Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=hlld20

Download by: [77.138.2.22] Date: 28 February 2017, At: 23:38

Language Learning and Development

ISSN: 1547-5441 (Print) 1547-3341 (Online) Journal homepage: http://www.tandfonline.com/loi/hlld20

Little Words, Big Impact: Determiners Begin toBootstrap Reference by 12 Months

Yarden Kedar, Marianella Casasola, Barbara Lust & Yisrael Parmet

To cite this article: Yarden Kedar, Marianella Casasola, Barbara Lust & Yisrael Parmet (2017):Little Words, Big Impact: Determiners Begin to Bootstrap Reference by 12 Months, LanguageLearning and Development

To link to this article: http://dx.doi.org/10.1080/15475441.2017.1283229

Published online: 28 Feb 2017.

Submit your article to this journal

View related articles

View Crossmark data

Little Words, Big Impact: Determiners Begin to BootstrapReference by 12 MonthsYarden Kedara, Marianella Casasolab, Barbara Lustb, and Yisrael Parmetc

aDepartment of Early Childhood Education, Beit Berl College; bDepartment of Human Development, CornellUniversity; cDepartment of Industrial Engineering and Management, Ben-Gurion University of the Negev

ABSTRACTWe tested 12- and 18-month-old English-learning infants on a preferential-looking task which contrasted grammatically correct sentences using thedeterminer “the” vs. three ungrammatical conditions in which “the” wassubstituted by another English function word, a nonsense word, or omitted.Our design involved strict controls on phonetic composition of functionwords and co-articulation effects in the stimuli tested, which were absent inprevious studies. The results show an overall significant effect of FunctionWord (FW), with no significant interaction effects between the two agegroups. Overall, infants oriented faster to a target image and had morecorrect first looks at target following grammatical sentences. Tests withinage groups show a significant effect of FW in 18-month-olds for bothlatency and correct first look, documenting a linguistic sensitivity, whichpersists over our new acoustic variables. In the 12-month-olds there was asignificant effect of FW on first look, although no significant effect of FW forlatency appeared. However, pairwise comparisons showed a significantdifference between grammatical and ungrammatical FWs in latency evenat this age. These findings suggest that some prototypical form of sensitiv-ity to determiners in sentence processing and its use in computing nounreference exists already during early developmental stages surrounding thechild’s first words. We speculate on potential explanations for developmen-tal changes between 12 and 18 months.

Researchers continue to ask: What is the nature of children’s representation of functional elementssuch as determiners (e.g., the, a) at early stages of language development, and how may it changeover time?

A wide range of behavioral research has now documented the emergence of fine-detailed phoneticrepresentations for functional words and morphemes (FWs) throughout children’s first year of life(e.g., Demuth, 1992, 1994; Gerken, Remez, & Landau, 1990; Katz, Baker, & MacNamara, 1974;Maratsos, 1982; Petretic & Tweney, 1977; Shipley, Smith, & Gleitman, 1969). More recently, an arrayof cross-linguistic studies has shown that infants encode FWs phonetically from about 6 months ofage. In fact, already before their first birthday, infants have been found to: (i) discriminate FWs fromcontent words, demonstrated in infants even a few days old (e.g., Shi, Werker, & Morgan, 1999); (ii)segment FWs from continuous speech (Höhle & Weissenborn, 2003; Shafer, Shucard, Shucard, &Gerken, 1998; Shi & Gauthier, 2005); (iii) discriminate and prefer well-formed FWs over nonsenseFWs (Hallé, Durand, & de Boysson-Bardies, 2008; Shady, 1996; Shi, Cutler, Werker, & Cruickshank,2006; Shi, Werker, & Cutler, 2006); (iv) use FWs to segment novel noun forms (Shi & Lepage, 2008);and (v) detect specific co-occurrence patterns between bare roots and their inflected variants(Marquis & Shi, 2012).

CONTACT Yarden Kedar Head, Department of Early Childhood Education, Beit Berl College, Kfar Saba, Israel.Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/hlld.© 2017 Taylor & Francis Group, LLC

LANGUAGE LEARNING AND DEVELOPMENThttp://dx.doi.org/10.1080/15475441.2017.1283229

This line of evidence has often figured in ongoing debates regarding children’s early representa-tion and processing of functional categories (FCs). These grammatical categories, which are reflectedin language by FWs, serve as phrasal heads that carry a critical role in establishing the structuralskeleton of sentences and enable essential syntactic operations (e.g., movement) in language (Abney,1987; Chomsky, 1995; Pollock, 1989). However, some researchers have argued that early discrimi-native effects of FWs in perception reflect infants’ excellence in acoustic processing, but no abstractgrammatical knowledge. Rather, the infant changes qualitatively at some later point in time,transforming from item-based (mainly phonetic) processes in infancy to grammatical, category-based incorporation of FWs in sentence processing. This non-continuous developmental trajectoryhas been attributed to infants’ early ability to discern various acoustic and distributional character-istics that FWs often share. For example, FWs are the most frequent items in every human language;typically have a monosyllabic lexical structure; and appear in fixed peripheral phrasal positions(Aslin, 1999; Morgan, 1986; Seidl & Johnson, 2006). Only later, it is argued, do children make aqualitative shift to a point where abstract grammatical representations of FCs become available; andthis is only after the child has gained sufficient exposure to content word categories (e.g., Pine &Lieven, 1997; Tomasello, 2000a, 2000b; Zangl & Fernald, 2007).

Other scholars hypothesize that infants’ discriminatory behaviors in response to particular FWsare guided by an inherent sensitivity to linguistic properties of the FCs to which these FWs belong,and thus reveal more than perceptual processing of acoustic elements of language (e.g., Lust, 2006;Shi, 2014; Valian, 2008, 2009; Yang, 2013). Focusing on determiners, Valian, Solt, and Stewart (2008)systematically analyzed thousands of natural speech utterances from 21 children aged 1.10–2.8 yearsand argued that “even at the onset of combinatorial speech, children have abstract functionalcategories” (773). An array of studies with L1 learners of Dutch, English, French, and Germanusing various methods and measures yielded evidence that young children not only detect FWs inthe linguistic input, but also consult various aspects of FWs and the syntactic contexts in which theyappear in sentences already during their second year of life (Bernal, Dehaene-Lambertz, Millotte, &Christophe, 2010; Bernal, Lidz, Millotte, & Christophe, 2007; Bobb & Mani, 2013; Cauvet et al., 2014;Cyr & Shi, 2013; Golinkoff, Hirsh-Pasek, & Schweisguth, 2001; Höhle, Schmitz, Santelmann, &Weissenborn, 2006; Kedar, Casasola, & Lust, 2006; Robertson, Shi, & Melançon, 2012; Santelmann &Jusczyk, 1998; Shady & Gerken, 1999; Van Heugten & Shi, 2009).

Some evidence for abstract linguistic knowledge has been obtained with even younger infantsclose to the 12-month age range (Höhle, Weissenborn, Kiefer, Schulz, & Schmitz, 2004; Kedar, 2009;Shady, 1996; Shi & Melançon, 2010; Soderstrom, White, Conwell, & Morgan, 2007; Zhang, Shi, & Li,2015). For example, Höhle et al. (2004) found that 14-month-old German-learners grasp thecategory information related to a novel (nonsense) word based on the particular type of FWpreceding it (e.g., a determiner, as in ein glamm); and then also generalize this information toother syntactic contexts in which the novel word appears (e.g., with other German determiners; dasglamm). That is, they use the determiner to syntactically categorize adjacent novel words. Similarly,Shi and Melançon (2010) have found that French-learning 14-month-olds can detect and use a set ofdistinct determiners preceding novel words to categorize these words as nouns. However, theseresults are somewhat inconsistent. While both studies have found that 14-month-olds use determi-ners to categorize novel words into nouns (but not 12–13 month-olds; Höhle et al., 2004), no suchevidence was found in terms of verb categorization (i.e., subject pronouns were not perceived as acommon syntactic class). In a recent study, Zhang et al. (2015) report a significant discrimination ofgrammatical vs. ungrammatical test trials based on functors in Mandarin-Chinese-Learning 12-month-old infants, however only in later stages of the test phase (Block 2).

In sum, it has been demonstrated that very young infants distinguish FWs based on their distinctphonetic forms in the linguistic input; and also consult this phonetic discrimination in onlinesentence processing. A central question in our view is how early during development do childrenprovide evidence that they are beginning to use FWs to link form and meaning? Specifically, withregard to determiners: How early do children begin to link their linguistic discrimination of this

2 Y. KEDAR ET AL.

grammatical category (DET) to (semantic) visual reference? We pursue this question beginning atthe age of 12 months, approximately the age of first productive word usage (e.g., Fenson et al., 1994)and some early aspects of word comprehension (e.g., Swingley, 2005; Thierry, Vihman, & Roberts,2003). We find the 12-month marker to be particularly intriguing because several broader develop-mental changes have been documented at this stage in the integration and reorganization of infants’phonological, prosodic, lexical, and semantic representations (e.g., DePaolis, Vihman, & Keren-Portnoy, 2014; Kuhl, Ramírez, Bosseler, Lin, & Imada, 2014; Lust, 2006; Pruden, Hirsh-Pasek,Golinkoff, & Hennon, 2006; Werker & Yeung, 2005; Yeung, Chen, & Werker, 2014).

If FWs begin to carry a role in the integration of grammatical form and reference already around12 months of age, as they are a few months later, infants may be able to utilize this sensitivity tofacilitate the acquisition and integration of different linguistic domains from very early stages oflexical and grammatical acquisition. Indeed, early access to FWs has often been assumed to benecessary for bootstrapping various aspects of language learning (Christophe, Millotte, Bernal, &Lidz, 2008; Gerken, 2007; Shi, 2007); and to assist infants in realizing certain formal aspects in theirlanguage (Gervain, Nespor, Mazuka, Horie, & Mehler, 2008; Valian, 2008). For example, preverbalinfants have been found to rely on well-formed FWs—but not on nonsense functors—to segmentadjacent nouns (11 months: Hallé et al., 2008; 8 months; Shi & Lepage, 2008). Gervain et al. (2008)demonstrated that 8-month-old learners of Italian and Japanese consult the frequency and locationpatterns of function and content elements in these two languages to determine the linear word orderrepresentation for their native language.

This study

Our study follows the design of Kedar et al. (2006), in which 18- and 24-month-old English-learnerstested on an intermodal preferential looking procedure (IPLP; Golinkoff, Hirsh-Pasek, Cauley, &Gordon, 1987) demonstrated different looking patterns to a target image in response to grammaticalsentences vs. ungrammatical sentences in which a single FW was manipulated (e.g., “Can you see thecup?” vs. “Can you see *and ball?”). These findings were replicated in a follow-up study with sixteen12-month-old infants (Kedar, 2009). These two studies replicated the findings of a seminal study byGerken and McIntosh (1993) in which two-year-olds were tested on a picture-pointing task with asimilar set of sentences.

However, the results attained in these studies allow some alternative explanations. Most critically,the comparison between the grammatical determiner “the” vs. the nonsense word “el” and theungrammatical English FW “and” is confounded by the variation in consonant-initial (CV) syllablestructure. Studies have documented infants’ difficulty in segmenting onsetless vs. onset words(Nazzi, Dilley, Jusczyk, Shattuck-Hufnagel, & Jusczyk, 2005; Seidl & Johnson, 2008); that is,vowel-initial (VC) as opposed to consonant-initial (CV) syllables and words, respectively. Thus,infants could have discriminated these two types of forms on this phonetic basis, potentiallypreferring the consonant-initial “the” because of depression in processing of vowel-initial words inutterance medial position.

In addition, co-articulation factors could have contributed to the results. Kedar and colleaguescreated their test sentences by splicing the prototype for the carrier phrase “can you see”—as wellas the prototype nouns following the FW (e.g., phone, dog)—from grammatical sentences. Infantshave been shown to be sensitive to subtle co-articulation cues, and to use such cues for wordsegmentation (e.g., Curtin, Mintz, & Byrd, 2001; Johnson & Jusczyk, 2001). Hence, the auditorystimuli could lead to subtle, yet detectable, acoustic differences between the test sentences, withonly grammatical sentences featuring an agreement in co-articulation between the carrier phrase(can you see) and the FW (the), and between the FW and the following noun, whereas the deviantco-articulatory nature of the ungrammatical sentences may have disrupted infants’ processing ofthose sentences.

LANGUAGE LEARNING AND DEVELOPMENT 3

These acoustic differences would not in themselves explain early use of FWs to determinereference. However, such uncontrolled acoustic variables, either individually or in combination,would allow both linguistic and acoustic factors as potential contributions to infants’ greater successat the correct determiner form in our task. To explore this issue further, we created new auditorystimuli that isolate grammatical from acoustic/phonetic properties in ways that allow an assessmentof their potential contributions to the observed results.

Our study shares key features with Kedar et al. (2006) and Kedar (2009), yet modifies theirexperimental designs to control several acoustic factors as potential explanations of the results.Similar to Kedar et al., our design contrasts grammatical sentences in which the Englishdeterminer “the” precedes a monosyllabic singular count noun (e.g., “Can you see the dog?”)with three types of similar yet ungrammatical conditions, in which “the” is substituted with anEnglish FW (by), which is itself well-formed, but non-grammatical in the sentence context inwhich it appears; substituted with a nonsense word (bo); or dropped (“null”). To test infants in amanner consistent with previous studies, we again used the IPLP, which has been successfullyapplied to test various aspects of language acquisition during children’s first three years of life(Golinkoff, Ma, Song, & Hirsh-Pasek, 2013). We also endorsed the methodological modificationsin Kedar (2009), taking into account infants’ more limited attentional resources as well as theirpartial lexical familiarity at 12 months. Lexical unfamiliarity has been shown to affect infants’performance on both production and perception of sentences (e.g., Boyle & Gerken, 1997; Zangl& Fernald, 2007). Thus, we reduced the size of the original set of test nouns used in Kedar et al.based on the lexical norms of the MacArthur-Bates Communicative Development Inventory(MBCDI; Dale & Fenson, 1996) at 16 months—to 8 nouns, which all have relatively highfamiliarity rates at 12 months according to the MBCDI for monosyllabic, singular count nouns.Moreover, we replicated the noun familiarization phase in Kedar (2009) to increase the likelihoodthat infants would be familiar with the appropriate linguistic label for each of the images. Thiscould then focus infants on the actual task rather than on deciphering what a certain noun refersto, as in a word-learning task. Lastly, following Kedar (2009), the overall time during which thetarget and distractor images were simultaneously presented was reduced from 6 sec to 4 sec, todecrease the likelihood that infants lose interest in the task.

Our study addresses the two potentially confounding acoustic variables mentioned above. First, inorder to eliminate an acoustic-based account whereby the results could alternatively be explainedbased on infants’ bias in favor of the grammatical condition (“the”, a CV syllable) vs. the vowel-initial ungrammatical FWs (“and”, “el”), we maintained “the” as the FW in the grammaticalcondition, but used consonant-initial FWs in the two ungrammatical FW-substitution conditionsas well (by, bo).

Second, to avoid the possibility that distinction between grammatical and ungrammatical sen-tences could potentially be explained in terms of distinct coarticulation cues, we changed the mannerin which the auditory stimuli were prepared and edited. In this study, the words comprising the final(edited) test sentences were all spliced from ungrammatical sentences (in grammatical sentences, theFW prototype for “the” was spliced from a grammatical sentence, but the carrier phrase and thenoun were taken from the ungrammatical sentences). Our manipulation produced identical coarti-culation cues across grammatical and non-grammatical forms.

In addition, the fact that the preposition “by” served as the English FW which is not grammaticalin the test sentences (in contrast to “was” and “and” previously), allows us to test whether younglanguage learners are able to detect yet another fine-detailed distinction between two legitimateEnglish function words, each fulfilling a different grammatical function and belonging to a differentfunctional category in language. Note that combinations of [by+Noun] may and do occur in English(e.g., “He is old enough to have a drink by cup”; “Those who arrive by car can park here”; “Someathletes are getting paid by shoe companies”). Furthermore, “Can you see by” is also a legitimatesequence in English (e.g., “Can you see by the smile on my face?”; “O! Say can you see by the dawn’searly light. . .”). Hence, if infants in both age groups (12 and 18 months) discriminate “the” from the

4 Y. KEDAR ET AL.

ungrammatical FWs and advantage this well-formed determiner in computing reference of the nounin a Det-N construction (as reflected in different looking behaviors)—there will be evidence thatspecific acoustic cues cannot alone account for such discrimination and computation; as well as forcontinuity in early form-meaning integration in language development involving FCs.

Two dependent variables that capture different aspects of infants’ looking response to thelinguistic manipulation were assessed to test this hypothesis. Based on Kedar et al. (2006) andKedar (2009), we predicted latency to target (duration of time passed during test trials from themoment the pair of images first appears to infants’ initial fixation to target) to be the most sensitivemeasure because it is continuous (time-based) and tracks infants’ very first looking response to thelinguistic manipulation. Specifically, in comparison to the ungrammatical conditions, infants pre-sented with grammatical sentences were predicted to orient faster to the target image. Second, adichotomous, first look to target measure, was based on whether infants correctly oriented to thetarget image on their first look. We predicted the highest proportion of correct first looks to appearfollowing grammatical sentences with the well-formed determiner.

Method

Participants

All infants who took part in the study were reported by their parents to be full-term and healthy,with no history of any auditory or visual impairment, and hearing only English regularly. Themajority of infants at each age were Caucasian. For the 18-month-old group, 16 infants (9 females; 7males) were recruited. The mean age (months; days) was 18;05 (SD = 7.66 days) ranging between 538and 564 days. Fourteen additional infants were not included in the final sample. Eleven infantsexpressed fussiness during most or all of the testing session (i.e., when crying or disinterest in thestimuli across several consecutive trials made it impossible to complete the session), while in threeother cases a technical problem in running the experiment occurred. For the 12-month group, 18infants (eight females, 10 males) were recruited. The mean age (months; days) was 12;01(SD = 9.36 days), ranging between 349 and 380 days. Fourteen additional infants were not includeddue to fussiness during most or all of the test session (n = 8), a technical problem (n = 5), or a childraised in a bilingual environment (n = 1).

Stimuli

Eight images with the highest familiarity proportion for their corresponding nouns among 12-month-old English-learners according to the MBCDI were chosen as the visual stimuli. Four pairswith approximately the same noun-familiarity proportion in each pair were created (See Table 1).Accordingly, the auditory stimuli consisted of 32 sentences based on all possible combinations of the

Table 1. (i) Mean percentage of infant familiarity at 12 months with the test nouns according to the Macarthur-BatesCommunicative Development Inventory (MBCDI); and (ii) Mean percentage of infant estimated familiarity with these nounsaccording to the parents of the infants tested in this study (only “Yes” responses were counted as indicating infant familiarity withthe nouns).

MBCDI:Lexical Norms at 12 Months

Parental Report: 18 Months Parental Report: 12 Months

Pair Nouns “Yes” “No” “Not Sure” “Yes” “No” “Not Sure”

1 Phone 58 91.7 8.3 0 69.2 23.1 7.7Cup 55.7 91.7 8.3 0 30.8 46.2 23

2 Ball 79.5 100 0 0 78.6 7.1 14.3Book 68.2 100 0 0 76.9 7.7 15.4

3 Cat 45.5 75 16.7 8.3 76.9 23.1 0Dog 69.3 83.3 16.7 0 92.3 7.7 0

4 Car 53.4 75 25 0 53.8 23.1 23.1Shoe 62.5 75 25 0 53.8 30.8 15.4

LANGUAGE LEARNING AND DEVELOPMENT 5

eight nouns and the four exemplars of FW used in this study (See Table 2). This set of 32 sentenceswas digitally recorded and edited in several stages to create test sentences that would be distin-guished only by their grammatical structure rather than by their phonological or prosodic char-acteristics. First, a female, native-English speaker produced several versions of each of the 32sentences in moderate speed so that each word would be clearly separated from the others. Next,to avoid a possible co-articulatory bias in favor of the grammatical sentences over the ungrammaticalsentences, two adult native-English speaking judges who were blind to the experiment’s goalslistened carefully and separately several times to the set of 24 ungrammatical sentences, and thenmarked the best exemplar of the sequence “can you see” in terms of vividness, naturalness andclarity. The specific “can you see” sequence which got the highest combined score was then spliced inas a prototype for all of the edited sentences that were eventually used in the test trials (grammaticaland ungrammatical alike). Similarly, the clearest exemplar of each noun was chosen from the set ofungrammatical sentences and then spliced in as a prototype in all sentences in which it appeared(grammatical and ungrammatical alike). Finally, for each type of FW, the best exemplar agreed on bythe two judges was spliced in to create a set of eight sentences for that condition. The edited testsentences lasted about 2 s (mean: 2071.9 msec; SD = 218.65 msec; range = 1620–2498 msec),depending on the FW used and mainly on the noun (mean length of noun = 457.1 msec;SD = 94.8 msec; range: 300–592 msec). In “null” sentences, in which “the” was omitted, a verybrief gap of silence was inserted between the verb “see” and the following noun in order to keep thesentence at the 2 sec timeframe. Lastly, four (other) adult native English speakers (2 females, 2 males.Mean age = 21;08 yy;mm) were presented (separately) with each of the 32 edited test sentences andwere asked to judge whether each sentence sounded natural and had normal prosodic characteristicsfor an interrogative sentence in English. All sentences were indeed confirmed as natural sounding byeach judge.

Apparatus

Adjacent experimental and control rooms were used. In the experimental room, three 20-in colormonitors were placed on a table approximately 76 cm from the floor. Infants were seated on theirparent’s lap approximately 127 cm from the monitors. Low lighting and a black wooden framesurrounding the monitors were used to focus infants’ attention on the monitors. An opening in theframe enabled a camcorder lens to be focused on the infant’s face. This camcorder was linked to amonitor in the control room, allowing the experimenter to observe the infant during the test session.All test sessions were video-recorded to allow offline frame-by-frame coding of each infant’s lookingbehavior using the SuperCoder software (Hollich, 2003). The Habit X software (Cohen, Atkinson, &Chaput, 2004) was used for controlling the order and the timing of the audio-visual presentationthroughout the test session.

Table 2. Lists of sentences.

Test Trial Pair Target Side of Target List 1 List 2 List 3 List 4

1 Cup-Phone Phone Left the phone? by phone? bo phone? _ phone?2 Car-Shoe Car Left by car? bo car? _ car? the car?3 Ball-Book Ball Right _ ball? the ball? by ball? bo ball?4 Cat-Dog Dog Right the dog? by dog? bo dog? _ dog?5 Cup-Phone Cup Left _ cup? the cup? by cup? bo cup?6 Car-Shoe Shoe Right bo shoe? _ shoe? the shoe? by shoe?7 Ball-Book Book Left by book? bo book? _ book? the book?8 Cat-Dog Cat Right bo cat? _ cat? the cat? by cat?

Note. All sentences began with the sequence “Can you see. . .”.

6 Y. KEDAR ET AL.

Procedure

All parents were blind to the experimental design and were only given a general description of theprocedure. Following a 15-min play session, during which the infant was accustomed to the lab settingand to the experimenters, and parental consent was obtained, the child and the parent entered theexperimental room, where the actual test occurred. Parents wore darkened sunglasses to avoid affectingtheir infant’s looking patterns. Infants were seated on their parent’s lap so that their eyes directly faced thecenter of three TVmonitors. Before the experimenter left the experimental room, parents were remindedto remain neutral and to avoid talking to their infant during the entire test session.

Infants were randomly assigned to one of four lists of sentences. In each list there were eight sentences,with two sentences corresponding to each of the test conditions (i.e., the, by, bo, null). Order of thedifferent four conditions was quasi-randomized across lists (see Table 2). Each infant was presented witheight consecutive test cycles in which a familiarization trial was followed by a baseline trial and then a testtrial. Each Familiarization-Baseline-Test cycle lasted about 40 sec, with the entire testing session lastingaround 6 min. As illustrated in Table 3, a flashing, chiming, green circle serving as an attention-getterappeared on the center monitor prior to beginning each test cycle. During familiarization, each imagewas presented separately three times on the center monitor, accompanied by its linguistic label, for 3 sec(e.g., “ball”, “book”, “ball”, “book”, “ball”, “book”). Thus, each image was presented for a total duration of9 s. During the subsequent baseline and test trials, the same two images were presented simultaneouslyon the left and right monitors. During baseline trials, infants first heard the recorded voice encouragetheir looking to the images (“Look! Look at these!” or “Look! Look at that!” counterbalanced across allbaseline trials) and then viewed the pair of images for 4 sec. After 2 sec, infants heard “wow!” to maintaintheir attention. During test trials, a test sentence was heard, followed by a presentation of the same pair ofimages for 4 sec, with each of the images placed in the same positions as in the preceding baseline trial(i.e., the left and right monitors). A 270 msec pause occurred in baseline and test trials from sentence endto first appearance of the images on the monitors. Following the first four test sessions (Cup-Phone, Car-Shoe, Ball-Book, Cat-Dog), infants viewed the same pairs again in the same order. Each image in a pairserved as the target image in one test trial and as the distractor in the other test trial in which that pair ofimages appeared. To control for side preference, the side in which the target image appeared was quasi-randomized across test sessions in the following order: left, left, right, right, left, right, left, and right.Therefore, infants who were in fact orienting to the target images based on the linguistic input wouldoften have to shift their gaze from one side to the other across the different image pairs. To avoid arecency bias during the baseline and test trials (based on whether the last image seen during familiariza-tion was the target image or the distractor)—the order in which the images were presented duringfamiliarization was quasi-randomized as well. Images seen last during familiarization trials were pre-sented in the following order: target, distractor, distractor, target, distractor, target, target, and distractor.Therefore, there was no direct relation between which image appeared last during familiarization andwhich image served as target during the subsequent baseline and test trials. After the experiment hadended, parents were asked to estimate their infant’s familiarity with each of the eight test nouns. Parentsalso reported whether their child produced any of the test nouns and/or other words, including FWs.

Coding

The videotaped test sessions were coded off-line using the SuperCoder software (Hollich, 2003). Thisprogram creates a 30-frame-per-second transcript of the session and allows a frame-by-frameanalysis of the infants’ looking behavior. Specifically, in the familiarization trials, infants’ lookingbehavior was coded during the 3 sec presentation of each image (x 3). During baseline and test trials,infants’ looking behavior was coded from first appearance of the two images until these imagesdisappeared. From a total of 272 test cells, there were 36 test cells from 16 different infants that werenot included in the analyses due to fussiness or non-attentiveness during a baseline trial and/or thesubsequent test trial. An experimenter who was blind to the experimental conditions coded all test

LANGUAGE LEARNING AND DEVELOPMENT 7

sessions. A second experimenter coded nine of the 34 test sessions (five 18-month-olds, four 12-month-olds). The average correlation between the two observers indicated high inter-observerreliability: 0.94 (range: 0.90–0.99).

Results

Parental reports

As Table 1 demonstrates, the parental reports estimating their 18-month-old infants’ familiarity withthe nouns used in the study indicated high familiarity levels: Mean = 86.46%; SD = 10.85%; Range = 75–100%. As expected, the parental reports for the 12-month-olds showed relatively lower and morevariable familiarity levels: Mean = 66.55%; SD = 19.4%; Range = 30.77–92.31%. In terms of speech

Table 3. Schematic illustration of the experimental procedure.

Trial Audio Left Monitor Center Monitor Right Monitor

Attention Getter(flashes)

Bell chiming

Familiarization(Each image + nounrepeated 3 times)

“book”

“ball”

Baseline “Look! Look atthese!”

“Wow!”

Test “Can you see theball?”

8 Y. KEDAR ET AL.

production, the 18-month-olds were reported to produce 47.73% of the test nouns on average(SD = 19.28%; Range = 18.18–81.82%). In regards to those infants’ general productive vocabulary,the mean number of words reported to be produced by the 18-month-old infants was 20, none of whichwere reported to be FWs (only one infant was reported to occasionally use a single FW, “that”). In the12-month age group, parents reported their infants to produce only 22.32% of the test nouns on average(SD = 20.36%; Range = 0.00–50.00%). The 12-month-old infants’ average productive vocabulary was 5according to the parental reports, none of which were FWs (as with the older age group, there was oneinstance of an infant who was reported to occasionally use “that”).

Attention ratio

Two attention ratio measures were used to estimate infants’ interest in the visual stimuli during thetest sessions. In familiarization trials, we calculated the time an infant looked towards the presentedimage, divided by the overall 9 sec during which this image appeared on the center monitor (i.e.,3 sec × 3). The mean overall attention ratio in both age groups suggests that infants were in factattentive enough to form an association between the noun labels and their visual representationsduring familiarization: In the 18-months group, the mean overall attention ratio was .81 (SD = .20);while in the 12-months group it was .69 (SD = .22). Similarly, in the baseline trials and the test trials,the attention ratio measure was based on the time an infant looked towards either image in a pair,divided by the overall 4 sec during which both images appeared on the left and right monitors. Theresults in both age groups indicate that the infants were in fact interested in the visual display: In the18-months age group, the overall mean attention ratio during baseline trials was .74 (SD = .18), and.64 during test trials (SD = .16). In the 12-months age group, the overall mean attention ratio duringbaseline trials was .76 (SD = .19), and .62 (SD = .21) during test trials.

Statistical analyses

The statistical analyses were completed using the IBM SPSS Statistics 20.0 software. In the statisticalanalyses mentioned henceforth, there was no significant interaction involving Sex nor was there amain effect for Sex. For this reason, Sex was omitted as a variable in subsequent analyses.

Latency to target

To test for the effects of Function Word and Age on Latency to Target, we ran a linear mixed-effectsmodel (LMM) which enables to fit linear mixed-effects models to data sampled from normaldistributions. Specifically, we used the Restricted Maximum Likelihood (REML) method, whichyields asymptotically efficient parameter estimators for unbalanced data, as was the case in ourexperiment (i.e., the two age groups were unequal in size, 16 vs. 18 infants; and some of the cellswere missing, that is, we did not get two measurements for each infant on each FW type). Our LMManalysis included participant (child) as the random effect (random subject intercepts); and theSatterthwaite approximation method for DF calculation. Following a log transformation (justifiedby a residual analysis), the analysis yielded a statistically significant effect for Function Word, F(3,204) = 5.80, p = .001, overall. As Figure 1 demonstrates, infants in both age groups respondedsimilarly to the visual stimuli and oriented to the target image faster following grammaticalsentences. No interaction between Function Word and Age, F (3,204) = .29, ns; or a main effectfor Age, F (1,32) = 1.06, ns., were found. Pairwise comparisons among the four FW conditionsyielded significant overall differences between the grammatical condition and each of the ungram-matical conditions: THE vs. BY, t (204) = 2.00, p < .05; THE vs. BO, t (205) = 3.70, p < .001; THE vs.NULL, t (202) = 3.46, p = .001. No significant differences among the ungrammatical conditions werefound.

LANGUAGE LEARNING AND DEVELOPMENT 9

Although we found no interaction with Age, we analyzed each age group separately. In the 18-months age group, a statistically significant effect for Function Word, F (3,204) = 3.83, p = .01 wasfound. A priori, theoretically motivated pairwise comparisons, revealed significant differencesbetween THE vs. BO, t (206) = 3.09, p < .01, and between THE vs. NULL, t (201) = 2.75, p < .01.The comparison between THE vs. BY approached significance, t (204) = 1.86, p = .06. No significantdifferences among the ungrammatical conditions were found.

However, for infants of 12 months, there was no significant effect for Function Word, F(3,204) = 2.01, ns. Nonetheless, specific pairwise comparisons were significant: THE vs. BO, t(204) = 2.11, p < .05; THE vs. NULL, t (203) = 2.12, p < .05. The comparison between THE vs.BY was not significant, t (204) = .94, ns. No significant differences among the ungrammaticalconditions were found.

First look to target

As Figure 2 demonstrates, in both age groups, the grammatical condition initiated more correct firstlooks in comparison to each of the ungrammatical conditions. Due to the binary nature of theoutcome (YES/NO), we used a Generalized Estimating Equation (GEE) logistic regression analysis(Liang & Zeger, 1986). We fitted the logistic regression in the framework of the GEE to account forthe personal differences among the infants. Testing for the effects of Function Word and Age oncorrect first look, the GEE analysis yielded only a significant effect of Function Word, x2 (3) = 20.14,p < .001 overall, hence replicating the Latency findings. No interaction between Function Word andAge, c5 (3) = .14, ns; or a main effect for Age, x2 (1) = .04, ns—were found. Pairwise comparisonsamong the four FW conditions yielded significant differences between THE vs. BO, x2 (1) = 9.52,p < .01; THE vs. NULL, x2 (1) = 18.70, p < .001; and between BY vs. NULL, x2 (1) = 7.11, p < .01. Thecomparison between THE vs. BY was not significant, x2 (1) = .79, ns.

Figure 1. Infants’ mean latency (±Std error) in seconds to the target image during test trials, as a function of Function Word(the/by/bo/null).

10 Y. KEDAR ET AL.

Again, although no interaction with Age was found, we analyzed the two age groups separately. Inthe 18-months age group, a significant effect for Function Word, x2 (3) = 8.62, p < .05, was found.Pairwise comparisons revealed significant differences between THE vs. BO, x2 (1) = 6.48, p < .05, andbetween THE vs. NULL, x2 (1) = 7.32, p < .01. The comparison between THE vs. BY was notsignificant, x2 (1) = .75, ns. No significant differences among the ungrammatical conditions werefound.

Within 12 months alone, a significant effect of Function Word, x2 (3) = 14.42, p < .01, was found.Pairwise comparisons revealed a significant difference between THE vs. NULL, x2 (1) = 12.99,p < .001; and between BY vs. NULL, x2 (1) = 5.49, p < .05. The comparison between THE vs. BOapproached significance, x2 (1) = 3.74, p = .055. The comparison between THE vs. BY was notsignificant, x2 (1) = .16, ns.

Discussion

This study replicates earlier findings with 18- and 24-month-old infants, confirming that previousresults were not solely due to acoustic properties of the stimuli, which were controlled in our newdesign. In this way, they suggest that some linguistic competence guides early discrimination and useof determiners in addition to acoustic sensitivities that the child may have. Recall that in this studywe only used consonant-initial (CV) FWs in both the grammatical and ungrammatical conditionsalike (the/by/bo). Thus, unlike Kedar et al. (2006) and Kedar (2009), in which the results could beaided by the fact that only the ungrammatical FWs (and, el) were vowel-initial (and consequentlyharder to process than the consonant-initial “the”), the set of FWs used in our study rules out suchan acoustic-based explanation for our findings. Nonetheless, infants distinguished grammaticalsentences from ungrammatical ones. These results may be interpreted as suggesting that even atthese early ages, some form of prototypical grammatical sensitivity to particular FWs may existindependent of their acoustic variation; and that the developing processes of (content) wordlearning, sentence processing, and determination of the semantic function of noun reference—aremediated to some degree by children’s integration of FWs such as determiners in their representa-tion of language.

Figure 2. Infants’ mean percentage of correct first look (+/- Std error) to the target image during test trials, as a function ofFunction Word (the/by/bo/null).

LANGUAGE LEARNING AND DEVELOPMENT 11

In addition, the current findings reveal continuity in infants’ developing language competence,already at 12 months of age and long before FWs are integrated in the child’s productive speechovertly, to distinguish sentences that only differ in a single English FW. The overall failure of ouranalyses to find a significant effect of age in the combined analysis of the 12 and 18 months groupsacross measures (neither Latency to Target nor First Look); the clear pattern similarities across agegroups (seen in Figures 1 and 2); and the significant within-age pairwise comparisons betweengrammatical and ungrammatical function words—all point to a basic continuity in these data.

At the same time, our results also point to developmental processes that occur during the12–18 months period. In general, these findings and those reported by Kedar et al. (2006) illustratethat some gradual development is taking place from 12–24 months of age in terms of attentiveness,lexical familiarity, and speed of processing. For example, the latency analyses in both studies suggestthat younger infants take more time to orient to target across all four conditions; and that foryounger infants, lexical familiarity and productive vocabulary are limited in comparison to older agegroups. More specifically, at 12 months of age, a significant effect of FW appears only in first look,not in latency to target. This result may be a function of infants’ overall slower latencies, combinedwith the fact that we lowered our stimulus presentation time to 4 sec to accommodate for thereduced attention span at 12 months. This may have counteracted a known increased slowness inprocessing at this age (e.g., Vihman, Nakai, DePaolis, & Hallé, 2004; Zhang et al., 2015). Theemerging literature on FW processing around 12 months of age points to infant sensitivity to subtledifferences in task, which relate to basic differences in processing timing. Thus, one challenge in thisline of inquiry now is to calibrate comparisons between specific age group around the 1-year markerwith regard to known variations in latency across development.

Furthermore, the pairwise comparisons showed that infants’ response to “by” did not seem to beconsistently distinct from “the” (compared to the clear distinction between “the” vs. “bo” and “null”).While in the Latency to Target analysis the comparison between “the” and “by” was statisticallysignificant, the within group latency analyses failed to yield a significant effect on this contrast. Interms of correct first look, the same comparison did not reach significance in any of the analysesperformed. Interestingly, the comparison between “by” and “null” was significant in the general FirstLook analysis as well as the 12-month first look analysis.

One possible explanation for this set of results is based on integration between two differentcomputational mechanisms. First, infants’ well-documented excellence in acoustic processing as wellas their sensitivity to transitional probabilities and co-occurrence patterns in their linguistic input asthey compute sentences (e.g., Cartwright & Brent, 1997; Chemla, Mintz, Bernal, & Christophe, 2009;Gerken, Wilson, & Lewis, 2005; Lany & Gomez, 2008; Mintz, 2006; Mintz, Newport, & Bever, 2002;Saffran, 2001). In particular, this study—as well as previous findings (see the “null” condition resultsin Kedar et al., 2006; Kedar, 2009; Petretic & Tweney, 1977; and Shipley et al., 1969)—demonstratesthat English-learning infants notice the omission of pre-nominal functors. That is, infants seem todevelop a general expectation for an article to precede a (count) noun; an expectation that ispresumably based on the transitional probability of such co-occurrences in spoken English.

Second, it has been shown that frequent FWs become part of the child’s lexicon earlier than less-frequent FWs, and play a more robust role in aiding infants in extracting novel lexical items fromcontinuous speech input. For example, Shi and Lepage (2008) found that the high-frequency Frenchfunctors “des” and “mes”—as opposed to a nonsense FW, “kes”; or an infrequent FW, “vos”—facilitated the phonetic segmentation of subsequent nouns by French-learning 8-month-olds (seealso Shi et al., 2006a, 2006b). Such bias towards higher-frequency and simple-structured lexicalitems, as found in word segmentation studies, also seems to be the case in infants’ incorporation ofFWs in sentence processing and word categorization (e.g., Höhle et al., 2004). Thus, it is possible thatby 12 months, English-learning infants may have already formed a form-meaning competence toNoun Phrases that contain the high-frequency FW “the”, but may not have had sufficient exposureto less-frequent Noun Phrases with “by” in order to integrate them in deriving the meaning andvisual reference of a Noun Phrase. Furthermore, note that “by”, as opposed to “the”, has a wide

12 Y. KEDAR ET AL.

variation in syntactic contexts (e.g, “by the way. . .”); therefore, is not only less frequent but also morevariable in terms of the syntactic and semantic contexts in which it may appear.

Given the substantial cross-linguistic variation in the typology of FWs and the FCs to which theybelong (Kayne, 2005), it would also make sense that a child learning a specific language must workout the lexical distinctions between different FWs and their corresponding syntactic and semanticroles. Thus, the delay in a contrast between two English FWs, “the” vs. “by” relative to othercontrasts such as “bo” (presumably recognized as “Not a FW”׳ in English) may suggest a period oflexical learning.

In general, our findings stand in contrast to a long held view according to which infants and evenmuch older children initially attend only to content words in the speech stream (e.g., Bowerman,1973; Grimshaw, 1981; Pine & Lieven, 1997; Radford, 1997; Schlesinger, 1971; Tomasello, 2000a)—and correspond with the large emerging literature that has challenged this earlier view on two maingrounds. First, although FWs are often absent from children’s first utterances in English-speakingcommunities, the timing and manner in which earliest productions of FWs occur cross-linguisticallyvary considerably.1 Second, in perception, infants have been found to possess fine-detailed phono-logical representations for both function and content words already during the first year of life;2 andto incorporate FWs in sentence computation during their second year. The current findings point todevelopmental continuity in this respect between 12 and 18 months of age (and beyond: See Gerken& McIntosh, 1993; Kedar et al., 2006).

More broadly, our results strengthen the view according to which infants’ early language proces-sing—much like adult speakers—is facilitated when function and content words are structurallycombined at the phrasal and sentential levels, rather than operating on a telegraphic, single-wordbasis (e.g., Aslin, Woodward, La Mendola, & Bever, 1996; Fernald & Hurtado, 2006; Gomez &LaKusta, 2004; Hallé et al., 2008; Höhle et al., 2004, 2006; Kedar et al., 2006; Soderstrom et al., 2007).Note that although the infants we tested could have simply ignored the pre-nominal FW and insteadlocated the visual target by attending to its corresponding noun—the looking differences across theexperimental conditions indicate that the infants consulted particular FWs which appeared (or wereabsent, “null”) in each of the test sentences. This was the case even though they could form a word-object association based solely on lexical labels (e.g., “book”). Instead, the infants’ looking patternssuggest that they had processed the grammatical structure of the test sentences, in particular, theDeterminer Phrase (DP)—and then used this information to facilitate the establishment of nounreference in their visual environment.

Children’s knowledge or processing of FC structure is clearly not fully available at 12 months.For example, Shady (1996) found that infants’ sensitivity to the exact sentential location andorder in which certain FWs may appear develops between 10 and 16 months of age. Another lineof work has shown progression during the first 2 years in category-based abstraction of FWs (e.g.,Gerken et al., 2005; Gomez & Lakusta, 2004). As noted above, other studies have revealeddevelopmental processes in the differentiation of FCs (e.g., Höhle et al., 2004; Shi & Melançon,2010). In fact, it may take children years to fully acquire their language-specific functional lexiconand link it to semantics as well as to syntax (e.g., the relatively late discrimination betweendefinite vs. indefinite articles by English-learners; Foley, Lust, Battin, Koehne, & White, 2000;Schaffer & Matthewson, 2005).

Still, our findings suggest that infants consulted a specific determiner (word; “the”) representingthe Determiner (category; DET) in determining reference; that this process is not solely determinedby acoustic factors such as coarticulation effects or phonetic property of syllable onset; and that it

1In fact, studies using various research strategies have provided evidence that children’s omission of FWs—when present—ispredominantly based on limitations in speech processing, planning and production, which are phonological in nature (e.g., Boyle& Gerken, 1997; de Lange et al., 2009; Demuth, 1992; Dye, 2011; Gerken, 1994; Gerken et al., 1990; Kirk & Seidl, 2004; Lust, 2006;Valian et al., 2008).

2Hence, it is unlikely that the obtained responses were based on infants’ rote memorization of holistic combinations of “the” andeach of the test nouns (e.g., “theball”, “thebook”).

LANGUAGE LEARNING AND DEVELOPMENT 13

appears to play a continuous role in this way by about 12 months and beyond. That is, our results donot provide empirical evidence for a qualitative change in infants’ incorporation of FWs in sentenceprocessing. Our findings thus cohere with the hypothesis that FWs and their corresponding FCsprovide the child with a syntactic skeleton that serves a crucial role in parsing the speech stream intoits phrasal constituents, as well as in linking between syntax and semantics—in line with theSyntactic Bootstrapping Hypothesis (Fisher, Gleitman, & Gleitman, 1991; Gerken, 2007; Hirsh-Pasek & Golinkoff, 1996; Landau & Gleitman, 1985).

In sum, this study suggests that language development between 12 and 18 months lies in thechild’s integration of parallel acoustic, distributional and grammatical knowledge. We have foundthat 12-month-olds store familiar words and their referents in such a way that involves computationof and referral to particular FWs. This means that at least some FWs must enter into the linguisticdata that the infant is accessing and storing. In particular, the DP seems to link syntax andsemantics, even before well-developed combinatorial language production. Although there is muchabout acquisition of FWs such as determiners yet to be developed subsequently (both semanticallyand grammatically), this early ability appears to be foundational.

Acknowledgments

We would like to thank Lou Ann Gerken and two anonymous reviewers for their most valuable suggestions on thedesign of this study and on earlier drafts of the manuscript. We thank Tywanquila Walker for her contribution insetting up the experiment in HABIT and training the RAs on data collection; Charlene Bohanon for recording theauditory stimuli; and Katherine Callaghan, Kelly Kubas, and Kristin Marrese for their help in testing and recording theinfants and performing reliability checks. Finally, we are extremely thankful to all the infants and their parents forparticipating in this study.

Funding

A National Science Foundation Award (BCS-0721238) and a PECASE Award (BCS-0349183) to Marianella Casasolasupported this research.

References

Abney, S. (1987). The English noun phrase in its sentential aspect (Ph.D. dissertation). Massachusetts Institute ofTechnology, Cambridge, MA.

Aslin, R. (1999). Utterance-final bias in word recognition by 8-month-olds. Poster presented at the SRCD biennialmeeting, Albuquerque, NM.

Aslin, R. N., Woodward, J. Z., La Mendola, N. P., & Bever, T. G. (1996). Models of word segmentation in fluentmaternal speech to infants. In J. L. Morgan, & K. Demuth (Eds.), Signal to syntax: Bootstrapping from speech togrammar in early acquisition. Mahwah, NJ: Lawrence Erlbaum Associates.

Bernal, S., Dehaene-Lambertz, G., Millotte, S., & Christophe, A. (2010). Two-year-olds compute syntactic structure on-line. Developmental Science, 13(1), 69–76. doi:10.1111/desc.2009.13.issue-1

Bernal, S., Lidz, J., Millotte, S., & Christophe, A. (2007). Syntax constrains the acquisition of verb meaning. LanguageLearning and Development, 3, 325–341. doi:10.1080/15475440701542609

Bobb, S. C., & Mani, N. (2013). Categorizing with gender: Does implicit grammatical gender affect semanticprocessing in 24-month-old toddlers? Journal of Experimental Child Psychology, 115, 297–308. doi:10.1016/j.jecp.2013.02.006

Bowerman, M. (1973). Early syntactic development: A cross-linguistic study with special reference to Finnish.Cambridge, England: Cambridge University Press.

Boyle, M. K., & Gerken, L. A. (1997). The influence of lexical familiarity on children’s function morpheme omissions:A nonmetrical effect? Journal of Memory and Language, 36, 117–128. doi:10.1006/jmla.1996.2478

Cartwright, T. A., & Brent, M. R. (1997). Early acquisition of syntactic categories: A formal model. Cognition, 63, 121–170. doi:10.1016/S0010-0277(96)00793-7

14 Y. KEDAR ET AL.

Cauvet, E., Limissuri, R., Millotte, S., Skoruppa, K., Cabrol, D., & Christophe, A. (2014). Syntactic context constrainslexical access in French 18-month-olds. Language Learning and Development, 10(1), 1–18. doi:10.1080/15475441.2012.757970

Chemla, E., Mintz, T. H., Bernal, S., & Christophe, A. (2009). Categorizing words using ‘frequent frames’: What cross-linguistic analyses reveal about distributional acquisition strategies. Developmental Science, 12(3), 396–406.doi:10.1111/desc.2009.12.issue-3

Chomsky, N. (1995). The minimalist program for linguistic theory. Cambridge, MA: MIT Press.Christophe, A., Millotte, S., Bernal, S., & Lidz, J. (2008). Bootstrapping lexical and syntactic acquisition. Language &

Speech, 51, 61–75. doi:10.1177/00238309080510010501Cohen, L. B., Atkinson, D. J., & Chaput, H. H. (2004). Habit X: A new program for obtaining and organizing data in

infant perception and cognition studies. University of Texas, Austin, Texas.Curtin, S., Mintz, T. H., & Byrd, D. (2001). Coarticulatory cues enhance infants’ recognition of syllable sequences in

speech. In A. H. J. Do, L. Dominguez, & A. Johansen (Eds.), BUCLD 25 proceedings (pp. 190–201). Sommerville,MA: Cascadilla Press.

Cyr, M., & Shi, R. (2013). Development of abstract grammatical categorization in infants. Child Development, 84, 617–629. doi:10.1111/cdev.2013.84.issue-2

Dale, P. S., & Fenson, L. (1996). Lexical development norms for young children. Behavior research methods.Instruments & Computers, 28, 125–127. doi:10.3758/BF03203646

Demuth, K. (1992). Accessing functional categories in Sesotho: Interactions at the morpho-syntax interface. In J.Meisel (Ed.), The acquisition of verb placement: Functional categories and V2 phenomena in language development(pp. 83–107). Dordrecht, the Netherlands: Kluwer Academic Publishers.

De Lange, J., Vasic, N. and Avrutin, S. (2009). Reading between the (head)lines: a processing account of articleomission in newspaper headlines and child speech. Lingua, 119, 1523–1540.

Demuth, K. (1994). On the ‘underspecification’ of functional categories in early grammars. In B. Lust, M. Suñer, & J.Whitman (Eds.), Syntactic theory and first language acquisition: Cross-linguistic perspectives (pp. 119–134). Hillsdale,NJ: Lawrence Erlbaum Associates.

DePaolis, R. A., Vihman, M., & Keren-Portnoy, T. (2014). When do infants begin recognizing familiar words insentences? Journal of Child Language, 41(1), 226–239. doi:10.1017/S0305000912000566

Dye, C. D. (2011) Reduced auxiliaries in early child language: Converging observational and experimental evidencefrom French. Journal of Linguistics, 47(2), 301–339. doi: 10.1017/S002222671000037X.

Fenson, L., Dale, P. S., Reznick, J. S., Bates, E., Thal, D., & Pethick, S. (1994). Variability in early communicativedevelopment. Monographs of the Society for Research in Child Development, 59(5), 1–173.

Fernald, A., & Hurtado, N. (2006). Names in frames: Infants interpret words in sentence frames faster than words inisolation. Developmental Science, 9, F33–F40. doi:10.1111/desc.2006.9.issue-3

Fisher, C., Gleitman, H., & Gleitman, L. (1991). On the semantic content of subcategorization frames. CognitivePsychology, 23, 331–392. doi:10.1016/0010-0285(91)90013-E

Foley, C., Lust, B., Battin, D., Koehne, A., & White, K. (2000). On the acquisition of an indefinite determiner: Evidencefor unselective binding. BUCLD 24 proceedings (pp. 286–298). Boston, MA.

Gerken, L. A. (1994). A metrical template account of children’s weak syllable omissions. Journal of Child Language, 21,565–584. doi:10.1017/S0305000900009466

Gerken, L. A. (2007). Acquiring linguistic structure. In E. Hoff, & M. Shatz (Eds.), Blackwell handbook of languagedevelopment. Oxford, UK: Blackwell Publishing.

Gerken, L. A., & McIntosh, B. J. (1993). Interplay of function morphemes and prosody in early language.Developmental Psychology, 29, 448–457. doi:10.1037/0012-1649.29.3.448

Gerken, L. A., Remez, R. E., & Landau, B. (1990). Function morphemes in young children’s speech-perception andproduction. Developmental Psychology, 26, 204–216. doi:10.1037/0012-1649.26.2.204

Gerken, L. A., Wilson, R., & Lewis, W. (2005). 17-month-olds can use distributional cues to form syntactic categories.Journal of Child Language, 32, 249–268. doi:10.1017/S0305000904006786

Gervain, J., Nespor, M., Mazuka, R., Horie, R., & Mehler, J. (2008). Bootstrapping word order in prelexical infants: AJapanese-Italian cross-linguistic study. Cognitive Psychology, 57, 56–74. doi:10.1016/j.cogpsych.2007.12.001

Golinkoff, R., Hirsh-Pasek, K., Cauley, K., & Gordon, L. (1987). The eyes have it: Lexical and syntactic comprehensionin a new paradigm. Journal of Child Language, 14, 23–46. doi:10.1017/S030500090001271X

Golinkoff, R., Hirsh-Pasek, K., & Schweisguth, M. (2001). A reappraisal of young children’s knowledge of grammaticalmorphemes. In J. Weissenborn, & B. Höhle (Eds.), Approaches to bootstrapping: Phonological, lexical, syntactic andneurophysiological aspects of early language acquisition (pp. 167–188). Amsterdam, the Netherlands: JohnBenjamins.

Golinkoff, R. M., Ma, W., Song, L., & Hirsh-Pasek, K. (2013). Twenty-five years using the intermodal preferentiallooking paradigm to study language acquisition: What have we learned? Perspectives on Psychological Science, 8,316–339. doi:10.1177/1745691613484936

LANGUAGE LEARNING AND DEVELOPMENT 15

Gomez, R. L., & LaKusta, L. (2004). A first step in form-based category abstraction by 12-month-old infants.Developmental Science, 7, 567–580. doi:10.1111/desc.2004.7.issue-5

Grimshaw, J. (1981). Form, function, and the language acquisition device. In C. L. Baker, & J. J. McCarthy (Eds.), Thelogical problem of language acquisition. Cambridge, MA: MIT Press.

Hallé, P., Durand, C., & de Boysson-Bardies, B. (2008). Do 11-month-old French infants process articles? Languageand Speech, 51, 23–44. doi:10.1177/00238309080510010301

Hirsh-Pasek, K., & Golinkoff, R. (1996). The intermodal preferential looking paradigm reveals emergent languagecomprehension. In D. McDaniel, C. McKee, & H. Cairns (Eds.), Methods for assessing children’s’ syntax. Cambridge,MA: MIT Press.

Höhle, B., Schmitz, M., Santelmann, L. M., & Weissenborn, J. (2006). The recognition of discontinuous verbaldependencies by German 19-month-olds: Evidence for lexical and structural influences on children’s early proces-sing capacities. Language Learning and Development, 2, 277–300. doi:10.1207/s15473341lld0204_3

Höhle, B., & Weissenborn, J. (2003). German-learning infants’ ability to detect unstressed closed-class elements incontinuous speech. Developmental Science, 6, 122–127. doi:10.1111/1467-7687.00261

Höhle, B., Weissenborn, J., Kiefer, D., Schulz, A., & Schmitz, M. (2004). Functional elements in infants’ speechprocessing: The role of determiners in the syntactic categorization of lexical elements. Infancy, 5, 341–353.doi:10.1207/s15327078in0503_5

Hollich, G. (2003). SuperCoder (Version 1.5) [Computer software]. Retrieved from http://hincapie.psych.purdue.edu/Splitscreen/SuperCoder.dmg

Johnson, E. K., & Jusczyk, P. W. (2001). Word segmentation by 8-month-olds: When speech cues count more thanstatistics. Journal of Memory and Language, 44, 548–567. doi:10.1006/jmla.2000.2755

Katz, N., Baker, E., & MacNamara, J. (1974). What’s in a name? A study of how children learn common and propernames. Child Development, 45, 469–473. doi:10.2307/1127970

Kayne, R. S. (2005). Movement and silence. New York, NY: Oxford University Press.Kedar, Y. (2009). The role of functional categories in early language acquisition. Saarbruecken, Germany: VDM Verlag

Dr. Müller.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, 325–338. doi:10.1111/cdev.2006.77.issue-2Kirk, C., & Seidl, A. (2004). Perception and production of unfooted syllables: Implications for lexical representations.

28th BUCLD proceedings (pp. 318–327). Boston, MA.Kuhl, P. K., Ramírez, R. R., Bosseler, A., Lin, J. L., & Imada, T. (2014). Infants’ brain responses to speech suggest

analysis by synthesis. Pnas, 111, 11238–11245. doi:10.1073/pnas.1410963111Landau, B., & Gleitman, L. (1985). Language and experience: Evidence from the blind child. Cambridge, MA: Harvard

University Press.Lany, J., & Gomez, L. R. (2008). Twelve-month-old infants benefit from prior experience in statistical learning.

Psychological Science, 19(12), 1247–1252. doi:10.1111/j.1467-9280.2008.02233.xLiang, K. Y., & Zeger, S. (1986). Longitudinal data analysis using generalized linear models. Biometrika, 73(1), 13–22.

doi:10.1093/biomet/73.1.13Lust, B. (2006). Child language: Growth and acquisition. Cambridge, UK: Cambridge University Press.Maratsos, M. (1982). The child’s construction of grammatical categories. In E. Wanner, & L. R. Gleitman (Eds.),

Language acquisition: The state of the art. Cambridge, UK: Cambridge University Press.Marquis, A., & Shi, R. (2012). Initial morphological learning in preverbal infants. Cognition, 122, 61–66. doi:10.1016/j.

cognition.2011.07.004Mintz, T. H. (2006). Finding the verbs: Distributional cues to categories available to young learners. In K. Hirsh-Pasek,

& R. Golinkoff (Eds.), Action meets word: How children learn verbs. New York, NY: Oxford University Press.Mintz, T. H., Newport, E. L., & Bever, T. G. (2002). The distributional structure of grammatical categories in speech to

young children. Cognitive Science, 26, 393–424. doi:10.1207/s15516709cog2604_1Morgan, J. L. (1986). From simple input to complex grammar. Cambridge, MA: MIT Press.Nazzi, T., Dilley, L. C., Jusczyk, A. M., Shattuck-Hufnagel, S., & Jusczyk, P. W. (2005). English-learning infants’

segmentation of verbs from fluent speech. Language & Speech, 48, 279–298. doi:10.1177/00238309050480030201Petretic, P., & Tweney, R. (1977). Does comprehension precede production? The development of children’s responses

to telegraphic sentences of varying grammatical adequacy. Journal of Child Language, 4, 56–73. doi:10.1017/S0305000900001604

Pine, J., & Lieven, E. (1997). Slot and frame patterns in the development of the determiner category. AppliedPsycholinguistics, 18, 123–138. doi:10.1017/S0142716400009930

Pollock, J. Y. (1989). Verb movement, Universal Grammar, and the structure of IP. Linguistic Inquiry, 20, 365–424.Pruden, S. M., Hirsh-Pasek, K., Golinkoff, R. M., & Hennon, E. A. (2006). The birth of words: Ten-month-olds learn

words through perceptual salience. Child Development, 77, 266–280. doi:10.1111/cdev.2006.77.issue-2Radford, A. (1997). Syntactic theory and the structure of English. Cambridge, UK: Cambridge University Press.

16 Y. KEDAR ET AL.

Robertson, E. K., Shi, R., & Melançon, A. (2012). Toddlers use the number feature in determiners during online nouncomprehension. Child Development, 83, 2007–2018. doi:10.1111/cdev.2012.83.issue-6

Saffran, J. R. (2001). Words in a sea of sounds: The output of infant statistical learning. Cognition, 81, 149–169.doi:10.1016/S0010-0277(01)00132-9

Santelmann, L., & Jusczyk, P. (1998). Sensitivity to discontinuous dependencies in language learners: Evidence forlimitations in processing space. Cognition, 69, 105–134. doi:10.1016/S0010-0277(98)00060-2

Schaeffer, J., & Matthewson, L. (2005). Grammar and pragmatics in the acquisition of article systems. NaturalLanguage and Linguistic Theory, 23, 53–201. doi:10.1007/s11049-004-5540-1

Schlesinger, I. M. (1971). Production of utterances in language acquisition. In D. I. Slobin (Ed.), The ontogenesis ofgrammar (pp. 63–101). San Diego, CA: Academic Press.

Seidl, A., & Johnson, E. (2006). Infant word segmentation revisited: Edge alignment facilitates target extraction.Developmental Science, 9, 565–573. doi:10.1111/desc.2006.9.issue-6

Seidl, A., & Johnson, E. (2008). Perceptual factors influence infants’ extraction of onsetless words from continuousspeech. Journal of Child Language, 35, 1–24.

Shady, M. E. (1996). Infants’ sensitivity to function morphemes (Doctoral dissertation). The State University of NewYork: University at Buffalo, Buffalo, NY.

Shady, M. E., & Gerken, L. A. (1999). Grammatical and caregiver cues in early sentence comprehension. Journal ofChild Language, 26, 163–175. doi:10.1017/S0305000998003730

Shafer, V. L., Shucard, D. W., Shucard, J. L., & Gerken, L. A. (1998). An electrophysiological study of infants’sensitivity to the sound patterns of English speech. Journal of Speech Language and Hearing Research, 41, 874–886. doi:10.1044/jslhr.4104.874

Shi, R. (2007). Infants recognition of function words in continuous speech. In The Proceedings of the 16th InternationalCongress of Phonetic Science. Saarbrücken, Germany.

Shi, R. (2014). Functional morphemes and early language acquisition. Child Development Perspectives, 8, 6–11.doi:10.1111/cdep.12052

Shi, R., Cutler, A., Werker, J., & Cruickshank, M. (2006a). Frequency and form as determinants of functor sensitivityin English-acquiring infants. Journal of the Acoustical Society of America, 119(6), EL61–EL67.

Shi, R., Werker, J., & Cutler, A. (2006b). Recognition and representation of function words in English-learning infants.Infancy, 10, 187–198.

Shi, R., & Gauthier, B. (2005). Recognition of function words in 8-month-old French-learning infants. The Journal ofthe Acoustical Society of America, 117, 2426–2427. doi:10.1121/1.4786583

Shi, R., & Lepage, M. (2008). The effect of functional morphemes on word segmentation in preverbal infants.Developmental Science, 11, 407–413. doi:10.1111/j.1467-7687.2008.00685.x

Shi, R., & Melançon, A. (2010). Syntactic categorization in French-learning infants. Infancy, 15, 517–533. doi:10.1111/j.1532-7078.2009.00022.x

Shi, R., Werker, J., & Cutler, A. (2006). Recognition and representation of function words in English-learning infants.Infancy, 10, 187–198. doi:10.1207/s15327078in1002_5

Shi, R., Werker, J., & Morgan, J. (1999). Newborn infants’ sensitivity to perceptual cues to lexical and grammaticalwords. Cognition, 72, B11–B21. doi:10.1016/S0010-0277(99)00047-5

Shipley, E. F., Smith, C. S., & Gleitman, L. (1969). A study in the acquisition of language: Free responses to commandsin Language. Journal of the Linguistic Society of America, 45, 322–342. doi:10.2307/411663

Soderstrom, M., White, K. S., Conwell, E., & Morgan, J. L. (2007). Receptive grammatical knowledge of familiarcontent words and inflection in 16-month-olds. Infancy, 12, 1–29. doi:10.1111/j.1532-7078.2007.tb00231.x

Swingley, D. (2005). 11-month-olds’ knowledge of how familiar words sound. Developmental Science, 8, 432–443.doi:10.1111/desc.2005.8.issue-5

Thierry, G., Vihman, M., & Roberts, M. (2003). Familiar words capture the attention of 11-month olds in less than 250ms. NeuroReport, 18, 2307–2310. doi:10.1097/00001756-200312190-00004

Tomasello, M. (2000a). Do young children have adult syntactic competence? Cognition, 74, 209–253. doi:10.1016/S0010-0277(99)00069-4

Tomasello, M. (2000b). The item based nature of children’s early syntactic development. Trends in Cognitive Sciences,4, 156–163. doi:10.1016/S1364-6613(00)01462-5

Valian, V. (2008). Innateness and learnability. In E. L. Bavin (Ed.), The Cambridge Handbook of Child Language.Cambridge, UK: Cambridge University Press.

Valian, V. (2009). Innate syntax - still the best hypothesis. Plenary Address, The 34th Annual Boston UniversityConference on Language Development, Boston, MA.

Valian, V., Solt, S., & Stewart, J. (2008). Abstract categories or limited-scope formulae? The case of children’sdeterminers. Journal of Child Language, 35, 1–36.

Van Heugten, M., & Shi, R. (2009). French-learning toddlers use gender information on determiners during wordrecognition. Developmental Science, 12(3), 419–425. doi:10.1111/desc.2009.12.issue-3

LANGUAGE LEARNING AND DEVELOPMENT 17

Vihman, M. M., Nakai, S., DePaolis, R. A., & Hallé, P. (2004). The role of accentual pattern in early lexicalrepresentation. Journal of Memory and Language, 50(3), 336–353. doi:10.1016/j.jml.2003.11.004

Werker, J. F., & Yeung, H. H. (2005). Speech perception bootstraps word learning in infancy. Trends in CognitiveScience, 9, 519–527. doi:10.1016/j.tics.2005.09.003

Yang, C. (2013). Ontogeny and phylogeny of language. Proceedings of the National Academy of Sciences, 110(16),6324–6327. doi:10.1073/pnas.1216803110

Yeung, H. H., Chen, L. M., & Werker, J. F. (2014). Referential labeling can facilitate phonetic learning in infancy. ChildDevelopment, 85(3), 1036–1049. doi:10.1111/cdev.2014.85.issue-3

Zangl, R., & Fernald, A. (2007). Increasing flexibility in children’s online processing of grammatical and noncedeterminers in fluent speech. Language Learning and Development, 3, 199–231. doi:10.1080/15475440701360564

Zhang, Z., Shi, R., & Li, A. (2015). Grammatical categorization in Mandarin-Chinese-Learning Infants. LanguageAcquisition, 22(1), 104–115. doi:10.1080/10489223.2014.943900

18 Y. KEDAR ET AL.