ArticlePDF Available

Abstract and Figures

To successfully understand spoken language, listeners need to determine how words within sentences relate to one another. Although the ability to compute relationships between word categories is known to develop early in life, little research has been conducted on infants’ early sensitivity to subcategorical dependencies, such as those evoked by grammatical gender (where the article form is dictated by the noun’s gender). This study therefore examines whether French-learning 18-month-olds track such relationships. Using the Visual Fixation Procedure, infants were presented with article-noun sequences in which the gender-marked article either matched (e.g., la-FEM poussette-FEM ‘the stroller’) or mismatched (e.g., le-MASC poussette-FEM) the gender of the noun. A clear preference for correct over incorrect co-occurrences was observed, suggesting that by 18 months of age, children’s storage and access of words is sufficiently sophisticated to include the means to track subcategorical dependencies. This early sensitivity to gender information may be greatly beneficial for constraining lexical access during online language processing.
Content may be subject to copyright.
Infants’ Acquisition of Grammatical Gender Dependencies
Marieke van Heugten and Anne Christophe
Laboratoire de Sciences Cognitives et Psycholinguistique (ENS / EHESS / CNRS)
Département d’Études Cognitives, École Normale Supérieure – PSL Research University
Author Note
Marieke van Heugten, Laboratoire de Sciences Cognitives et Psycholinguistique (ENS /
EHESS / CNRS), Département d’Études Cognitives, École Normale Supérieure – PSL Research
University; Anne Christophe, Laboratoire de Sciences Cognitives et Psycholinguistique (ENS /
EHESS / CNRS), Département d’Études Cognitives, École Normale Supérieure – PSL Research
This work was supported by a Fyssen postdoctoral grant, as well as grants from the
French Agence Nationale de la Recherche (ANR-2010-BLAN-1901, ANR-13-APPR-0012,
ANR-10-0001-02 PSL* and ANR-10-LABX-0087), the Fondation de France, and the Région
Ile-de-France. We thank all children and their families for their participation.
Correspondence should be addressed to Marieke van Heugten, Department of
Psychology, University at Buffalo, State University of New York, Buffalo, NY 14260, United
States. E-mail:
To successfully understand spoken language, listeners need to determine how words within
sentences relate to one another. Although the ability to compute relationships between word
categories is known to develop early in life, little research has been conducted on infants’ early
sensitivity to subcategorical dependencies, such as those evoked by grammatical gender (where
the article form is dictated by the noun’s gender). This study therefore examines whether French-
learning 18-month-olds track such relationships. Using the Visual Fixation Procedure, infants
were presented with article-noun sequences in which the gender-marked article either matched
(e.g., laFEM poussetteFEM ‘the stroller’) or mismatched (e.g., leMASC poussetteFEM) the gender of
the noun. A clear preference for correct over incorrect co-occurrences was observed, suggesting
that by 18 months of age, children’s storage and access of words is sufficiently sophisticated to
include the means to track subcategorical dependencies. This early sensitivity to gender
information may be greatly beneficial for constraining lexical access during online language
Keywords: grammatical gender, determiners, morphosyntax, language development
Infants’ Acquisition of Grammatical Gender Dependencies
Within their first few years of life, children rapidly and without much instruction acquire
their native language. By the time they reach their first birthday, infants typically start
pronouncing their first words and their productive vocabulary quickly expands to an approximate
average of 300 words around age two (Dale & Fenson, 1996). Over time, toddlers start to
combine words into short phrases (e.g., Brown, 1973) and by the preschool period, most children
speak in full sentences. The building blocks for such language skills are, however, in place long
before evidence for its productive use is observed. As early as by six months of age, for instance,
children map frequently occurring words in their language input onto visual representations of
these items (Bergelson & Swingley, 2012; Tincoff & Jusczyk, 2012) and in the ensuing months,
infants start tuning into the formal relationships among linguistic elements as a function of the
arrangement of words and phrases within an utterance.
Relationships between words within an utterance are often conveyed by means of
function morphemes, such as determiners (e.g., the) or pronouns (e.g., she). Although these
frequently occurring, closed-class words tend to be omitted from children’s early productions,
and have traditionally been thought to be overlooked until long after children start talking
(Brown, 1973), more recent work testing the perception of these items has revealed that children
do process them. Towards the end of the first year of life, for example, infants encode and
recognize the sound forms of function words (e.g., Shafer, Shucard, Shucard, & Gerken, 1998;
Shi, Cutler, Werker, & Cruickshank, 2006) and use this knowledge to segment phrases into its
individual components (Hallé, Durand, & de Boysson-Bardies, 2008; Shi et al., 2006; Shi &
Lepage, 2008). Given that functors tend to be indicative of the word class of subsequent
linguistic material (e.g., the is more likely to be followed by nouns such as ball than by verbs
such as walked; Mintz, 2003; Redington, Chater, & Finch, 1998), knowledge of function words
can furthermore constrain word candidates to the items of a specific syntactic category. Indeed,
by 18 months of age, infants readily exploit function words both to deduce the word class of
novel words (Höhle, Weissenborn, Kiefer, Schulz, & Schmitz, 2004; Shi & Melançon, 2010) and
to facilitate the recognition of familiar words (Cauvet et al., 2014; Gerken & McIntosh, 1993;
Kedar, Casasola, & Lust, 2006; Van Heugten & Johnson, 2011; Zangl & Fernald, 2007).
Relationships between words do not merely exist between major word classes, but can
also involve more subtle subcategories. In French, for instance, the form of singular articles
depends on the gender of the associated noun (e.g., leMASC doudouMASC ‘the security blanket’ but
laFEM poussetteFEM ‘the stroller’). Although the dependency between article form and noun
gender is largely independent of the semantics of the word,1 French learners nonetheless exploit
this co-occurrence pattern during linguistic processing by 25 months of age (Van Heugten & Shi,
2009; also see Johnson, 2005; Lew-Williams & Fernald, 2007 for similar results with slightly
older Dutch- and Spanish-learners). However, not much is known regarding the developmental
trajectory of this sensitivity to grammatical gender. In particular, it is currently unclear whether
children start acquiring this article-noun dependency early on in life or whether it is only learned
around their second birthday, though there is reason to suspect that tracking gender co-
occurrences can be challenging prior to age two. Specifically, artificial language work suggests
that English-learning 17-month-olds gain sensitivity to an unfamiliar (though relatively simple)
gender system only when multiple correlated phonological cues are available (Gerken, Wilson,
& Lewis, 2005). In addition, acquiring the abstract grammatical categories of new words has
1French does have some semantically transparent words, mostly nouns referring to human
entities (e.g., laFEM filleFEM ‘the girl’ vs. leMASC garconMASC ‘the boy’), but these are considered
the exception (Corbett, 1991).
been shown to be hard for young children when task demands are relatively high (Cyr & Shi,
2013). In that study, children were taught four novel word forms preceded by a gender-marked
indefinite article as the sole cue denoting its gender (e.g., unMASC ravoleMASC; ‘a
<pseudoword>’). At test, children were presented with the same pseudowords preceded either by
the correct or by the incorrect definite article (e.g., correct: leMASC ravoleMASC ‘the
<pseudoword>’; incorrect: laFEM ravoleMASC ‘the <pseudoword>’). Using this design, children
were found to differentiate between the two types of test trials only by 30 months of age,
although 20-month-olds were subsequently shown to succeed when task demands were reduced.
What does this mean for learning gender dependencies in the case of naturalistic
language input? On the one hand, children’s language input lacks the simplicity associated with
laboratory experiments and learning gender from everyday input may be relatively hard
compared to laboratory tasks. On the other hand, experiments typically last only a few minutes
(compared to a more extensive time span available for learning dependencies in natural
languages), potentially making it easier to learn novel dependencies from the input. To
differentiate between these two possibilities, this study examines whether French-learning
children track and store the subcategorical gender information of familiar words before their
second birthday. Using the Visual Fixation Procedure, 18-month-olds were presented with lists
of grammatical and ungrammatical noun phrases (NPs), the grammatical lists containing correct-
gender definite articles (e.g., leMASC doudouMASC ‘the security blanket’; laFEM poussetteFEM ‘the
stroller’) and the ungrammatical lists containing incorrect-gender definite articles (e.g., laFEM
doudouMASC ‘the security blanket’; leMASC poussetteFEM ‘the stroller’). Following previous work
pitting grammatical and ungrammatical utterances using similar procedures (e.g., Santelmann &
Jusczyk, 1998; Van Heugten & Johnson, 2010; Van Heugten & Shi, 2010), a preference for the
correct over the incorrect NPs will be taken to indicate children’s sensitivity to grammatical
gender cues.
A total of 24 typically developing monolingual French-learning 18-month-olds from the
Paris area participated in this study (mean age: 18 months, 7 days; age range: 17 months, 21 days
-19 months, 9 days; 12 girls). None of these children were known to have any language or
hearing problem. An additional three infants were tested, but excluded from the analyses due to
extreme fussiness. Participating children received a diploma as a token of appreciation.
Twelve nouns (six monosyllabic and six bisyllabic), generally known by 18-month-olds,
were used in this study. Half of these nouns were masculine and the other half feminine
(masculine: chat ‘cat’, chien ‘dog’, pain ‘bread’, bébé ‘baby’, biberon ‘bottle’, doudou ‘security
blanket’; feminine: bouche ‘mouth’, couche ‘diaper’, main ‘hand’, cuillère ‘spoon’, compote
‘stewed fruit’, poussette ‘stroller’). A female native speaker of French recorded the materials. To
avoid recording ungrammatical utterances (potentially resulting in unnatural tokens), the
materials were cross-spliced. For each NP token used in the study, three versions were recorded.
One version contained the noun preceded by the correct definite article (e.g., la poussette ‘the
stroller’). From this version, the noun was spliced. In the other two versions, the noun was
replaced by a word starting with the same consonant-vowel sequence. One of these nouns was
masculine and the other one feminine. Both were preceded by the correct definite article (e.g., le
poulet ‘the chicken’; la poupée ‘the doll’). From these two tokens, the article was selected. In
order to prevent mismatches in co-articulation between the article vowel and noun onset, articles
and nouns were spliced immediately following the noun onset consonant, just before the first
vowel of the noun. The combination of the articles with the noun resulted in two NPs, one
containing the correct definite article (e.g., la poussette) and one containing the incorrect definite
article (e.g., le poussette), that were similar in duration and intonation (average correct NP
length: 855 ms; average incorrect NP length: 865 ms).
Eight lists were created, four of which contained the correct NPs and four of which
contained incorrect NPs. Lists contained two unique tokens of each of the twelve NPs. These
NPs were ordered differently across each of the four correct and incorrect lists, but the order was
identical across these two conditions. Interstimulus pauses were approximately 750 ms long. All
lists lasted 38.3 s.
In addition to these auditory materials, which formed the core of the experiment, two
movies were used. One movie, a silent clip from the Flurry screensaver featuring colorful swirls,
was used to capture the children’s looks toward the screen during stimulus presentation. The
other movie, a red flashing light accompanied by a cartoon boing, was used as an attention getter
during the intertrial period.
Infants were individually tested for their spontaneous listening preferences using the
Visual Fixation Procedure. During test, they sat on the parent’s lap in front of a TV screen in a
sound-attenuated booth. At the front of the booth, next to the TV, two loudspeakers were
positioned at the child’s ear level. Stimuli were presented using the Lincoln Infant Lab Package
1.0 software package (Meints & Woodford, 2008). The procedure started by displaying the
attention getter on the screen. Once the child oriented towards the screen, the experimenter,
monitoring the child’s looking behavior through a videocamera, initiated the first trial. During
each trial, the experimenter pushed a button when the child oriented towards the screen. When
the child looked away, the experimenter released the button. If the child reoriented towards the
screen within two seconds, the trial continued to play, but the time looked away was subtracted
from the orientation time. Trials lasted until the child looked away for more than two seconds or
until the maximum list length was reached. The experiment finished when the child had been
presented with all eight lists. All children listened to the same lists, though the order in which
these lists appeared was randomized separately for each child.
Parental reports confirmed that children generally understood most of the words used in
the study (mean number of words comprehended: 10.88 out of 12; range: 6-12). Children
oriented towards the screen for an average of 21.76 s in Correct trials and for an average of 17.59
s in Incorrect trials (see Figure 1), with 18 out of 24 children listening longer to Correct than to
Incorrect trials. A 2-tailed paired samples t-test revealed that these two values differ significantly
from each other (4.17 s; SEM = 1.52 s; t(23) = 2.747; p = 0.011; d = .581). Neither the child’s
age nor the number of test words parents reported their child to know correlated with this
difference score. Thus, by 18 months of age, infants display a robust preference for correctly
used over incorrectly used definite articles, indicating that they track the co-occurrence between
article and noun during language processing.
This study shows that French-learning 18-month-olds track (and store) the subtle co-
occurrence patterns between article form and noun gender. By presenting infants with correct
and incorrect article-noun sequences, we identified a listening preference for correct over
incorrect phrases, suggesting that children are sensitive to gender dependencies in the months
prior to their second birthday. Children’s daily language input thus provides them with the
information necessary to learn subcategorical relationships early in life.
How could this early sensitivity to gender cues help children process spoken language?
Although the current study does not explicitly test whether the NPs were segmented into its
individual components, infants’ early sensitivity to the form and usage of function words (Hallé
et al., 2008; Shi et al., 2006; Shi & Lepage, 2008) makes it unlikely that 18-month-olds consider
article-noun sequences to be unanalyzed single words (e.g., lapoussette). Instead, the recognition
of an article (e.g., la) will allow children to infer the presence of a noun (Cauvet et al., 2014;
Höhle et al., 2004; Shi & Melançon, 2010), and its associated gender cues may subsequently
prevent (or eliminate) the activation of candidates that do not belong to the same subcategorical
gender category. For example, upon hearing la pou, children may continue to consider both
feminine-gender poussette ‘stroller’ and poupée ‘doll’ as potential nouns, but masculine-gender
poulet ‘chicken’ should no longer be an option (cf. Johnson, 2005; Lew-Williams & Fernald,
2007; Van Heugten & Shi, 2009 for similar patterns with older children). Note that the current
study does not test whether the acquired gender knowledge is grammatical (i.e. masculine nouns
should be preceded by masculine articles; feminine nouns should be preceded by feminine
articles) or distributional (i.e. doudou ‘security blanket’ should be preceded by le; poussette
‘stroller’ should be preceded by la; though see, Cyr & Shi, 2013 for evidence for grammatical
abstraction). It may thus be either the surface form of the articles or their associated gender cues
that could drive the potential advantage for online language processing. Future work could tease
apart these two possibilities.
In sum, this study reveals that infants as young as 18 months track the co-occurrence
patterns between articles and nouns in their daily language input in a robust fashion and use this
information during speech perception. This early sensitivity to such a subtle subcategorical cue
could be greatly beneficial for constraining lexical access and underlines the tremendous
sophistication of early syntactic processing.
Bergelson, E., & Swingley, D. (2012). At 6–9 months, human infants know the meanings of
many common nouns. Proceedings of the National Academy of Sciences, 109(9), 3253–
Brown, R. (1973). Development of the first language in the human species. American
Psychologist, 28(2), 97–106.
Cauvet, E., Limissuri, R., Millotte, S., Skoruppa, K., Cabrol, D., & Christophe, A. (2014).
Function words constrain on-line recognition of verbs and nouns in French 18-month-
olds. Language Learning and Development, 10(1), 1–18.
Corbett, G. G. (1991). Gender. Cambridge, UK: Cambridge University Press.
Cyr, M., & Shi, R. (2013). Development of abstract grammatical categorization in infants. Child
Development, 84(2), 617–629.
Dale, P. S., & Fenson, L. (1996). Lexical development norms for young children. Behavior
Research Methods, Instruments, & Computers, 28(1), 125–127.
Gerken, L., & McIntosh, B. J. (1993). Interplay of function morphemes and prosody in early
language. Developmental Psychology, 29(3), 448–457.
Gerken, L., Wilson, R., & Lewis, W. (2005). Infants can use distributional cues to form syntactic
categories. Journal of Child Language, 32(2), 249–268.
Hallé, P. A., Durand, C., & de Boysson-Bardies, B. (2008). Do 11-month-old French infants
process articles? Language and Speech, 51(1-2), 23–44.
Höhle, B., Weissenborn, J., Kiefer, D., Schulz, A., & Schmitz, M. (2004). Functional elements in
infants’ speech processing: The role of determiners in the syntactic categorization of
lexical elements. Infancy, 5(3), 341–353.
Johnson, E. K. (2005). Grammatical gender and early word recognition in Dutch. In Proceedings
of the 29th annual Boston University conference on language development (Vol. 1, pp.
320–330). Somerville, MA: Cascadilla Press.
Kedar, Y., Casasola, M., & Lust, B. (2006). Getting there faster: 18- and 24-month-old infants’
use of function words to determine reference. Child Development, 77(2), 325–338.
Lew-Williams, C., & Fernald, A. (2007). Young children learning Spanish make rapid use of
grammatical gender in spoken word recognition. Psychological Science, 18(3), 193–198.
Meints, K., & Woodford, A. (2008). Lincoln Infant Lab Package 1.0: A new programme
package for IPL, Preferential Listening, Habituation and Eyetracking. [WWW document:
Computer software & manual]. URL:
Mintz, T. H. (2003). Frequent frames as a cue for grammatical categories in child directed
speech. Cognition, 90(1), 91–117.
Redington, M., Chater, N., & Finch, S. (1998). Distributional Information: A powerful cue for
acquiring syntactic categories. Cognitive Science, 22(4), 425–469.
Santelmann, L. M., & Jusczyk, P. W. (1998). Sensitivity to discontinuous dependencies in
language learners: Evidence for limitations in processing space. Cognition, 69(2), 105–
Shafer, V. L., Shucard, D. W., Shucard, J. L., & Gerken, L. (1998). An electrophysiological
study of infants’ sensitivity to the sound patterns of English speech. Journal of Speech,
Language and Hearing Research, 41(4), 874–886.
Shi, R., Cutler, A., Werker, J., & Cruickshank, M. (2006). Frequency and form as determinants
of functor sensitivity in English-acquiring infants. The Journal of the Acoustical Society
of America, 119(6), EL61–EL67.
Shi, R., & Lepage, M. (2008). The effect of functional morphemes on word segmentation in
preverbal infants. Developmental Science, 11(3), 407–413.
Shi, R., & Melançon, A. (2010). Syntactic categorization in French-learning infants. Infancy,
15(5), 517–533.
Tincoff, R., & Jusczyk, P. W. (2012). Six-month-olds comprehend words that refer to parts of
the body. Infancy, 17(4), 432–444.
Van Heugten, M., & Johnson, E. K. (2010). Linking infants’ distributional learning abilities to
natural language acquisition. Journal of Memory and Language, 63(2), 197–209.
Van Heugten, M., & Johnson, E. K. (2011). Gender-marked determiners help Dutch learners’
word recognition when gender information itself does not. Journal of Child Language,
38(1), 87–100.
Van Heugten, M., & Shi, R. (2009). French-learning toddlers use gender information on
determiners during word recognition. Developmental Science, 12(3), 419–425.
Van Heugten, M., & Shi, R. (2010). Infants’ sensitivity to non-adjacent dependencies across
phonological phrase boundaries. The Journal of the Acoustical Society of America,
128(5), EL223–EL228.
Zangl, R., & Fernald, A. (2007). Increasing flexibility in children’s online processing of
grammatical and nonce determiners in fluent speech. Language Learning and
Development, 3(3), 199–231.
Figure 1. Mean orientation times in seconds (error bars indicate the standard error of the mean
difference score) to correct and incorrect NPs.
Orientation time (in s)
Correct NPs
Incorrect NPs
... To address this question, we decided to test infants at the earliest stage of gender agreement acquisition. We chose the age of 17 to 18 months, the age when the earliest learning of the gender of some familiar nouns was observed (Van Heugten & Christophe, 2015). ...
... The child needs to first learn the gender of some nouns in order to process gender agreement in structures such as (1) and (2). Given that French-learning infants begin to learn the genders of nouns from around 18 months of age (Cyr & Shi, 2013;Van Heugten & Christophe, 2015), we decided to test French-learning infants of this age on the same experiment in which the 30-month-olds in our previous study had shown gender agreement. ...
... Moreover, gender agreement depends on noun gender, but the genders of nouns must be learned one by one as the child builds a vocabulary. Our study shows that at an age when infants are just beginning to learn the gender of some nouns (Van Heugten & Christophe, 2015), they immediately track gender agreement and do so in a structure-dependent fashion. The amazing timing of the emergence of these two aspects of acquisition suggests that hierarchical phrase structures must already be available to toddlers, and that the child is likely constrained by the principle of structure dependence of the universal grammar, consistent with the view in Chomskyan theory (Chomsky, 1965(Chomsky, ,1988. ...
... A novel word that co-occurs with determiners is expected to be used later on with another determiner, not a pronoun. Toddlers have also been shown to be sensitive to their native language's sub categorical dependencies (Van Heugten & Christophe, 2015) as well as to non-adjacent dependencies (Höhle, Schmitz, Santelmann, & Weissenborn, 2006;Santelmann & Jusczyk, 1998;van Heugten & Johnson, 2010). For instance, in van Heugten & Christophe (2015), French-learning 18-month-olds distinguished between grammatical and ungrammatical lists in which a determiner either matched or not the gender of the co-occurring familiar noun (e.g. ...
... In Experiment 2, only common verbs were used, either in grammatical contexts (preceded by pronouns) or in ungrammatical contexts (preceded by determiners). If infants are indeed sensitive to noun and verb grammatical dependencies, a preference for the grammatical over the ungrammatical lists is expected in both experiments, based on previous work pitting grammatical and ungrammatical sentences using similar procedures (e.g., Santelmann & Jusczyk, 1998;Van Heugten & Christophe, 2015;Van Heugten & Johnson, 2010;Van Heugten & Shi, 2010). We hypothesize that infants from both age groups can succeed in this task based on their proven word recognition skills (e.g., Hallé & Boysson-Bardies, 1994;Shi & Lepage, 2008), and their notorious ability to track simple cooccurrence patterns (e.g. ...
Full-text available
During their first months of life, infants can already distinguish function words (e.g., pronouns and determiners) from content words (e.g., verbs and nouns). Little research has explored preverbal infants' sensitivity to the relationships between these word categories. This preregistered study examines whether French‐learning 8‐ and 11‐month‐olds track the grammatical dependencies between determiners and nouns as well as pronouns and verbs. Using the Visual Fixation Procedure, infants were presented with lists containing either grammatical (e.g., tu manges “you eat”, des biberons “some bottles”) or ungrammatical (e.g., des manges “some eat”, tu biberons “you bottle”) phrases. In Experiment 1 (N = 59), the lists involved common nouns and verbs, while in Experiment 2 (N = 28), only common verbs were used. Eleven‐month‐olds showed a clear preference for correct over incorrect co‐occurrences in both experiments, while 8‐month‐olds showed a trend in the same direction. These results suggest that before their first birthday, infants' storage and access of words and word sequences are sufficiently sophisticated to include the means to track categorical dependencies. This early sensitivity to co‐occurrence patterns may be greatly beneficial for constraining lexical access and later on for learning novel words' syntactic and semantic properties.
... Thanks to behavioural online techniques that measure the process of comprehension while it is occurring without any need for overt responses by the child, some studies have demonstrated that comprehension and use of gender agreement in syntactical processing may occur even in children younger than 3 years of age. For example, French toddlers as young as 18 months track co-occurrence patterns of gender agreement between articles and nouns and use this information during speech perception, showing a preference in listening to a condition in which determiners agree in gender with nouns (or pseudonouns) (Cyr & Shi, 2013;van Heugten & Christophe, 2015). Interestingly, other findings reported that as soon as their second birthday, children can extract abstract features (i.e., gender) from function words or bound morphemes and use them to make a prediction (Arias-Trejo & Alva, 2012;Johnson, 2005;Lew-Williams & Fernald, 2007;Melançon & Shi, 2015;Smolík & Bláhová, 2019). ...
... At age 18 months, English-acquiring children use their knowledge of determiners and function words in sentence processing, orienting faster and more accurately to the visual target following the correct determiner (Kedar et al., 2006) and using function words to categorise new words as nouns or verbs (Echols & Marti, 2004). In addition, toddlers at 18 months of age can specifically understand gender agreement, preferring a match condition in which articles agree in gender with nouns (or pseudo-nouns, Cyr & Shi, 2013;van Heugten & Christophe, 2015). To our knowledge, the present study provides the first evidence that children at 20 months of age, and even at 12 months of age, do not only understand the gender feature in determiners, but also use it to recognise the upcoming target word. ...
We investigated online early comprehension in Italian children aged 12 and 20 months, focusing on the role of morphosyntactic features (i.e., gender) carried by determiners in facilitating comprehension and anticipating upcoming words. A naturalistic eye-tracking procedure was employed, recording looking behaviours during a classical Looking-While-Listening task. Children were presented with sentences and pictures of two objects representing nouns characterised by either the same gender (determiner was uninformative) or a different gender (determiner was informative). As expected, 20-month-old children recognised the target picture when this was named, and they were faster in the different-gender condition. Interestingly, 12-month-old infants identified the target picture only when presented with an informative determiner (different-gender condition). These results suggest that, as early as 12 months of age and with an improvement seen at 20 months of age, toddlers can extract and use determiner gender features to enhance comprehension and make predictions about upcoming words.
... First, the semantic seed assumption proposes that when they approach the categorization task, infants have already succeeded in learning the meaning of a few words (frequent, referring to concrete objects and actions, presented in pragmatically helpful situations), and are able to group them into semantic classes: object referents and action referents (both parts of the assumption well supported by the infant literature, as seen above). Second, the model supposes that infants are able to keep track of bi-and trigram frequencies: a number of experiments support this assumption, showing that infants as young as 12 months pay attention to this type of distributional information, both when exposed to artificial languages (e.g., Gomez and Gerken, 1999;Marchetto and Bonatti, 2013), or when listening to sentences in their mother tongue (e.g., Santelmann and Jusczyk, 1998;Höhle et al., 2006;van Heugten and Johnson, 2010;van Heugten and Christophe, 2015). Note that the model is mimicking comprehension, since it attempts to categorize words from its input (on the basis of their linguistic context), in the hope of guessing their potential meanings, just as an infant would do when attempting to decode language. ...
Full-text available
While many studies have shown that toddlers are able to detect syntactic regularities in speech, the learning mechanism allowing them to do this is still largely unclear. In this article, we use computational modeling to assess the plausibility of a context-based learning mechanism for the acquisition of nouns and verbs. We hypothesize that infants can assign basic semantic features, such as “is-an-object” and/or “is-an-action,” to the very first words they learn, then use these words, the semantic seed , to ground proto-categories of nouns and verbs. The contexts in which these words occur, would then be exploited to bootstrap the noun and verb categories: unknown words are attributed to the class that has been observed most frequently in the corresponding context. To test our hypothesis, we designed a series of computational experiments which used French corpora of child-directed speech and different sizes of semantic seed. We partitioned these corpora in training and test sets: the model extracted the two-word contexts of the seed from the training sets, then used them to predict the syntactic category of content words from the test sets. This very simple algorithm demonstrated to be highly efficient in a categorization task: even the smallest semantic seed (only 8 nouns and 1 verb known) yields a very high precision (~90% of new nouns; ~80% of new verbs). Recall, in contrast, was low for small seeds, and increased with the seed size. Interestingly, we observed that the contexts used most often by the model featured function words, which is in line with what we know about infants' language development. Crucially, for the learning method we evaluated here, all initialization hypotheses are plausible and fit the developmental literature (semantic seed and ability to analyse contexts). While this experiment cannot prove that this learning mechanism is indeed used by infants, it demonstrates the feasibility of a realistic learning hypothesis, by using an algorithm that relies on very little computational and memory resources. Altogether, this supports the idea that a probabilistic, context-based mechanism can be very efficient for the acquisition of syntactic categories in infants.
... This shows that children did not merely reproduce article-noun combinations and could not rely on co-occurrence patterns but had to select the appropriate form of the article themselves. This goes beyond what can be shown by observational studies (Mills, 1986;Szagun et al., 2007) and what has been found by most of the perception studies (Johnson, 2005;Lew-Williams & Fernald, 2007;van Heugten & Christophe, 2015;van Heugten & Shi, 2009), although in two studies abstract gender knowledge was necessary because the gender markers used in the test phase were different from the ones used during familiarization (Cyr & Shi, 2013;Melançon & Shi, 2015). ...
Conference Paper
Full-text available
Learning novel nouns in a language with grammatical gender requires the learner to memorize gender on an item-by-item basis and/or to use cues for gender assignment. While gender assignment in German is said to be largely arbitrary, several cues to gender do exist. Amongst them are phonological cues which are probabilistic not deterministic (e.g., Köpcke & Zubin, 1996). Adults and 7- to 8-year-old children have been shown to exploit these regularities when learning pseudowords (Mills, 1986). Szagun and colleagues (2007) analyzed spontaneous speech samples of German-learning children (1.3–2.8 years of age) and found that nouns with phonologically consistent gender were marked more accurately for gender than those with phonologically inconsistent gender. The current study investigates experimentally children’s reliance on phonological cues for gender assignment in a group of 3.5-year olds (N=21). We hypothesized that children are more accurate to learn the gender of novel nouns when phonological cues support gender assignment. We picked the strongest probabilistic phonological cues, with 92% and 78% predictive power for the associated gender, respectively (Szagun, 2019): the final Schwa syllable for feminine nouns and the ending -er for masculine nouns. We created 16 disyllabic pseudowords. In addition, 16 existing animal names were selected (8 feminine, 8 masculine), half of which conformed to the phonological cues, half of which did not. An elicitation task was administered as a game where “Bruno the bear” was looking for his friends and the child had to help him. On each trial, a familiar and a novel animal were labeled by the experimenter three times each such that the gender marking appeared on indefinite determiners, adjectives, or demonstratives. After that, one of the animals was hidden and the child was asked to tell “Bruno the bear” which animal that was. This scenario triggered the use of a definite article which is unambiguously gender marked in German but was never used during the introduction. Accuracy scores were analyzed in an ANOVA with word type (words/pseudowords), gender (feminine/masculine) and consistency to the phonological regularity as within-subject factors. Gender did not have any effect (p>.5). Gender marking for pseudowords was less accurate than for words (p<.001) and phonologically consistent labels were more accurately marked than inconsistent ones (p<.001). An interaction between word type and consistency showed that consistency had an effect within the pseudowords (p<.001) but not within the words (p>.2). The vast majority of errors (75%) was due to regularizing gender according to the phonological cue. We conclude that 3-year-old German children make use of phonological cues when learning the grammatical gender of novel nouns. This corroborates findings from previous corpus studies (Szagun et al., 2007) and experimental studies with older children (Mills, 1986). The association of gender with some phonological noun properties aids children in learning grammatical gender such that grammatical acquisition can proceed smoothly. This is an important step for further syntactic development in German, especially because case marking interacts with gender.
Full-text available
Although L1 French speakers (FS) acquire the formal features of gender and number early, agreement appears to take longer, leading to persistent difficulties even for cases of straightforward agreement within a nominal or verbal phrase. This begs the questions of how adult FSs (n = 168) may fare with idiosyncratic cases of agreement such as nominal affective constructions and past participles as measured by a written grammaticality judgment /correction task and preference/grammaticality judgment task. The findings showing that participants performed better at correctly accepting than rejecting stimuli, are consistent with an increasing number of empirical studies revealing individual differences among adult L1 speakers. The findings are discussed from a generative perspective and the usage-based perspective of the Basic Language Cognition-High Language Cognition theory of L1 proficiency (Hulstijn, 2015).
Adult processing of other-accented speech is fast, dependent on lexical access, and readily generalizable to new words. But what does children's processing of other-accented speech look like? Although many acquisition researchers have emphasized how other-accented speech presents a formidable challenge to young children, we argue that the field has perhaps underestimated children's early accent processing abilities. In support of this view, we present evidence that 2-year-olds’ accent processing abilities appear to be in many respects adult-like, and discuss the growing literature on children's ability to cope with multi-accent input in the natural world. We outline different theoretical outlooks on the transition children make from infancy to later childhood, and discuss how the growing sophistication of infants’ accent processing abilities feeds into their social perception of the world (and perhaps vice versa). We also argue that efficient processing and meaningful interpretation of accent variation are fundamental to human cognition, and that early proficiency with accent variation (along with all of the implied representational and learning capacities) is difficult to explain without assuming the early emergence of abstract speech representations. Expected final online publication date for the Annual Review of Linguistics, Volume 8 is January 2022. Please see for revised estimates.
Each language has its unique way to mark grammatical information such as gender, number and tense. For example, English marks number and tense/aspect information with morphological suffixes (e.g., ‐s or ‐ed). These morphological suffixes are crucial for language acquisition as they are basic building blocks of syntax, encode relationships, and convey meaning. Previous research shows that English‐learning infants recognize morphological suffixes attached to nonce words by the end of the first year, although even 8‐month‐olds recognize them when they are attached to known words. These results support an acquisition trajectory where discovery of meaning guides infants’ acquisition of morphological suffixes. In this paper, we re‐evaluated English–learning infants’ knowledge of morphological suffixes in the first year of life. We found that 6–month–olds successfully segmented nonce words suffixed with –s,–ing,–ed and a pseudo‐morpheme ‐sh. Additionally, they related nonce words suffixed with –s, but not ‐ing,‐ed or a pseudo‐morpheme –sh with stems. By 8–months, infants were also able to relate nonce words suffixed with –ing and stems. Our results show that infants demonstrate knowledge of morphological relatedness from the earliest stages of acquisition. They do so even in the absence of access to meaning. Based on these results, we argue for a developmental timeline where the acquisition of morphology is, at least, concurrent with the acquisition of phonology and meaning.
Full-text available
We assessed monolingual Spanish and bilingual Spanish-Basque toddlers’ sensitivity to gender agreement in correct vs. incorrect Spanish noun phrases (definite article + noun), using a spontaneous preference listening paradigm. Monolingual Spanish-learning toddlers exhibited a tendency to listen longer to the grammatically correct phrases (e.g., la casa; “the house”), as opposed to the incorrect ones (e.g., * el casa). This listening preference toward correct phrases is in line with earlier results obtained from French monolingual 18-month-olds (van Heugten & Christophe, 2015). Bilingual toddlers in the current study, however, tended to listen longer to the incorrect phrases. Basque was not a source of interference in the bilingual toddler’s input as Basque does not instantiate grammatical gender agreement. Overall, our results suggest that both monolingual and bilingual toddlers can distinguish between the correct and incorrect phrases by 18 months of age; however, monolinguals and bilinguals allocate their attention differently when processing grammatically incorrect forms.
Full-text available
In this experiment using the conditioned head-turn procedure, 18-month-old French-learning toddlers were trained to respond to either a target noun (“la balle”/the ball) or a target verb (“je mange”/I eat). They were then tested on target word recognition in two syntactic contexts: the target word was preceded either by a correct function word (“une balle”/a ball or “on mange”/they eat), or by an incorrect function word, signaling a word from the other category (∗“on balle”/they ball or ∗“une mange”/a eat). We showed that 18-month-olds exploit the syntactic context on-line to recognize the target word: verbs were recognized when preceded by a personal pronoun but not when preceded by a determiner and vice-versa for nouns. These results suggest that 18-month-olds already know noun and verb contexts. As a result, they might be able to exploit them to categorize unknown words and constrain their possible meaning (nouns typically refer to objects whereas verbs typically refer to actions).
Full-text available
This study examined abstract syntactic categorization in infants, using the case of grammatical gender. Ninety-six French-learning 14-, 17-, 20-, and 30-month-olds completed the study. In a preferential looking procedure infants were tested on their generalized knowledge of grammatical gender involving pseudonouns and gender-marking determiners. The pseudonouns were controlled to contain no phonological or acoustical cues to gender. The determiner gender feature was the only information available. During familiarization, some pseudonouns followed a masculine determiner and others a feminine determiner. Test trials presented the same pseudonouns with different determiners in correct (consistent with familiarization gender pairing) versus incorrect gender agreement. Twenty-month-olds showed emerging knowledge of gender categorization and agreement. This knowledge was robust in 30-month-olds. These findings demonstrate that abstract, productive grammatical representations are present early in acquisition.
Full-text available
Previous studies have demonstrated that children are aware of the function morphemes in their language despite their failure to produce them. However, none of these studies tested whether children are aware of the linguistic contexts in which particular function morphemes occur. Only if children are aware of such co-occurrence patterns could they use function morphemes to determine the linguistic categories of words and phrases. Young 2-yr-olds demonstrated their awareness of function morpheme co-occurrence patterns by performing better in a picture identification task when the target word was preceded by a grammatical article than an ungrammatical auxiliary. Children who heard the sentences produced in a female voice performed better than those who heard a male voice, and this was especially true for sentences exhibiting the most regular co-occurrence patterns. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Gender is a fascinating category, central and pervasive in some languages and totally absent in others. In this new, comprehensive account of gender systems, over 200 languages are discussed, from English and Russian to Archi and Chichewa. Detailed analysis of individual languages provides clear illustrations of specific types of system. Gender distinction is often based on sex; sometimes this is only one criterion and the gender of nouns depends on other factors (thus 'house' is masculine in Russian, feminine in French and neuter in Tamil). Some languages have comparable distinctions such as human/non-human, animate/inanimate, where sex is irrelevant. No other textbook surveys gender across this range of languages. Gender will be invaluable both for class use and as a reference resource for students and researchers in linguistics.
Discusses psycholinguistic research on the acquisition of sentence construction rules by children. Field work in 1st language development using tape recordings of preschool children speaking more than 30 different languages is noted. The systematic errors made by children tend to reflect either inconsistencies in morphology and syntax within the language or the complexity of the rules. The early methods used by transformational linguistic studies are criticized and the inadequacies of the various descriptive characterizations (e.g., telegraphic speech and pivot grammar) are discussed. Semantic development at the various stages is analyzed in detail. Similarities and differences between the learning of a 2nd language by an adult and 1st language acquisition in the child are considered. The implications of the proposal that automatic internal programs of structure extraction determine language development is noted. (26 ref.) (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Comprehending spoken words requires a lexicon of sound patterns and knowledge of their referents in the world. Tincoff and Jusczyk (1999) demonstrated that 6-month-olds link the sound patterns “Mommy” and “Daddy” to video images of their parents, but not to other adults. This finding suggests that comprehension emerges at this young age and might take the form of very specific word-world links, as in “Mommy” referring only to the infant’s mother and “Daddy” referring only to the infant’s father. The current study was designed to investigate if 6-month-olds also show evidence of comprehending words that can refer to categories of objects. The results show that 6-month-olds link the sound patterns “hand” and “feet” to videos of an adult’s hand and feet. This finding suggests that very early comprehension has a capacity beyond specific, one-to-one, associations. Future research will need to consider how developing categorization abilities, social experiences, and parent word use influence the beginnings of word comprehension.
Recent work showed that infants recognize and store function words starting from the age of 6–8 months. Using a visual fixation procedure, the present study tested whether French-learning 14-month-olds have the knowledge of syntactic categories of determiners and pronouns, respectively, and whether they can use these function words for categorizing novel words to nouns and verbs. The prosodic characteristics of novel words stimuli for noun versus verb uses were balanced. The only distinguishing cue was the preceding determiners versus subject pronouns, the former being the most common for nouns and the latter the most common for verbs, i.e., Det + Noun, Pron + Verb. We expected that noun categorization may be easier than verb categorization because the co-occurrence of determiners with nouns is more consistent than that of subject pronouns with verbs in French. The results showed that infants grouped the individual determiners as one common class, and that they appeared to use the determiners to categorize novel words into nouns. However, we found no evidence of verb categorization. Unlike determiners, pronouns were not perceived as a common syntactic class.
How do children determine the syntactic category of novel words? In this article we present the results of 2 experiments that investigated whether German children between 12 and 16 months of age can use distributional knowledge that determiners precede nouns and subject pronouns precede verbs to syntactically categorize adjacent novel words. Evidence from the head-turn preference paradigm shows that, although 12- to 13-month-olds cannot do this, 14- to 16-month-olds are able to use a determiner to categorize a following novel word as a noun. In contrast, no categorization effect was found for a novel word following a subject pronoun. To understand this difference we analyzed adult child-directed speech. This analysis showed that there are in fact stronger co-occurrence relations between determiners and nouns than between subject pronouns and verbs. Thus, in German determiners may be more reliable cues to the syntactic category of an adjacent novel word than are subject pronouns. We propose that the capacity to syntactically categorize novel words, demonstrated here for the first time in children this young, mediates between the recognition of the specific morphosyntactic frame in which a novel word appears and the word-to-world mapping that is needed to build up a semantic representation for the novel word.
We examined infants' recognition of functors and the accuracy of the representations that infants construct of the perceived word forms. Auditory stimuli were “Functor + Content Word” versus “Nonsense Functor + Content Word” sequences. Eight-, 11-, and 13-month-old infants heard both real functors and matched nonsense functors (prosodically analogous to their real counterparts but containing a segmental change). Results reveal that 13-month-olds recognized functors with attention to segmental detail. Eight-month-olds did not distinguish real versus nonsense functors. The performance of 11-month-olds fell in between that of the older and younger groups, consistent with an emerging recognition of real functors. The three age groups exhibited a clear developmental trend. We propose that in the earliest stages of vocabulary acquisition, function elements receive no segmentally detailed representations, but such representations are gradually constructed so that once vocabulary growth starts in earnest, fully specified functor representations are in place to support it.