choi, 2000; Choi & Gopnik, 1995; Fernald & Morikawa, 1993; Kim, McGregor, & Thompson, 2000; Ogura, Dale, Yamashita, Murase, & Mahieu, 2006). The coding of maternal speech in terms of its pragmatic intention has indicated the emergence of two categories: naming-oriented/object-oriented utterances and activity-oriented utterances. In general, naming-oriented utterances include those used to elicit an object from the child or an object label directly or indirectly or those encouraging the child to focus on an object or an entity. Activity-oriented utterances encourage the child to focus on an action or state or suggesting some type of action (Choi, 2000, 2006; Choi & Gopnik, 1995; Kim et al., 2000; Masur, Flynn, & Lloyd, 2013; Ogura et al., 2006). Results of these analyses have shown stylistic as well as cultural differences among mothers: Korean caregiver speech, for example, includes more activity-oriented utterances than English caregiver speech (Choi & Gopnik, 1995; Kim et al., 2000).

Contextual factors affecting input

These pragmatic differences mentioned above are accentuated in some specific contexts. Context is considered as activity context, i.e. the activities that participants are engaged in while a conversation takes place. A number of studies have found evidence for large variations in maternal speech as a function of the ongoing activity type (Goldfield, 1993; Hoff, 2010; Leaper & Gleason, 1996; Lucariello & Nelson, 1986; Luo, Snow, & Chang, 2011; O’Brien & Nagle, 1987). Of these, Goldfield’s study (1993) led to the suggestion to consider the context of parent–child interaction as an important factor in the domain of noun and verb use. She studied 12 English mother–child dyads in two different contexts: toy-play and non-toy-play. She found that the toy-play context, in itself, generated more nouns as mothers labelled objects during their play with their children. However, in the non-toy-play setting (physical play), mothers used more varied verbs. The production of a higher proportion of noun types in a book context than in a toy context by caregivers appears to be a very widespread phenomenon, occurring even in languages considered to be verb biased. Gelman and Tardif (1998) compared caregivers’ input in picture-book reading (caregivers described the pictures of a book without text) and toy-play contexts in English and Mandarin. They found differences in the proportion of nouns and verbs in caregivers’ speech between the two languages and between contexts. Overall, English-speaking mothers provided more nouns than Mandarin-speaking mothers, whereas Mandarin-speaking mothers provided more verbs than English-speaking mothers. There was also a context effect: mothers in both groups produced more nouns than verbs in the book-context but they exhibited the reverse pattern in the toy-context. More recently, Choi (2000), in her comparative study based on 20 English and 20 Korean mother–child dyads, obtained a similar context effect: whereas English-speaking mothers emphasized nouns in both picture-book reading and toy-play contexts, Korean-speaking mothers did so only in the book-context. In the toy-play context, they provided more verbs and focused more on actions.

Objectives of this study

The findings reviewed above have highlighted the importance of pragmatic and contextual factors on the use of verb or nouns in child directed speech (CDS). However, three limitations are apparent. First, only very few studies have considered these factors simultaneously. Second, some results were directly compared without taking into account heterogeneity in data collection methods. Third, the majority of studies concerned the same language comparison: English was compared to Asian languages. This limited set of language comparisons poses a problem for generalization of findings.

Our goal in this study is to describe patterns of input in two typologically diverse languages that have not been previously compared: French and Turkish. The main purpose is to compare Turkish and French caregiver speech systematically in two contexts (picture-book reading and toy-play), with a specific focus on (1) mothers’ use of verb vs noun and (2) mothers’ pragmatic intention, in their talk to their children. The following questions will be addressed:

To what extent will Turkish- and French-speaking mothers differ in terms of noun vs verb use?

To what extent will context play a role in their linguistic choice?

The comparison between French and Turkish has been chosen due to their typological differences in the domain of noun/verb use differences. In French, nouns are acquired earlier than verbs as reported both in longitudinal naturalistic studies (Bassano, 2000; Bassano, Maillochon, & Eme, 1998; Parisse & Le Normand, 2000) and in checklist-based studies (Kern, 2007; Kern & Gayraud, 2010). Parisse and Le Normand (2000) recorded the symbolic play of 27 French-speaking children of 2 years of age. They reported a larger number of nouns than verb types: 29% of nouns were produced in one word utterances and 44% in longer utterances vs only 18 and 22% of verbs. In a longitudinal study of one French child between 14 and 30 months as well as a cross-sectional study of 12 children at 20 months and 12 children at 30 months, Bassano et al. (1998) showed that nouns and paralexical items (interjections, particles, etc.) are the most frequent types until 20 months, prior to the emergence of predicates (verbs and adjectives) and grammatical words (e.g. articles, prepositions, conjunctions). Results obtained with the French Communicative Development Inventory (Kern, 2007; Kern & Gayraud, 2010) confirms this result: between 8 and 16 months, 80% of words produced were nouns. From 16 to 30 months, nouns are well but less represented (58.5%) because of the increase of verbs and adjectives in children’s repertoire.

A relatively large number of studies have also accumulated in the literature about the early acquisition of nouns and verbs by Turkish children. These studies focus on various aspects such as distribution, frequency, morphological and positional properties of nouns and verbs. Ketrez and Aksu-Koç (2003) have investigated the inflectional paradigms of nouns and verbs in one child’s spontaneous speech sample. They found that these two categories are acquired at around the same age. Türkay (2005), in a longitudinal study over one year with five children, found that verbs are acquired at the same frequency as nouns. No regular pattern of noun distribution has been observed in distributional analysis of nouns and verbs in CDS (Ketrez & Aksu-Koç, 2003; Küntay & Slobin, 1996). These two studies, however, seem to contradict each other in terms of the verb category. Küntay and Slobin (1996) suggest that distributional characteristics of verbs in Turkish CDS do not represent an informative pattern to the Turkish children. In contrast, Ketrez and Aksu-Koç (2003) conclude that verbs exhibit a regular distributional scheme.

The limited number of cross-linguistic comparisons of Turkish provide some challenging findings. In their cross-linguistic and longitudinal study based on four French- and five Turkish-speaking children and their mothers, Kern and Türkay (2006) suggested that nouns (in both type and token) were dominant in both languages before and after the lexical spurt. However, the difference between nouns and verbs in Turkish children was less important than in French children. Moreover, Kauschke, Lee, and Pae (2007), in a comparison among Turkish, German and Korean, reported that Turkish and Korean children were better than German children at verb-naming abilities. In summary, though descriptive features of Turkish may place Turkish in the group of verb-friendly languages, the patterns in the actual Turkish input call into question assumptions about the verb-friendliness of Turkish. That is, when verb- and noun-friendliness are regarded as a bipolar continuum, Turkish should not be positioned at the frontier of verb-friendliness end of that continuum. According to Kauschke and Hofmeister’s perspective (2002), in particular languages, a weaker version of the noun-bias hypothesis should be considered.

Some linguistic and caregiver speech characteristics can be considered as possible explanations for these differences in French-learning and Turkish-learning children. French can be considered as a noun-friendly language, whereas Turkish presents some common properties with verb-friendly languages such as Korean. Subject-Verb-Object (SVO) word order is one of the noun-friendly characteristics of French. Canonical word order in French is SVO even if various different orders are permitted in some limited contexts (Harris, 1990). For instance, SOV order can be found in the case of object pronominalization or in object relative clauses. However, in most of the cases, the syntactic function of the object is at the end of the utterances which may contribute to the appearance of nouns in the salient utterance-final position and explain early noun learning in French. Turkish canonical word order is Subject-Object-Verb (Göksel & Kerslake, 2005; Slobin, 1982) but because of the flexibility of the SOV order, this parameter may not be a reliable tenet in Turkish to keep a positional and distributional track of any lexical category. This diminishes verb saliency in Turkish and can potentially differentiate Turkish from other verb-friendly languages (e.g. Korean) in which word order is more rigid. In addition, noun phrases can be omitted as long as their referent is clear in the discourse in Turkish: a single verb with appropriate morphology can stand for a complete sentence. The parameter of pro-dropness can favour verbs and their early acquisition in Turkish.

Nominal and verbal morphology are also very different in French and Turkish. There are large differences in the diversity, transparency and regularity of marking across the two languages. Morphology is poorer in French than in Turkish in both nouns and verbs. In French, nouns are inflected only for number (singular and plural) and gender (feminine and masculine), whereas in Turkish, nouns are inflected for number (singular and plural), case (accusative, dative, locative, ablative and genitive) and possession (1st, 2nd or 3rd person, singular or plural). In spoken French, neither the plurality marker –s nor the gender marker –e on nouns is pronounced but a large number of nouns are preceded by determiners with gender and/or number marking. Turkish verbs are even more inflected than Turkish nouns. Verbs in Turkish are marked for person, number, tense, aspect, modality, voice, negation and interrogation. Verbs in French are only marked for number, tense and gender. French has three groups of verbs that are conjugated differently. To summarize, the Turkish morphological system (and particularly the verb morphological system) is more complex (more different morphemes) than the French system. However, the Turkish morphological system is transparent and regular. Inflections appear in the form of suffixes in Turkish. Both these characteristics seem to play a positive role in the acquisition of morphology.

Hypotheses

Cross-linguistic and as well as cross-contextual hypotheses are tested in this study. The examination of frequency of word classes (noun vs verb) and the pragmatic emphasis on object or action will help us determine if mothers are noun or verb dominant according to their native language. French can be considered a noun-friendly language with syntax and morphology favouring nouns whereas Turkish is very close to a verb-friendly language, with verbs mostly at the end of utterances and a particularly rich but regular and transparent morphology. This leads us to predict noun dominance in French with more nouns and more object-oriented utterances than in Turkish. For Turkish, we predict a different trend, with more verbs and verb-oriented utterances than in French. However, due to a certain degree of freedom in Turkish word order, as well as a relatively complex morphological noun system, the verb dominance could be weaker (with a smaller difference between verb and noun frequencies than in French). The design enables also to evaluate the influence of context on caregivers’ speech. We expect to have more nouns and more object-oriented utterances in picture-book reading and more verbs and action- oriented utterances in toy-play.

Method

Participants

Twenty French-speaking mothers and 20 Turkish-speaking mothers and their children participated in this study. The French children were between 1;4 and 2;11, with an average age of 23 months (9 boys and 11 girls). The Turkish children were between 1;4 and 2;8, with an average age of 24 months (10 boys and 10 girls). In both groups, the participants had monolingual backgrounds. The Turkish data were collected in Turkey with families living in the same city centre (Adana) and the French data were collected in France (Lyon). Both French and Turkish families were representatives of mid-socioeconomic classes. In this study, maternal educational attainment was strictly kept under control. All Turkish mothers were university graduates and already working and all French mothers had their diplomas in different specializations. All children were reported as mentally and physically well-developed by their parents, with no indications of developmental problems.

Materials

Two different sets of materials were used to elicit mothers’ input. For the picture-book reading session, we used The farm picture book (Amery & Cartwright, 1986), a wordless book with five unrelated pictures of a farm. In this booklet, there are scenes of a farm and a group of people who are busy doing different farm-related activities. This book was chosen as it included scenes similar to those depicted in the main reference studies (Choi, 2000; Kim et al., 2000). For the toy-play session, we used a miniature house with separable entities (table, bed, etc.) and figures (a father, a mother, a boy and a girl).

Procedure

The interactions between the mother and the child during a picture-book reading session and a toy-play session, each lasting 10 minutes, were videotaped by the same experimenter both in France and Turkey. All maternal and children utterances were transcribed from the videotape in the CHAT format for analysis by the Computerized Language Analysis (CLAN) programs (MacWhinney, 1991) by native speakers of both languages. Partially intelligible utterances, exclamations utterances and sentence fragments that contained no nouns or verbs were not included in the coding. Two types of analysis were performed: frequency of nouns and verbs (type and token) and frequency of pragmatic utterance type.

Lexical measures

The lexical coding identified all nouns and verbs in each caregiver’s utterance. The category of noun included common nouns, proper names and kinship terms. All vocative uses of nouns were excluded from the analysis of nouns. This rule concerned mainly the use of the child’s name by the mother to get the child’s attention.

The category of verb included only main verbs of a clause. Only lexical dynamical verbs indicating an action, a process, or a sensation (e.g. courir in French and koş- in Turkish [to run], manger in French and ye- in Turkish [to eat]) were analysed. State verbs were not considered (e.g. penser in French and düşün- in Turkish [to think], préférer in French and tercih et- in Turkish [to prefer], dormir in French and uyu- in Turkish [to sleep]). In the case of verbs which could be either action or state verbs depending on the context, they were considered only when they referred to an action (e.g. sentir in French and kokla- in Turkish [to smell] in ‘il sent la fleur’ [he smells the flower] but not sentir in ‘ça sent bon’ [it smells good]). In addition, the verb category excluded all auxiliary verbs in French, modal words, various forms of the copula (e.g. ‘c’est, il y a’ [there is] in French, ‘köpek var’ [there is a dog] in Turkish) as well as verbs such as ‘to look’ (regarde [look] in French, and bak in Turkish) when they were used as attention getting terms and consequently did not refer to an action. In the caregiver samples, both types and tokens were counted for nouns and verbs.

Pragmatic measures

All maternal utterances were categorized into three categories, replicating coding schemes in related studies (Choi, 2000; Kim et al., 2000; Ogura et al., 2006): object-oriented utterances (OOU), action-oriented utterances (AOU) and others. Object-oriented utterances were the ones reinforcing the child to concentrate on an object or an entity in their environment by means of their mother’s utterance. These were: labelling utterances in a declarative form (e.g. ‘c’est une histoire’ [this is a story] in French and ‘bu bir kukla’ [this is a puppet] in Turkish), object-identification questions (e.g. ‘qu’est-ce-que c’est?’ [what is this?] in French and ‘bu ne?’ [what is this?] in Turkish) and object-focus commands (e.g. ‘regarde ce petit poulet!’ [look at this small chicken] in French and ‘şu tavsana bak!’ [look at this rabbit] in Turkish). Action-oriented utterances constituted utterances reinforcing the child to pay attention to a punctual or durative action. These were action-description utterances (e.g. ‘cet enfant est en train de donner quelque chose à manger’ [this child is giving something to eat] in French and ‘çocuk zıplıyor’ [the child is jumping] in Turkish), action-asking utterances (e.g. ‘qu’est-ce-qu’il fait?’ [what is he doing?] in French and ‘adam ne yapıyor?’ [what is the man doing?] in Turkish), action-focus commands (e.g. ‘tu mets ça dans la cuisine!’ [you put this into the kitchen] in French and ‘tavukları çağırıyor’ [she is calling the chicken] in Turkish). Finally, the ‘other’ category included some ambiguous utterances (e.g. ‘et là’ [and here]) and discourse fragments without verbs or nouns (e.g. ‘oh oui bravo’ [oh yes, congratulations]). For the pragmatic analyses, all transcriptions were independently categorized by two native language speaker coders. The inter-rater variability for this measure for Turkish was 87% and for French 90%. The disagreements were resolved through discussion.

Design and analyses

Each set of mother–child participants took part in both a picture-book reading and toy-play session. Only mothers’ productions were considered. Productions’ length in utterance were calculated. As lexical measures, we calculated the ratio n/(n+v), where n is the number of noun types and v is the number of verb types for mothers’ speech. This ratio will be symbolized by N-typ/(n+v). The parallel ratio based on noun and verb tokens was also computed for mothers. It is symbolized by N-tok/(n+v). In both cases (type and token), a score greater than 0.50 indicates that mothers produced more noun types than verb types and more noun tokens than verb tokens. To observe the pragmatic orientation of mothers’ productions, we calculated the ratio of OO/(OO+AO), where OO is the number of object-oriented utterances and AO the number of action-oriented utterances. A score greater than 0.50 indicates that mothers produced more object-oriented utterances than action-oriented utterances. We evaluated the effect of language and context on these three ratios using repeated measure 2 × 2 ANOVA with Language (French – Fr and Turkish – T) as a between-subjects variable and Context (picture-book reading – BR and toy-play – TP) as a within-subjects variable. In addition, in order to evaluate effect sizes, we calculated partial eta-squares (Cohen, 1988).

Results

The results of mothers’ production length and lexical and pragmatic measures according to languages and contexts are given in Table 1.

Table 1.

Length in utterances, noun/verb type ratio, noun/verb token ratio and object-oriented utterances/action-oriented utterances ratio (mean, standard deviation, min-max) in mothers’ directed speech according to languages (French and Turkish) and context (book-reading, BR and toy-play, TP).

Languages/Contexts	BR	TP	Total languages
Length in utterances
French	175.5, 41.5, 112–239	175.8, 43.4, 87–288	175.7, 34.24, 87–288
Turkish	178.1, 36, 128–248	182.3, 50.8, 82–244	180.3, 30.5, 82–248
Total contexts	177.3, 31.9, 113–248	178.7, 33, 82–288	178, 32.5, 82–288
N-typ/(n+v) ratio
French	0.53, 0.04, 0.41–0.6	0.45, 0.07, 0.37–0.58	0.49, 0.07, 0.37–0.58
Turkish	0.51, 0.04, 0.44–0.58	0.45, 0.06, 0.31–0.59	0.48, 0.06, 0.31–0.59
Total contexts	0.52, 0.05, 0.41–0.59	0.45, 0.07, 0.31–0.59	0.49, 0.06, 0.31–0.59
N-tok/(n+v)ratio
French	0.54, 0.07, 0.32–0.63	0.45, 0.07, 0.37–0.58	0.49, 0.08, 0.32–0.63
Turkish	0.53, 0.06, 0.43–0.62	0.43, 0.07, 0.33–0.63	0.48, 0.08, 0.43–0.63
Total contexts	0.54, 0.06, 0.32–0.63	0.44, 0.07, 0.33–0.63	0.49, 0.08, 0.32–0.63
OO/(OO+AO)
French	0.57, 0.07, 0.42–0.69	0.39, 0.1, 0.28–0.75	0.48, 0.12, 0.28–0.75
Turkish	0.56, 0.07, 0.39–0.68	0.37, 0.09, 0.14–0.67	0.46, 0.1, 0.14–0.68
Total contexts	0.56, 0.07, 0.39–0.69	0.38, 0.11, 0.14–0.75	0.47, 0.13, 0.14–0.75

Mothers’ production length

In the book-reading context, French and Turkish mothers produced 175.5 (SD = 41.5, range: 113–239) and 178.1 (SD = 36, range: 128–248) utterances, respectively. In the toy-play context, French and Turkish mothers produced averages of 175.8 (SD = 43. 4, range: 87–288) and 182.3 (SD = 50.8, range: 82–244) utterances, respectively. A repeated measure 2 × 2 ANOVA, with Language (French – Fr and Turkish – T) as a between-subjects variable and Context (picture-book reading – BR and toy-play – TP) as a within-subjects variable, showed no effect of language, of context and no significant interaction between language and context on production length.

Lexical measures: Mothers’ use of noun vs verb types

Table 1 summarizes the analyses of N-typ/(n+v) ratios as a function of language and context. Only an effect of context on N-typ/(n+v) ratios was found, F(2,38) = 23.353, p < .001, η_p² = .55) with a ratio of 0.52 in BR and only 0.45 in the TP context, which means that there are more noun types than verb types in the book context as compared to the play context.

Lexical measures: Mothers’ use of noun vs verb tokens

Table 1 presents the results concerning the N-tok/(n+v) ratio according to language and context (mean and SD). A repeated measure 2 × 2 ANOVA revealed an effect of context, F(2,38) = 43.716, p < .001, η_p² = .70, but no effect of language on the N-tok/(n+v) ratio. In both languages, the N-tok/(n+v) ratio was higher in the BR context than in TP, indicating a higher use of noun (vs verbs) in the BR context than in TP. Both languages combined, the N-tok/(n+v) ratio is 0.54 in BR and only 0.44 in TP.

Pragmatic measures: Mothers’ use of object-oriented vs action-oriented utterances

The results concerning the OO/(OO+AO) ratio according to language and context (mean and SD) are given in Table 1. Only context had a statistically significant influence on the ratio. The OO/(OO+AO) ratio was significantly higher, F(2,38) = 43.9, p < .001, η_p² = .70) in the BR context than in the TP context (0.56 vs 0.38) in both languages indicating a higher use of object-oriented utterances (vs action-oriented utterances) in BR than in TP. Moreover, the OO/(OO+AO) ratio was higher than 0.5 in BR and lower than 0.5 in TP in French and Turkish, which shows a preference for object-oriented utterances in BR and action-oriented utterances in TP.

Discussion

Our main objectives were to observe the possible influence of language typology and context on two lexical measures (the use of verbs vs nouns in type and token) and one pragmatic measure (the use of object- vs action-oriented utterances by mothers).

Because French can be described as a more noun-friendly language as compared to Turkish, we expected more nouns (types and tokens) and object-oriented utterances in French CDS than in Turkish CDS. Moreover, considering the influence of context shown in previous studies, we expected a noun and an object-oriented dominance in the BR context as compared to the TP context. This predicts an interaction between language and context.

There was no language effect either on the linguistic or on the pragmatic measures. Independent of context, both groups of mothers tended to favour slightly the use of verbs (types and tokens) over nouns (types and tokens) and the use of action-oriented utterances over object-oriented utterances in their CDS. Based on these results, our first hypothesis is not confirmed: lexical and pragmatic measures indicate no real noun orientation in French as compared to Turkish. These results are different from results reported in the literature for French: we did not replicate the noun bias in French CDS. The difference observed in French mothers can be explained by methodological differences: the studies generally involve either naturalistic data which are very heterogeneous in terms of recorded situations or parental reports which are known to favour noun acquisition, whereas, in our case, two similar and specific situations were used to collect data in both languages. However, our results are in accordance with the findings of both Goldfield (1993) and Tardif et al. (1997). The latter compared the relative emphasis of nouns or verbs in CDS for English, Italian and Mandarin. They found that English and Italian mothers used as many noun types as verb types whereas Mandarin mothers produced more verb types. They were also able to show that adult-to-child speech in all three groups contained more verb than noun tokens.

In comparison to language, context had an influence on all measures. Our data revealed higher use of noun types and tokens (vs verb types and tokens) in the BR than in the TP context, in which conversely mothers produced more verb types and tokens. The same effect is seen in the pragmatic measures: in both languages, mothers used more object-oriented utterances in the BR context and more verb-oriented utterances in the TP context. As seen through the measure of effect sizes, context explains 70% of the variance on noun/verb ratio type and object-oriented/action-oriented ratio and 55% of the variance on noun/verb ratio tokens, which are very large effects in all three cases. This result confirms our second expectation and shows clearly that context played a more important and regular role in mothers’ linguistic behaviours than language typology. The importance of context on the use of verbs or nouns has already been described for Mandarin, Japanese, English and Korean caregiver speech. Ogura et al. (2006) found for Japanese mothers a significant effect of context with higher N-typ/(n+v) and N-tok/(n+v) ratios in a picture-book reading context. In their cross-linguistic study, Tardif, Gelman, and Xu (1999) compared the noun type ratio to the verb type ratio in the CDS of two groups of mothers (Mandarin and Japanese). They were able to show in the three groups a preference for nouns over verbs in picture-book reading and the opposite trend in toy-play. This behaviour was also present in our French and Turkish mothers. However, in comparison with our population, the difference between verb type and noun type frequencies was more important in Tardif’s population: N-typ/(n+v) = 0.30 in TP and 0.55 in BR for Mandarin; 0.33 and 0.55 in Japanese; 0.45 and 0.52 for French and 0.44 and 0.51 for Turkish. In addition, a larger amount of noun tokens over verb tokens in the picture-book reading context has been revealed in Korean caregiver speech (Choi, 2000). The production of more nouns in the picture-book reading context appears to be a very widespread phenomenon, occurring even in languages which could be considered as verb biased.

However, frequency is not the entire story. Other input characteristics have to be considered, such as perceptual and structural saliency of verbs and nouns as well as morphological variations in verbal and nominal systems. These analyses will be the next step in our Turkish/French comparison. We will focus on the position of verbs and nouns in CDS as well as on the morphological richness and transparency of the linguistic systems.

Mean ages of the children in both the French and the Turkish samples were respectively 23 and 24 months with relatively large ranges. Much of the literature on nouns vs verbs in the speech of young children has focused on the single word period. In addition, a significant number of studies on CDS have shown different speech characteristics of this register according to age. One possible extension would be to analyse the speech directed to younger as well as to older children to consider the developmental trajectory.

So far, early trajectories of children’s first language acquisition and especially early lexical composition have been explained by two main perspectives: a cognitive perspective which postulates that language acquisition trajectories are constrained by general cognitive development (noun advantage view) and an input dependent view in which linguistic development is constrained by typological characteristics of languages to be acquired. This second perspective grants importance to CDS which should provide the role of a model to follow. These two main perspectives have been integrated by Gentner and Boroditsky (2001), who claim that both cognitive and linguistic factors interact during children’s language acquisition. Our findings seem to indicate a third perspective: the context dependent view. Context of data collection seems to play a more regular and important role in mothers’ linguistic behaviour (Hoff, 2010; Murase, Dale, Ogura, Yamashita, & Mahieu, 2005). To understand cross-linguistic differences in lexical composition, systematic studies of cognitive development of children have to be conducted but also systematic studies of input across contexts.

Footnotes

Acknowledgements

The data presented in the article are original data collected in the framework of the project ‘Language-general and language-specific processes between Turkish and French speaking children’s early lexical development and effects of child-directed speech’, main investigators: N. Feyza Altınkamış and Sophie Kern.

Funding

The project was financially supported by Tübitak (Turkey) and CNRS (France) (2008–2010).

References

Amery

Cartwright

(1986). The farm picture book. London, England: Usborn.

Bassano

(2000). Early development of nouns and verbs in French: Exploring the interface between lexicon and grammar. Journal of Child Language, 27, 521–559.

Bassano

Maillochon

Eme

(1998). Developmental changes and variability in early lexicon: A study of French children’s naturalistic productions. Journal of Child Language, 25, 493–531.

Bornstein

M. H.

Cote

L. R.

Maital

Painter

Park

S. Y.

Pascual

. . .Vyt

(2004). Cross-linguistic analysis of vocabulary in young children: Spanish, Dutch, French, Hebrew, Italian, Korean, and American English. Child Development, 75, 1115–1139.

Caselli

M. C.

Bates

Casadio

Fenson

Sanderl

Weir

(1995). A cross-linguistic study of early lexical development. Cognitive Development, 10, 159–200.

Choi

(2000). Caregiver input in English and Korean: Use of nouns and verbs in book-reading and toy-play contexts. Journal of Child Language, 27, 69–96.

Choi

(2006). Influence of language-specific input on spatial cognition: Categories of containment. First Language, 26, 207–232.

Choi

Gopnik

(1995). Early acquisition of verbs in Korean: A cross-linguistic study. Journal of Child Language, 22, 497–529.

Cohen

(1988). Statistical power analysis for the behavioral sciences (2nd ed.). New York, NY: Lawrence Erlbaum.

10.

De Houwer

Gillis

. (1998). Dutch child language: An overview. In Gillis

De Houwer

(Eds.), The acquisition of Dutch (pp. 1–100). Amsterdam, The Netherlands: John Benjamins.

11.

Dromi

(1987). Early lexical development. New York, NY: Cambridge University Press.

12.

Fenson

Dale

P. S.

Reznick

J. S.

Thal

Bates

Hartung

J. P.

. . .Reilly

J. S.

(1993). MacArthur Communicative Development Inventories. San Diego, CA: Singular.

13.

Fernald

Morikawa

(1993). Common themes and cultural variations in Japanese and American mothers’ speech to infants. Child Development, 64, 637–656.

14.

Gathercole

V. C. M.

Min

(1997). Word meaning biases or language-specific effects? Evidence from English, Spanish and Korean. First Language, 17, 31–56.

15.

Gelman

Tardif

(1998). Acquisition of nouns and verbs in Mandarin and English. In Clark

(Ed.), The Proceedings of the Twenty-ninth Annual Child Language Research Forum (pp. 27–36). Stanford, CA: Stanford University Press.

16.

Gentner

(1981). Some interesting differences between verbs and nouns. Cognition and Brain Theory, 4, 161–178.

17.

Gentner

(1982). Why nouns are learned before verbs: Linguistic relativity versus natural partitioning. In Kuczaj

S. A.

(Ed.), Language development. Volume 2: Language, thought, and culture (pp. 301–334). Hillsdale, NJ: Lawrence Erlbaum.

18.

Gentner

Boroditsky

(2001). Individuation, relativity, and early word learning. In Bowerman

Levinson

S. C.

(Eds.), Language acquisition and conceptual development (pp. 215–256). Cambridge, UK: Cambridge University Press.

19.

Göksel

Kerslake

(2005). Turkish: A comprehensive grammar. London, England and New York, NY: Routledge.

20.

Goldfield

B. A.

(1993). Noun bias in maternal speech to one-year-olds. Journal of Child Language, 20, 85–99.

21.

Goldfield

B. A.

(2000). Nouns before verbs in comprehension vs. production: The view from pragmatics. Journal of Child Language, 27, 501–520.

22.

Gopnik

Choi

(1990). Do linguistic differences lead to cognitive differences? A cross-linguistic study of semantic and cognitive development. First Language, 10, 199–215.

23.

Harris

(1990). French. In Comrie

(Ed.), The world’s major languages (pp. 210–235). New York, NY: Oxford University Press.

24.

Hoff

(2010). Context effects on young children’s language use: The influence of conversational setting and partner. First Language, 30, 461–472.

25.

Jackson-Maldonado

Thal

Marchman

Bates

Guitierrez-Clellen

(1993). Early lexical development of Spanish-speaking infants and toddlers. Journal of Child Language, 20, 523–549.

26.

Kauschke

Hofmeister

(2002). Early lexical development in German: A study on vocabulary growth and vocabulary composition during the second and third year of life. Journal of Child Language, 29, 735–757.

27.

Kauschke

Lee

W. H.

Pae

(2007). Similarities and variation in noun and verb acquisition: A cross-linguistic study of children learning German, Korean, and Turkish. Language and Cognitive Processes, 22, 1045–1072.

28.

Kern

(2007). Lexicon development in French-speaking infants. First Language, 27, 227–250.

29.

Kern

Gayraud

(2010). IFDC. Grenoble, France: Les éditions La Cigale.

30.

Kern

Türkay

F.N.

(2006). Influence cognitive générale et/ou langagière spécifique sur les premiers mots des enfants: étude pilote chez des enfants français et des enfants turcs [Language general and language specific effects on children’s first words: pilot study on Turkish and French children]. Poster presented at ARCO, Bordeaux, 6–8 December.

31.

Ketrez

Aksu-Koç

(2003). Acquisition of noun and verb categories in Turkish. In Özsoy

A. S.

Akar

Nakıpoğlu-Demiralp

Erguvanlı-Taylan

Aksu-Koç

(Eds.), Proceedings of the Tenth International Conference in Turkish Linguistics (pp. 239–245). İstanbul, Turkey: Boğaziçi University Press.

32.

Kim

McGregor

Thompson

C. K.

(2000). Early lexical development in English and Korean speaking children: Language-general and language-specific patterns. Journal of Child Language, 27, 225–254.

33.

Küntay

Slobin

D. I.

(1996). Listening to a Turkish mother: Some puzzles for acquisition. In Slobin

D. I.

Gerhardt

Kyratzis

Guo

(Eds), Social interaction, social context, and language (pp. 265–286). Mahwah, NJ: Lawrence Erlbaum.

34.

Leaper

Gleason

J. B.

(1996). The relationship of play activity and gender to parent and child sex-typed communication. International Journal of Behavioral Development, 19, 689–703.

35.

Lucariello

Nelson

(1986). Context effects on lexical specificity in maternal and child discourse. Journal of Child Language, 13, 507–522.

36.

Luo

Snow

E. C.

Chang

(2011). Mother–child talk during joint book reading in low-income American and Taiwanese families. First Language, 32, 494–511.

37.

MacWhinney

(1991). The CHILDES project: Tools for analyzing talk. Hillsdale, NJ: Lawrence Erlbaum.

38.

Maital

Dromi

E. L.

Sagi

Bornstein

M. H.

(2000). The Hebrew communicative development inventory: Language specific properties and cross-linguistic generalizations. Journal of Child Language, 27, 43–67.

39.

Maratsos

(1991). How the acquisition of nouns may be different than that of verbs. In Krasnegor

Rumbaugh

Schiefelbusch

Studdert-Kennedy

(Eds.), Biological and behavioral determinants of language development (pp. 67–88). Hillsdale, NJ: Lawrence Erlbaum.

40.

Markman

E. M.

(1987). Constraints children place on word meanings. Cognitive Science, 14, 57–77.

41.

Masur

E. F.

Flynn

Lloyd

C. A.

(2013). Infants’ behaviors as antecedents and consequents of mothers’ responsive and directive utterances. First Language, 33, 354–371.

42.

Murase

Dale

S. P.

Ogura

Yamashita

Mahieu

(2005). Mother–child conversation during joint picture book reading in Japan and the USA. First Language, 25, 197–218.

43.

O’Brien

Nagle

K. J.

(1987). Parents’ speech to toddlers: The effect of play context. Journal of Child Language, 14, 269–279.

44.

Ogura

Dale

P. S.

Yamashita

Murase

Mahieu

(2006). The use of nouns and verbs by Japanese children and their caregivers in book-reading and toy-playing context. Journal of Child Language, 33, 1–29.

45.

Parisse

Le Normand

M. T.

(2000). Beyond the realm of noun and verb: The cognitive lexicon of the young child. Retrieved from http://cogprints.org/2114/

46.

Salerni

Assanelli

D’Odorico

Rossi

(2007). Qualitative aspects of productive vocabulary at the 200- and 500-word stages: A comparison between spontaneous speech and parental report data. First Language, 27, 75–87.

47.

Slobin

D. I.

(1982). Universal and particular in in the acquisition of language. In Wanner

Gleitman

L. R.

(Eds.), Language acquisition: The state of art (pp. 128–170). Cambridge, UK: Cambridge University Press.

48.

Tardif

(1996). Nouns are not always learned before verbs: Evidence from Mandarin speakers’ early vocabulary. Developmental Psychology, 32, 492–504.

49.

Tardif

Gelman

S. A.

(1999). Putting the ‘noun bias’ in context: A comparison of English, Italian and Mandarin. Journal of Child Language, 24, 535–565.

50.

Tardif

Shatz

Naigles

(1997). Caregiver speech and children’s use of nouns versus verbs: A comparison of English, Italian and Mandarin. Journal of Child Language, 24, 535–565.

51.

Türkay

(2005). Children’s early lexicon in terms of noun/verb dominance (Unpublished doctoral dissertation). Çukurova University, Balcalı/Adana, Turkey.

52.

Verlinden

Gillis

(1988). Nouns and verbs in the input: Gentner (1982) reconsidered. ABLA Papers, 12, 163–187.