Semantic categorization: A comparison between deaf and hearing children

Available online at www.sciencedirect.com

Journal of Communication Disorders 43 (2010) 347–360

Semantic categorization: A comparison between deaf and hearing children Ellen A. Ormel a,b,*, Martine A.R. Gijsel d, Daan Hermans c, Anna M.T. Bosman b, Harry Knoors b,c, Ludo Verhoeven b a

Department of Linguistics, Centre for Language Studies, Radboud University Nijmegen, The Netherlands b Behavioural Science Institute, Radboud University Nijmegen, The Netherlands c Royal Dutch Kentalis, St. Michielsgestel, The Netherlands d Expertisecentrum Nederlands, Nijmegen, The Netherlands Received 5 April 2008; received in revised form 22 January 2010; accepted 20 March 2010

Abstract Learning to read is a major obstacle for children who are deaf. The otherwise significant role of phonology is often limited as a result of hearing loss. However, semantic knowledge may facilitate reading comprehension. One important aspect of semantic knowledge concerns semantic categorization. In the present study, the quality of the semantic categorization of both deaf and hearing children was examined for written words and pictures at two categorization levels. The deaf children performed better at the picture condition compared to the written word condition, while the hearing children performed similarly at pictures and written words. The hearing children outperformed the deaf children, in particular for written words. In addition, the results of the deaf children for the written words correlated to their sign vocabulary and sign language comprehension. The increase in semantic categorization was limited across elementary school grade levels. Learning outcomes: Readers will be able to: (1) understand several semantic categorization differences between groups of deaf and hearing children; (2) describe factors that may affect the development of semantic categorization, in particular the relationship between sign language skills and semantic categorization for deaf children. # 2010 Elsevier Inc. All rights reserved.

1. Introduction The fundamental role of phonology in word recognition and reading comprehension of hearing children and adults is accepted almost unequivocally (Bosman & de Groot, 1996; Bosman & van Hell, 2002; Unsworth & Pexman, 2003; Van Orden, 1987). Studies of the role of phonology in the reading of deaf children and adults, however, have produced mixed results. In cases of limited phonology use by deaf readers, it is assumed that semantic knowledge may provide critical reading support (e.g., Kyle & Harris, 2006). The important role of semantic knowledge may therefore be particularly true for deaf children and adults. After all, semantic organization is an integral part of language learning.

* Corresponding author at: Radboud University Nijmegen, Linguistics Department, Erasmusplein 1, 6525 HT, Nijmegen, The Netherlands. Tel.: +31 024 3616028; fax: +31 024 3611070. E-mail address: [email protected] (E.A. Ormel). 0021-9924/$ – see front matter # 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.jcomdis.2010.03.001

348

E.A. Ormel et al. / Journal of Communication Disorders 43 (2010) 347–360

For example, semantic knowledge is closely related to subsequent word-learning ability (Borovsky & Elman, 2006). Reversely, semantic knowledge may increase as a consequence of reading experience. One important part of semantic knowledge is the proficiency of semantic categorization (Jerger & Damian, 2005). This has been examined in terms of both superordinate/subordinate relations (e.g., the bee as part of the superordinate category of insects) and at the level of exemplar relations (e.g., both the bee and the mosquito are part of the category of insects). In fact, semantic categorization has been widely used to gain insight into the organization of the semantic memory system and the relevant categorical relations have been found to become increasingly elaborate during the first few years of elementary school (e.g., Lucariello, Kyratzis, & Nelson, 1992; Nguyen & Murphy, 2003). Not only knowledge of categorical relations but also the speed of access for this categorical information appear to affect the extraction of semantic information from memory, and also the extraction of semantic information while reading (Howell & Manis, 1986), because effective retrieval of semantic information and particularly the categorical associations between words is needed for fluent reading (e.g., Hagtvet, 2003; Howell & Manis, 1986). In line with this need for effective retrieval of semantic information and the categorical associations between words for fluent reading, several researchers have documented the associations between semantic categorization and word decoding (Howell & Manis, 1986; Vellutino, Scanlon, & Spearing, 1995). Moreover, in addition to associations between semantic categorization and word decoding, the associations between semantic categorization and reading comprehension have been studied. Ben-Dror, Bentin, and Frost (1995), for example, found children with poor reading comprehension to perform worse on a number of semantic categorization tasks than children with normal reading comprehension, which suggests an association between underlying categorical knowledge and reading comprehension (see also Vellutino, Fletcher, Snowling, & Scanlon, 2004; Vellutino et al., 1995). When Nation and Snowling (1998) found evidence of an association between semantic categorization skill and reading comprehension, they further suggested that poor reading comprehension may be due to the non-automatic access of underlying semantic information. Nation, Adams, Bowyer-Crane, and Snowling (1999) showed that 10% of the poor reading comprehenders had perfect phonological decoding skills but clear semantic difficulties. This finding is in line with the assumption that not just the phonological skills, but also the semantic skills of otherwise normally developing children are strongly related to their level of reading proficiency. 1.1. Deaf bilingual children Due in part to the limited phonological information available to them, many deaf children experience major difficulties learning to read (Hermans, Knoors, Ormel, & Verhoeven, 2008; Marschark, Lang, & Albertini, 2002; Ormel, Hermans, Knoors, & Verhoeven, in press; Perfetti & Sandak, 2000; Wauters, van Bon, & Tellings, 2006). Similarly, the development of semantic categorization is delayed for most deaf children due to a lack of full access to a language—including sign language—during the first few years of their lives (Chamberlain & Mayberry, 2008; Mayberry, 2002, 2006). This is particularly true for deaf children growing up in hearing families because the acquisition of a sign language appears to be a difficult task for the hearing parents and hearing teachers of deaf children (Fortgens, 2003; Marschark et al., 2002). Nevertheless, sign language is the more natural and accessible language for the vast majority of deaf children (Klatter-Folmer, van Hout, Kolen, & Verhoeven, 2006; Knoors, 2007). Moreover, semantic categorization may play a more important role in relation to reading comprehension of deaf children when compared to hearing children. As said, semantic categorical structure seems to be closely related to subsequent word-learning ability (Borovsky & Elman, 2006). The causal relationship between semantic categorization and reading skills could go in either direction. Information regarding deaf children’s semantic knowledge to date involves the frequent mentioning of poor vocabulary skills (Marschark et al., 2002). In addition to vocabulary knowledge, more knowledge is needed on semantic categorization in deaf people, in particular in deaf children. Only a limited number of studies examined semantic knowledge in deaf people, including semantic categorization knowledge (MacSweeney, Gossi, & Neville, 2004). This is especially true for young deaf children in bilingual education programs, who are just learning to read. A number of studies showed similar semantic categorization skills for deaf and hearing participants. When Courtin (1997) examined the semantic categorization skills of second generation deaf children who were fluent signers at the age of six, their performance on a semantic categorization task was found to be similar to that of hearing children, but clearly affected by sign language structures. In the seventies, Tweney, Hoemann, and Andrews (1975) already showed that deaf and hearing children view simple, highly familiar objects in similar ways, when asked to sort items into

E.A. Ormel et al. / Journal of Communication Disorders 43 (2010) 347–360

349

groups. Similarly, Liben (1979) found that deaf children used semantic clustering (both spontaneously and after instruction) in a free recall task with a list containing semantic categorically related items, as much as hearing children. However, lower overall recall scores by the deaf children could suggest inadequate knowledge of categories. More recently, in a semantic judgment task with deaf people who were exposed to Cued Speech (a mode of communication for visually conveying traditionally spoken languages at the phonemic level), D’Hondt and Leybaert (2003) found similar semantic accuracy results for deaf and hearing children. They also showed a similar brain lateralization pattern for the deaf and the hearing participants; a left hemisphere specialization for semantic processing of written language. In contrast to the studies showing similar patterns of semantic knowledge between deaf and hearing participants, several studies also revealed differences. Green and Shepherd (1975) studied knowledge of antonymous pairs of adjectives, such as good–bad and slow–fast, in both deaf and hearing children. They found a more restrictive semantic system in the deaf children. Likewise, the categorical organization of semantic information and automatic processing of this information have been shown to differ between deaf adults and that of hearing adults (MacSweeney et al., 2004; Marschark, Convertino, McEvoy, & Masteller, 2004; McEvoy, Marschark, & Nelson 1999). When automatic processing was required, in the use of a masked semantic priming task, MacSweeney et al. (2004) showed that deaf adults showed no behavioral semantic priming effect (e.g., for cat-dog) and also complementary ERP measures showed no N400 effects for deaf adults, whereas the hearing adults did show semantic effects (both in the behavioral and ERP measures). The results of the masked priming task showed less automatic semantic processing in the group of deaf adults when compared to hearing adults. These findings are in line with the results by Neville et al. (1998), who showed some differences in neural activity between deaf and hearing individuals during processing of written English sentences. However, the underlying semantic processing system involved seemed similar for deaf and hearing adults, which was examined by controlled reading of familiar words in an unmasked priming task, again complemented by using ERP measures. McEvoy et al. showed a more heterogeneous conceptual organization and weaker associative relations among concepts for deaf students when compared to hearing students. It was argued that this less coherently organized semantic knowledge may result in reduced support for comprehension and learning. Marschark et al. (2004) examined different types of categorical relationships and found an asymmetry in exemplar-category relations for deaf students but not for hearing students. Knowledge of category superordinates (e.g., animal or fruit) was not as readily available by deaf students as it was by hearing students. Deaf students correctly solved more analogies in which they had to provide a subordinate term (category member) than analogies in which they had to provide a superordinate term (category label), whereas the hearing children showed no difference between providing subordinate and superordinate terms. Part of the variability in the deaf students’ results is related to their reading skills. Deaf students who were better readers showed patterns closer to the hearing students than deaf students who had more limited reading skills. This finding supports the notion of the importance to study (and train) semantic categorization knowledge in deaf children. Holowka, Brosseau-Lapre, and Petitto (2002) studied bilingual hearing babies of deaf parents who learned sign language and French and showed that the semantic organization for these babies showed similar patterns to those of bilingual hearing babies who learned two spoken languages. In other words, the semantic differences found between deaf and hearing participants seem to be related to deafness, and not to the use of sign language. 1.2. Present study In several past studies, differences in semantic categorization between deaf and hearing participants were found for adults. In the present study, we looked further into the differences between bilingual deaf children and hearing children, who took part in two different semantic-categorizations tasks. The deaf children were being taught in bilingual education settings, which include Sign Language of the Netherlands in addition to (Sign Supported) Dutch. Sign Language of the Netherlands is a full-blown sign language with its own grammatical rules, whereas Sign Supported Dutch is a combination between Dutch and Sign Language of the Netherlands; the Dutch words are generally accompanied by signs. The semantic categorization of the deaf children was investigated in order to gain insight into this possible source for reading difficulties, in addition to the already assumed limited phonological knowledge in many deaf children (Miller, 2006; Waters & Doehring, 1990). Additionally, deaf and hearing children of different ages were included in the present study to enable us to examine the development of semantic categorization skills in both deaf and hearing children. One of the reasons why many deaf people showed weaker semantic categorization than others may lie in the use of written words to assess semantic categorization (see Marschark et al., 2004). Pictures may be easier to process than

350

E.A. Ormel et al. / Journal of Communication Disorders 43 (2010) 347–360

words for young readers or readers with otherwise limited reading skills (Vellutino, Scanlon, DeSetto, & Pruzek, 1981). Similarly, the severe decoding difficulties of many deaf children when confronted with written words—but expectedly not pictures—may severely impede assessment of their semantic categorical knowledge, at least when the child has received sufficient (sign) language input starting from a very young age. We therefore used pictures, in addition to written words, to assess the semantic skills of the deaf children and thereby avoid full reliance on their word-recognition skills. The stimuli were presented in two conditions to all children: (a) pictures and (b) written words. To further assess the organization of the children’s underlying semantic categorization knowledge, the semantic information was assessed using two levels of categorization and thus in two different experiments. In Experiment 1, the children’s semantic categorization was assessed at the level of the exemplar. In Experiment 2, the children’s semantic categorization was assessed at the level of the subcategory (i.e., superordinate and subordinate relations), in line with the division in different categorization levels as proposed by Marschark et al. (2004) and Borghi and Caramelli (2003). 2. Experiment 1: exemplar categorization In Experiment 1, the children’s semantic categorization was assessed at the level of the exemplar (e.g., both the bee and the mosquito are insects). The stimuli were presented in two conditions: (a) pictures, and (b) written words. The performances of the deaf and hearing children were then compared within the different conditions. In other words, whether or not the deaf children in the present study showed the same categorical exemplar knowledge as the hearing children was examined. The knowledge of the deaf children was then compared across the different conditions and different grade levels. We expected the hearing children to outperform the deaf children in the written words condition but not, or to a lesser extent, in the picture condition. Given the reading difficulties experienced by most deaf children, the deaf children were expected to encounter larger difficulties in the written word condition relative to the picture condition. 2.1. Method 2.1.1. Participants Deaf and hearing elementary school students from grades three through six participated in the present study. The deaf children constituted the experimental group with 39 boys and 20 girls from three different schools for deaf education in the Netherlands (n = 59). The mean age for the deaf children ranged from 103 months (SD = 8 months) in grade three to 149 months (SD = 6 months) in grade six. Of the 59 deaf children, 24 were in grades three and four (i.e., the lower grades hereafter) and 35 were in grades five and six (i.e., the upper grades hereafter). The mean age for the 24 children in the lower grades was 116 months (SD = 7.6 months) and the age range was between 105 and 127 months. The mean age for the 37 children in the upper grades was 137 months (SD = 6.9 months) and the age range was between 127 and 150 months. Forty children performed a test for nonverbal intelligence (NVIQ Raven-CPM), which provided a global estimation of their intelligence. Given that NVIQ scores were only available for 40 of the deaf children it cannot be assumed that the other children were in the normal range. However, this is expectedly the case. The Raven-CPM consists of 36 items. From this test we can conclude that the deaf children performed within the range of a normal intelligence. Fourteen of the deaf children had a cochlear implant.1 Four of the deaf children had deaf parents.2 The schools all provided bilingual deaf education and support for children with auditory and communicative difficulties, and the curriculum involved alternation between Sign Language of The Netherlands (SLN) and (Sign Supported) Dutch (SSD). 1

The number of deaf children who enrol in mainstream education programs in the Netherlands is still growing. Most of these children who enrol in mainstream education have cochlear implants. Some studies have shown that many of these children have acquired reasonably good production and comprehension skills in spoken Dutch, and acquire higher levels of literacy in comparison to deaf children without cochlear implants (e.g., Vermeulen, van Bon, Schreuder, Knoors, & Snik, 2007). The deaf children in bilingual education programs who have cochlear implants are not a representative sample of all the deaf children with cochlear implants in the Netherlands. As a consequence, it is not really informative to investigate the language skills of deaf children with and without a cochlear implant in the present study. 2 We conducted an ANOVA when the four deaf children of deaf parents were excluded in order to see if the significant semantic effects remained the same. This was indeed the case.

E.A. Ormel et al. / Journal of Communication Disorders 43 (2010) 347–360

351

The hearing children constituted the control group with 42 boys and 46 girls from two regular elementary schools in the Netherlands (n = 88). The mean age for the hearing children ranged from 102 months (SD = 6.5 months) in grade three to 152 months (SD = 6.5 months) in grade six. Of the 88 hearing children, 41 were in the lower grades and 47 were in the upper grades of elementary school. The mean age for the 41 children in the lower grades was 115 months (SD = 6.8 months) and the age range was between 103 and 126 months. The mean age for the 47 children in the upper grades was 137 months (SD = 6.8 months) and the age range was between 123 and 153 months. A number of variables could arguably affect the results for the deaf children. Sign language fluency is one of these possibly influencing variables (e.g., Mayberry, 2002). As part of sign fluency, one can distinguish sign vocabulary knowledge and sign language comprehension as important components, amongst other sign language skills. Moreover, all of the children came from hearing families and grew up in homes where spoken Dutch was the primary language used. Several studies have shown that family hearing status may affect the deaf child’s language development (e.g., Mayberry, 2002). The same is true for language preference of the child. Four of the children in the present study had deaf family members. Language variables are available for 41 of the participating deaf children. The scores on the semantic tests were compared for the deaf children who preferred to use sign language, the deaf children who preferred sign supported Dutch, and the children who preferred spoken language in the classroom, based on their teachers’ judgments. Language preference in the classroom was defined by four categories: (1) Sign language (9 children), (2) Sign Supported Dutch (18 children), (3) equal preference for Sign language and Sign supported Dutch (8 children), and (4) equal preference for Speech and Sign Supported Dutch (5 children). In all analyses, age was controlled for. ANOVA’s were carried out in General Linear Model to compare groups of children with different language preferences. Neither of the semantic tests revealed differences between the groups (F < 1 for pictures and written words). Note that group sizes are rather small. Subsequently, in order to assess the extent to which our results may be influenced by these arguably affecting variables, correlations were calculated between sign vocabulary (a combined score for passive and active vocabulary performance, Hermans, Knoors, & Verhoeven, 2010) and the semantic tests, and also between sign language comprehension (a combined score for passive sign language comprehension performance, Hermans et al., 2010) and the semantic tests (see Sections 2.2.2.2 and 2.2.2.3). 2.1.2. Materials The materials presented in the two exemplar categorization conditions involved either pictures or written words. In both conditions, four pictures were presented for the respondent to select that picture which best matches the stimulus. The words in the exemplar experiment were matched for log frequencies per million words, number of letters, and number of neighbor words using the CELEX counts (Baayen, Piepenbrock, & van Rijn, 1993). The majority of the stimuli were one syllable CVC, CCVC, CVCC, or CCCVC words. The pictures were taken from the Dutch Leesladder (i.e. ‘‘Reading Ladder’’), which is a computer program for children with reading disabilities (Irausquin & Mommers, 2001). The pictures in the exemplar experiment were colored line-drawings and represented nouns that could be assumed to be familiar to most 6-year-old Dutch children (Schaerlaekens, Kohnstamm, & Lejaegere, 1999; familiarity rating  .80 along a scale of 0–1). In addition, only high-imageability words were selected for use (Van LoonVervoorn, 1985; imageability rating > 5.5 along a seven-point scale). Detailed descriptions of the conditions in Experiment 1 are presented below. 2.1.2.1. Pictures. This condition consists of 20 experimental trials, preceded by three practice trials. A target picture is presented (e.g., a picture of an orange) and then four pictures consisting of the target response and three distracters are presented simultaneously. The target response represents a concept from the same taxonomic category (e.g., a picture of a cherry) and was the correct answer. The semantic categories of the included target pictures and target responses (the correct answer) were: Insects, predators, mammals, rodents, reptiles, sense organs, vegetables, fruit, furniture, transport, clothes, jewels, body parts, kit/tools, toys, and buildings. The different trials involved exemplars from one of these different semantic categories. Each of these semantic categories was used once, with the exception of the categories of vegetables, fruits, body parts, and buildings which were used in two items. One of the distracter pictures was a semantic distracter and was included for half of the experimental items: A semantic distracter was a concept (e.g., egg) that belongs to a higher superordinate category (e.g., food) and not, thus, the category represented by the target stimulus (e.g., fruit in the case of a picture of an orange). A second distracter picture is a phonological distracter such as ‘‘beer’’ for the target stimulus ‘‘ear’’ (used on 11 of the 20 trials) or a perceptual distracter such as

352

E.A. Ormel et al. / Journal of Communication Disorders 43 (2010) 347–360

‘‘ball’’ for the target stimulus ‘‘orange’’ (used on 9 of the 20 trials). The criterion for phonological similarity was sharing the end-rime with the stimulus. Perceptual similarity was created in terms of similar contours and colors. A fourth type of distracter picture was an unrelated picture such as ‘‘chair’’ for the target stimulus ‘‘orange’’. An unrelated picture was used with all of the stimuli. For half of the items with no semantic distracter, two unrelated pictures were included among the response options. The criterion for the unrelated picture was absence of a semantic (i.e., taxonomic or associative) relation, perceptual relation, or phonological similarity. The accuracy of responding was measured. The items in the picture condition had an average of 4.6 letters, 14.5 neighbor words (i.e., words differing 1 letter from the target word), and a log frequency of 1.29. 2.1.2.2. Written words. This condition consisted of 20 experimental trials, preceded by three practice trials. The stimuli were drawn from the same categories as in the picture condition. The written word stimuli were presented as written letter strings. The response options were again four pictures presented simultaneously. The exact stimuli and response options differed from those used in the picture condition. The types of distracters and their frequencies were identical to those in the picture condition. The accuracy of responding was measured. And the items in this condition had an average of 4.6 letters, 13.9 neighbor words, and a log frequency of 1.29. 2.1.2.3. Sign vocabulary knowledge. The sign vocabulary test was developed and administered as part of the development of an assessment instrument for SLN (Hermans et al., 2010). In the vocabulary test, children saw a sign that was followed by four pictures. Children were instructed to select the picture that matched the sign in meaning. The vocabulary task consisted of 61 items. 2.1.2.4. Sign language comprehension. The sign language comprehension test was developed and administered as part of the development of the same assessment instrument for SLN (Hermans et al., 2010). In the story comprehension task, children saw five stories in SLN, one at a time. After each of these stories, children had to answer four questions about the story. Some of these questions referred to information literally given in the stories. Other questions were either so-called gap-filling or text-connecting questions (Cain & Oakhill, 1999). 2.1.3. Apparatus and procedure All of the conditions were implemented in E-prime (Schneider, Eschman, & Zuccolotto, 2002), which is a psychology software tool. The trials were presented on a laptop. First, a fixation stimulus (i.e., a ‘‘+’’ presented using a 50-point Times New Roman font) was presented on the screen for 1000 ms. Immediately thereafter, the target stimulus (i.e., picture or letter string, depending on the two different conditions) was presented. The target stimulus remained visible until the participant pressed the space bar. Thereafter, the four response options (i.e., pictures) appeared on the laptop screen simultaneously and the participant had to decide which picture best matched the target stimulus, and belonged to the same semantic category. The participants made a response by using one of four keys on the laptop keyboard. The position of the four keys corresponded to the (horizontal) position of the response pictures on the screen (i.e., the keys ‘‘c,’’ ‘‘b,’’ ‘‘m,’’ and ‘‘.’’ were used, corresponding to Pictures 1, 2, 3 and 4 on the screen). The relevant keys on the laptop were marked with white stickers. The participant was asked to keep his or her hands in front of the keyboard in between their responses. The word stimuli were presented in white on a black background in the center of the screen in a lowercase 24-point Courier New font. Similarly, the pictures were presented on a black background in the center of the screen. There was a 1500 ms delay between responding by the participant and the onset of the next trial. For each of the participants, the items were presented in a different (i.e., random) order. The children were instructed in sign language to choose one picture out of four pictures of which they thought it belonged to the same group (for example things or animals) as the word or picture presented first. Several examples of the experiment then provided additional insight into the experiment. After the examples, the children were asked if the instructions were clear to them. The children performed the matching tasks in groups of 6–8 students. The order of administration for the conditions was varied across participants such that each condition appeared an equal number of times as first or second. Those children tested at the same time received the same order of conditions but not the same order of items within a given condition. The children were administered both conditions within a single session. The children were instructed to press the space bar after reading the target word when performing the written word matching task. Once the space bar was pressed in the written word matching task as well as in the picture matching

E.A. Ormel et al. / Journal of Communication Disorders 43 (2010) 347–360

353

task, the four response options (i.e., pictures) were presented simultaneously on the screen. The accuracy of the matching response provided by the child was recorded. 2.2. Results The first set of analyses was aimed at answering the question of whether deaf children show equally specific semantic categorization as hearing children for different ages (i.e., in both grades 3/4 and grades 5/6), and under different matching conditions (i.e., pictures or written words as stimuli). The second set of analyses concerned the question of whether the deaf children showed differences in semantic skill when the stimuli were pictures or written words and in the lower versus higher elementary grades. 2.2.1. Semantic categorization of the deaf and hearing children In the first set of analyses, univariate analyses of variance and repeated measures were conducted, using General Linear Model in SPSS. In the presentation of the outcomes of the analyses, F-values refer to participant analyses. Grade (lower vs. upper) and hearing status (deaf vs. hearing) were treated as between-subjects’ factors, and condition (pictures vs. written words) as a within-subjects’ factors. As already mentioned the results for the deaf and hearing children were compared in each of the stimulus conditions separately. 2.2.1.1. Accuracy data. There was no three-way interaction between condition, grade, and hearing status (F > 1). Two-way interactions were found between condition and hearing status (F(1,134) = 12.74, p < .001), not for condition and grade (F < 1), or between grade and hearing (F < 1). Main effects were found of condition (F(1,134) = 24.26, p < .001), hearing status (F(1,134) = 58.55, p < .001), and grade (F(1,134) = 5.44, p < .05). The results for the individual conditions showed highly significant differences between the children in favor of the hearing children in both conditions: picture condition (F(1,138) = 19.32, p < .001; written word condition (F(1,142) = 82.52, p < .001). The largest general difference between the deaf and hearing children occurred in the written word condition (see Table 1), in favor of the hearing participants. In the Levene’s Test of Equality of Error significantly different standard deviations were found for the results of the deaf children versus the hearing children. The difference was largest for the written word test. 2.2.2. Modality and semantic skills of the deaf children The following part concerns deaf children only. In a repeated-measures analysis based upon a General Linear Model, condition was treated as within-subjects’ factor and grade level as between-subjects’ factor. 2.2.2.1. Accuracy data. In the analysis on the accuracy of the deaf children’s responding2, the interaction between condition and grade level was not significant (F(2,49) = 4.62, p < .05). In both grades, written words were responded to less accurately than pictures. A main effect of condition was found (F(1,48) = 20.43, p < .001), but not of grade level (F(1,48) = 1.98, p > 1). Responses in the picture condition were more accurate when compared to the written word condition. Table 1 Means and standard deviations for accuracy of responding for deaf versus hearing children in the two conditions from Experiment 1 (also according to grade level). Grade level Lower elementary Deaf (n = 24), accuracy

Upper elementary Hearing (n = 41), accuracy

Deaf (n = 35), accuracy

Hearing (n = 47), accuracy

Written word M 52.7 SD 20.5

76.8 9.6

57.9 19.7

80.6 11.3

Picture M SD

78.8 12.6

71.1 18.2

82.5 13.0

65.5 21.5

354

E.A. Ormel et al. / Journal of Communication Disorders 43 (2010) 347–360

2.2.2.2. Sign language comprehension test. The semantic written word test showed a marginal correlation with sign language comprehension scores when controlling for the children’s age (written words: r(41) = .27, p = .08; pictures: r(41) = .14, p = .37). 2.2.2.3. Sign vocabulary. One of the two semantic tests correlated with sign vocabulary knowledge when controlling for the children’s age: written words (written words: (r(41) = .36, p = .02; pictures: r(41) = .06, p = .69). 2.3. Conclusion In sum, the data showed largest differences between the deaf and hearing children in the written word condition, which might be expected. Responses in the other condition were also less accurate for the deaf children. For these children, performance in the picture condition was clearly easiest. Sign language skills correlated with the semantic performance, but only in the written word condition. 3. Experiment 2: superordinate categorization In Experiment 2, the deaf and hearing children’s knowledge of such superordinate/subordinate relations as a ‘‘bee’’ being part of the category of ‘‘insects’’ or a ‘‘chair’’ being part of the category of ‘‘furniture’’ was tested using written words. Pictures could not be used due to difficulties with the pictorial representation of superordinate/subordinate category information. Similar to Experiment 1, the accuracy of the deaf versus hearing children’s performances was first compared. We expected the hearing children to outperform the deaf children. The performance of the deaf children in the two grade levels was next examined. The deaf children were expected to encounter difficulties in the written word condition, in line with the often found reading difficulties for deaf children, and in contrast to the performance of the hearing children. 3.1. Method 3.1.1. Participants The participants were the same as in Experiment 1. The children with various language preferences in the classroom differed significantly in the written condition (F(1,30) = 4.13, p < 05). The best scores in the written condition were found for the group who had equal preference for Sign Language and for Sign Supported Dutch. Also in Experiment 2, correlations were calculated between sign vocabulary and the semantic test, and between sign language comprehension and the semantic test (see Sections 3.2.2.2 and 3.2.2.3). 3.1.2. Materials The children’s knowledge of taxonomic relations (i.e., superordinate categories) was tested using written words. 3.1.2.1. Written words. The condition consisted of 25 experimental trials, preceded by two practice trials. The name of a (superordinate) semantic category was presented as a written word, followed by the simultaneous presentation of four pictures representing the target response and three distracters. The target response involved depiction of a member of the previously presented superordinate category. The superordinate categories included: Residence, toys, jobs, transport, sports, pets, fruit, furniture, vegetables, mammals, numbers, and clothes. Each of the superordinate categories has one unambiguous sign. Of the 25 trials, 15 of the trials included two semantic distracters (i.e. response options that did not depict a subordinate member of the superordinate category mentioned just prior, but, however, a semantically related response option) and one completely unrelated response option. The remaining 10 trials included three unrelated response options. The criteria for the selection of the distracters were the same as in Experiment 1. The accuracy of the children’s responding was measured. The items had 5.1 letters on average, 11.4 neighbor words on average, and an average log frequency of 1.35. 3.1.2.2. Sign vocabulary and sign language comprehension. See Experiment 1.

E.A. Ormel et al. / Journal of Communication Disorders 43 (2010) 347–360

355

Table 2 Means and standard deviations for accuracy of responding for deaf versus hearing children in the written word condition from Experiment 2 (also according to grade level). Grade level Lower elementary Deaf (n = 24), accuracy Written word M 59.4 SD 16.7

Upper elementary Hearing (n = 41), accuracy

Deaf (n = 35), accuracy

Hearing (n = 47), accuracy

86.7 7.7

61.6 20.4

91.1 6.4

3.1.3. Apparatus and Procedure The same equipment and procedures were used as in Experiment 1. Evidently, the content of the instructions was somewhat different. The children were instructed to choose the picture out of four pictures of which they thought it belonged to the group (for example fruit or insects) as the word presented first. Several examples of the experiment then provided additional insight into what was meant by the instruction. After the instruction, the children were asked if the task was clear to them. 3.2. Results The first set of analyses concerned the question of whether significant differences exist in the semantic categorization of written words of the deaf versus hearing children (see Table 2). The possible differences between the deaf and hearing children depending on grade level were also examined. The second set of analyses concerned the question of whether the deaf children showed differences in their semantic categorization depending on grade level. 3.2.1. Semantic categorization of the deaf and hearing children Univariate analyses of variance and repeated measures were conducted in the participant analyses, using General Linear Model. Grade level and hearing status were treated as between-subjects’ factors. 3.2.1.1. Accuracy data. Analyses of the accuracy of the performance showed that no two-way interaction was found between hearing and grade (F < 1). A main effect was found for hearing (F(1,130) = 148.95, p < .001), and a marginal effect was found for grade (F(1,130) = 2.94, p < .1). Analyses of the accuracy of the performances of the deaf versus hearing showed the hearing children to consistently and clearly outperform the deaf children in the written word condition, and children in the higher grades performed slightly better than the children in the lower grades. 3.2.2. Semantic categorization of the deaf children In a repeated-measures analysis based upon a General Linear Model, grade level was treated as a between-subjects’ factor. 3.2.2.1. Accuracy data. The deaf children2 showed no effect of grade in the written words (F < 1). 3.2.2.2. Sign language comprehension test. The semantic test showed marginal correlations with sign language comprehension scores when controlling for the children’s age (written words: r(29) = .30, p = .098). 3.2.2.3. Sign vocabulary. The sign semantic test was significantly correlated with sign vocabulary knowledge when controlling for the children’s age (written words: r(29) = .41, p = .02). 3.3. Conclusion To summarize the results of Experiment 2, the deaf children found the written word condition involving taxonomic semantic relations to be more difficult than the hearing children did. The deaf children showed very similar accuracy

356

E.A. Ormel et al. / Journal of Communication Disorders 43 (2010) 347–360

levels across grade levels. The results of the semantic test at the superordinate level were related to sign language comprehension (marginally) and sign vocabulary. 4. General discussion The purpose of the present study was twofold. On the one hand, the semantic categorization of deaf children who were educated bilingually was compared to that of hearing children using different types of stimuli. On the other hand, the increase of those skills across grades was explored. In Experiment 1, the children’s knowledge of exemplar-level semantic relations was assessed. In Experiment 2, the children’s knowledge of superordinate-level semantics relations was assessed. In both experiments, the hearing children outperformed the deaf children. The differences between the deaf and hearing children were smallest in the picture condition from Experiment 1. The deaf children performed better in the picture condition in this experiment than in the written word condition. These results support the assumption that using written words to assess the semantic-categorization skills of deaf children may underestimate their knowledge. Nevertheless, the results of Experiment 1 also show the deaf children to lag behind same-age hearing children when pictures are used to assess their semantic categorization skills. That is, the semantic-categorical knowledge of deaf children appears to be less precise or less finely differentiated than the semantic-categorical knowledge of hearing children. Additionally, the results of the deaf children on the written words correlated to their sign vocabulary and sign language comprehension skills. Children who had large sign vocabularies were better at the semantic categorization of written words than children who had small sign vocabularies. The same pattern was found for children who had large sign language comprehension in comparison to children who had little sign language comprehension. 4.1. Factors that affect the development of semantic categorization Nation and Snowling (1998) suggested that the reading experience of poor readers may result in limited semantic knowledge. More recently, however, it has been suggested that reading difficulties can stem from semantic deficits but, at the same time, semantic deficits can stem from reading difficulties (see Vellutino et al., 2004). Impoverished reading experiences can restrict an otherwise rich semantic environment and thereby lead to reduced semantic knowledge and elaboration. Note, however, Vellutino et al. suggested that for all bilingual (including deaf) children and special populations, the reading difficulties may be the result of, rather than a cause of, semantic deficits. Limited reading experience resulting from poor semantic categorization may indeed hold for deaf children. Interactions in the home, at school, and in other settings can generally contribute to children’s semantic categorization but may be very different for deaf children as opposed to hearing children. That is, most deaf children grow up in hearing families in which their linguistic interactions are often limited. When parents learn to sign fluently, however, their deaf children are generally raised bilingually and their linguistic skills may be more proficient than for children who are raised in a monolingual, non-signing, environment. Deaf children who grow up in deaf families are usually part of an entire deaf culture and may thus experience very few limitations on and very different experiences with their linguistic interactions. Semantic categorization can be affected by both cultural and bilingual influences (Pen˜a, Bedore, & Zlatic-Giunta, 2002; Unsworth, Sears, & Pexman, 2005). The bilingual influences involve the fact that children in a bilingual language environment often learn two words for the same referent. The deaf children in the present study supposedly had access to both the sign and written Dutch word for a single referent. However, most bilingual deaf children only become bilingual after entrance into a bilingual deaf education setting around the age of three. The parents of most deaf children also still have to learn Sign Language after the discovery that their child is deaf, and this can take a considerable amount of time and have considerable consequences for the deaf child’s development (Mayberry, 2002). The deaf child’s semantic-categorical knowledge may certainly not emerge as it naturally does in hearing (bilingual) children. In the present study, even the oldest deaf children did not have their semantic-categorical knowledge organized in such a manner that they could accurately process the information conveyed not only by words, but also by pictures. In this light, investigation of the development of semantic skills in deaf children growing up with signing parents or caregivers who may or may not be deaf themselves is an important topic for future research. Moreover, given the important findings by Borovsky and Elman (2006) that semantic category structure seems to be closely related to subsequent word-learning ability, it seems essential to teach deaf children (in particular those who have hearing parents) semantic category structures in, for example, sign. The limited semantic results for the deaf children

E.A. Ormel et al. / Journal of Communication Disorders 43 (2010) 347–360

357

in the present study may actually stress the importance of high-quality basic (sign) vocabulary knowledge, which is needed in order to develop semantic knowledge, but additionally, it may stress the importance of explicit instructions of semantic relationships between concepts. Moreover, it can result in the reverse effect as well, whereby semantic knowledge allows for the development of a larger vocabulary. In other research, Courtin (1997) found that native-deaf signers did not perform worse than hearing controls on a semantic categorization task. Unfortunately, we do not know if the deaf signers were fluent readers or not. Whether or not Courtin’s findings reflect patterns that hold for older deaf children or adolescents with presumably improved signing skills is also a question that needs to be answered. It is also possible that the deaf children’s signing skills provide them with enhanced knowledge of categorical relations, which can subsequently transfer to their reading and writing skills. Nation and Snowling (1998) have suggested that impaired categorical knowledge may be mediated by poor phonological knowledge. If phonological knowledge is thus required or at least preferable to improve semanticcategorical knowledge, then deaf children may be at a disadvantage. Deaf children must possibly acquire semantic categorization information via other means. One possible limitation of the present study is that the relations between children’s semantic categorization and their reading proficiency are still unclear. It would be interesting to study the associations between semantics and reading under different conditions and for different grade levels. As yet, thus, it remains to be seen whether: (a) the development of reading skills promotes the semantic development of deaf children, (b) the development of semantic categorization contributes to the development of reading in deaf children, or (c) semantic categorization and reading possibly interact throughout the deaf child’s development. A second limitation of the present study is related to the possibility that pictures or words, which were presumably unrelated, were actually related at the level of sign phonology (hand shape, movement, location, and orientation). In the present study, we assured that sign phonology did not overlap to a large extent. However, we did not control for the presence of some overlapping parts (e.g., hand shape, location, or movement). Knowledge about the role of sign phonology in sign perception and production is very limited thus far, but is recently increasing (e.g., Dye & Shih, 2006; Emmorey & Corina, 1990; Ormel, Hermans, Knoors, & Verhoeven, 2009). In the future, a study such as this one could be controlled more precisely for the presence of sign distracters. A third limitation of the present study is the missing language background information for a third of the deaf children. Preferably, the conclusions on relationships between semantic categorization and sign language proficiency should have been based on all participants. To conclude, the present study showed limited semantic improvement across grade levels and rather low levels at the semantic tasks for the deaf children, in particular when using words. As can be expected from other studies, better reading can follow from improved semantic categorization. For that reason, attempts should be made to make semantic information more explicit to young deaf children. One starting point could be to help deaf children develop both large written and sign vocabularies. The next step, then, is to link either written or signed words together in such a manner that the underlying semantic relations are made more transparent. Such elaborated semantic networks can subsequently help the child learn to read more adequately and fluently, which can then—in turn—further promote the child’s semantic categorization. Deaf people may develop semantic networks that are different from conventional semantic networks, and this domain of inquiry therefore urgently requires further investigation. The importance of semantic categorization, however, should not be underestimated. Role of the funding source Financial support for this study was provided by Viataal and the Koninklijke Effatha Guyot Groep (part of the Royal Kentalis Group), as well as the Mgr. J. C. van Overbeekstichting. The funding sources agreed on the content of the present study. The Royal Kentalis Group was most closely involved in the actual design, collection, analyses, data interpretation, writing of the report, and the decision to submit the paper for publication. Acknowledgments We thank Jikke Planting, Corina Michielsen, Annelies Leechburch-Auwers who collected part of the data and assisted in creating the test materials. We are also grateful to all children that participated in the study and to the teachers who gave the permission to run the experiments.

358

E.A. Ormel et al. / Journal of Communication Disorders 43 (2010) 347–360

Appendix A. Continuing education Continuing education questions: 1. Do hearing children outperform deaf children at semantic categorization at the exemplar level and at the superordinate level? a. No, only at the exemplar level. b. No, only at the superordinate level. c. Yes, both at exemplar and superordinate level, but only shown in the lower grades. d. Yes, both at exemplar and superordinate level in the lower and in the upper grades. 2. Do deaf children perform equally as accurate on the written word condition at the exemplar level and the picture condition at the exemplar level at both grade levels? a. No, they showed better scores at the written word condition than the picture condition in the upper grades; in the lower grades the reverse pattern was found. b. No, they showed better scores at the picture condition than the written condition in the upper grades; in the lower grades the reverse pattern was found. c. No, they showed better scores at the picture condition than the written condition at both grade levels. d. Yes, they showed equal scores at the written word condition and the picture condition at both grade levels. 3. Do the deaf children show semantic improvements across grades at the exemplar level and the superordinate level? a. No, deaf children showed limited improvements, at the exemplar level and at the superordinate level. b. No, deaf children showed limited improvements at the exemplar level, however, improvements were significant at the superordinate level. c. No, deaf children showed limited improvements at the superordinate level, however, improvements were significant at the exemplar level. d. Yes, the deaf children showed significant improvements across grades, both at the exemplar level and at the superordinate level. 4. What do the authors argue when they suggest that semantic organization is an integral part of language learning? a. Semantic knowledge is somewhat related to subsequent word-learning ability. b. Semantic knowledge is closely related to subsequent word-learning ability. Reversely, semantic knowledge may also increase as a consequence of reading experience. c. Semantic knowledge may coincidentally increase in parallel to the acquisition of reading experience. d. Semantic knowledge may increase as a consequence of reading experience but the reverse, an increase of semantic knowledge as a consequence of reading experience, is not to be expected. 5. What family circumstances may affect social and linguistic interactions, and subsequently semantic categorization, according to the authors? a. Most deaf children grow up in deaf families in which the linguistic interactions are often limited. Interactions can contribute to semantic categorization. Also in the present study, most children grew up in deaf families, which may have affected the results. b. Most deaf children grow up in hearing families in which the linguistic interactions are often limited. Interactions can contribute to semantic categorization. In the present study, most children grew up in deaf families, which may have affected the results. c. Most deaf children grow up in deaf families in which the linguistic interactions are often limited. Interactions can contribute to semantic categorization. In the present study, most children grew up in hearing families, which may have affected the results. d. Most deaf children grow up in hearing families in which the linguistic interactions are often limited. Interactions can contribute to semantic categorization. Also in the present study, most children grew up in hearing families, which may have affected the results. References Baayen, R., Piepenbrock, R., & van Rijn, H. (1993). The CELEX Lexical Database. Technical report. Philadelphia, PA: Linguistic Data Consortium, University of Pennsylvania.

E.A. Ormel et al. / Journal of Communication Disorders 43 (2010) 347–360

359

Ben-Dror, I., Bentin, S., & Frost, R. (1995). Semantic, phonologic, and morphologic skills in reading disabled and normal children: Evidence from perception and production of spoken Hebrew. Reading Research Quarterly, 30, 876–893. Borghi, A. M., & Caramelli, N. (2003). Situation bounded conceptual organization in children: from action to spatial relations. Cognitive Development, 18, 49–60. Borovsky, A., & Elman, J. (2006). Language input and semantic categories: A relation between cognition and early word learning. Journal of Child Language, 33, 759–790. Bosman, A., & de Groot, A. (1996). Phonologic mediation is fundamental to reading: Evidence from beginning readers. The Quarterly Journal of Experimental Psychology, 49A(3), 715–744. Bosman, A. M. T., & van Hell, J. G. (2002). Orthography, phonology, and semantics. Concerted action in word perception. In L. Verhoeven, C. Elbro, & P. Reitsma (Eds.), Precursors of functional literacy (pp. 165–187). Amsterdam, the Netherlands: John Benjamins. Cain, K., & Oakhill, J. V. (1999). Inference making ability and its relation to comprehension failure. Reading and Writing, 11, 489– 503. Chamberlain, C., & Mayberry, R. I. (2008). American Sign Language syntactic and narrative comprehension in good and poor deaf readers: Bilingual-Bimodal evidence for the linguistic basis of reading. Applied Psycholinguistics, 39, 367–388. Courtin, C. (1997). Does sign language provide deaf children with an abstraction advantage? Evidence from a categorization task. Journal of Deaf Studies and Deaf Education, 2, 161–171. D’ Hondt, M., & Leybaert, J. (2003). Lateralization effects during semantic and rhyme judgement tasks in deaf and hearing subjects. Brain and Language, 87, 227–240. Dye, M. W. G., & Shih, S. (2006). Phonological priming in British Sign Language. In L. M. Goldstein, D. H. Whalen, & C. T. Best (Eds.), Laboratory phonology 8. Berlin, Germany: Mouton de Gruyter. Emmorey, K., & Corina, D. (1990). Lexical recognition in sign language: Effects of phonetic structure and morphology. Perceptual and Motor skills, 71, 1227–1252. Fortgens, C. (2003). Taalkeuze van dove kinderen. Language choice of deaf children]. Gouda: Koninklijke Auris groep. Green, W. B., & Shepherd, D. C. (1975). The semantic structure in deaf children. Journal of Communication Disorders, 8(4), 357–365. Hagtvet, B. E. (2003). Listening comprehension and reading comprehension in poor decoders: Evidence for the importance of syntactic and semantic skills as well as phonological skills. Reading and Writing: An Interdisciplinary Journal, 16, 505–539. Hermans, D., Knoors, H., Ormel, E., & Verhoeven, L. (2008). A model of reading vocabulary acquisition for deaf children in bilingual education programs. Journal of Deaf Studies and Deaf Education, 13(2), 155–174. Hermans, D., Knoors, H., & Verhoeven, L. (2010). Assessment of Sign Language development: The case of deaf children in the Netherlands. Journal of Deaf Studies and Deaf Education, Advance Access, 15(2), 107–119. Holowka, S., Brosseau-Lapre´, F., & Petitto, L. A. (2002). Semantic and conceptual knowledge underlying bilingual babies’ first signs and words. Language Learning, 52(2), 205–262. Howell, M. J., & Manis, F. R. (1986). Developmental and reader ability differences in semantic processing efficiency. Journal of Educational Psychology, 78(2), 124–129. Irausquin, R., & Mommers, C. (2001). Leesladder. Een programma voor kinderen met leesmoeilijkheden [Reading ladder. A program for children with reading difficulties]. Zwijsen. Jerger, S., & Damian, M. F. (2005). What’s in a name? Typicality and relatedness effects in children. Journal of Experimental Child Psychology, 92, 46–75. Klatter-Folmer, J., van Hout, R., Kolen, E., & Verhoeven, L. (2006). Language development in deaf children’s interactions with deaf and hearing adults: A longitudinal study. Journal of Deaf Studies and Deaf Education, 11(2), 238–251. Knoors, H. (2007). Educational responses to varying objectives of deaf parents of deaf children: A Dutch perspective. Journal of Deaf Studies and Deaf Education, 12(2), 243–253. Kyle, F., & Harris, M. (2006). Concurrent correlates and predictors of reading and spelling achievement in deaf and hearing school children. Journal of Deaf Studies and Deaf Education, 11(3), 273–288. Liben, L. S. (1979). Free recall by deaf and hearing children: Semantic clustering and recall in trained and untrained groups. Journal of Experimental Child Psychology, 27(1), 105–119. Lucariello, J., Kyratzis, A., & Nelson, K. (1992). Taxonomic knowledge: What kind and when? Child Development, 63, 978–998. MacSweeney, M., Grossi, G., & Neville, H. (2004). Semantic priming in deaf adults: An ERP study. In Proceedings at the Cognitive Neuroscience Society Annual Meeting. Marschark, M., Lang, H. G., & Albertini, J. A. (2002). Educating deaf students. From research to practice. New York: Oxford University Press. Marschark, M., Convertino, C., McEvoy, C., & Masteller, A. (2004). Organization and use of the mental lexicon by deaf and hearing individuals. American Annals of the Deaf, 149(1), 51–61. Mayberry, R. I. (2002). Cognitive development in deaf children: The interface of language and perception in neuropsychology. In Segalowitz, S. J., & Rapin, I. Eds. Handbook of neuropsychology, Part II. vol. 8 (pp.71–107). Amsterdam: Elsevier. Mayberry, R. I. (2006). Second language learning of sign languages. In B. Woll (Ed.), Sign Language. Encyclopedia of Language and Linguistics (2nd ed., pp. 743–746). Oxford: Elsevier. McEvoy, C., Marschark, M., & Nelson, D. L. (1999). Comparing the mental lexicons of deaf and hearing individuals. Journal of Educational Psychology, 91, 312–320. Miller, P. (2006). What the processing of real words and pseudohomophones can tell us about the development of orthographic knowledge in prelingually deafened individuals. Journal of Deaf Studies and Deaf Education, 11(1), 21–38. Nation, K., Adams, J. W., Bowyer-Crane, C. A., & Snowling, M. J. (1999). Working memory deficits in poor comprehenders reflect underlying language impairments. Journal of Experimental Child Psychology, 73, 139–158.

360

E.A. Ormel et al. / Journal of Communication Disorders 43 (2010) 347–360

Nation, K., & Snowling, M. J. (1998). Semantic processing and the development of word-recognition skills: Evidence from children with reading comprehension difficulties. Journal of Memory and Language, 39, 85–101. Neville, H. J., Bavelier, D., Corina, D., Rauschecker, J., Karni, A., Lalwani, A., et al. (1998). Cerebral organization for language in deaf and hearing subjects: Biological constraints and effects of experience. Proceedings of the National Academy of Sciences, 95, 922–929. Nguyen, S., & Murphy, G. L. (2003). An apple is more than just fruit: Cross-classification in children’s concepts. Child Development, 27, 1783– 1806. Van Orden, G. C. (1987). A ROWE is a ROSE: Spelling, sound, and reading. Memory and Cognition, 15, 181–198. Ormel, E., Hermans, D., Knoors, H., & Verhoeven, L. (2009). The role for sign phonology and sign iconicity during sign processing: The case of deaf children. Journal of Deaf Studies and Deaf Education, 14(4), 436–448. Ormel, E., Hermans, D., Knoors, H., & Verhoeven, L. (in press). Phonological activation during visual word recognition in deaf and hearing children. Journal of Speech, Language, and Hearing Reserach. Pen˜a, E. D., Bedore, L. M., & Zlatic-Giunta, R. (2002). Category-generation performance of bilingual children: The influence of condition, category, and language. Journal of Speech, Language, and Hearing research, 45, 938–947. Perfetti, C. A., & Sandak, R. (2000). Reading optimally builds on spoken language: Implications for Deaf readers. Journal of Deaf Studies and Deaf Education, 5(1), 32–50. Schaerlaekens, A., Kohnstamm, G. A., & Lejaegere, M. (1999). Streeflijst woordenschat voor zesjarigen: Derde herziene versie gebaseerd op nieuw onderzoek in Nederland en Belgie¨ [Target list vocabulary for 6-year-old children]. Lisse: Swets & Zeitlinger. Schneider, W., Eschman, A., & Zuccolotto, A. (2002). E-Prime. Psychology software tools. Pittsburgh: Learning Research and Development Centre, University of Pittsburgh. Tweney, R. D., Hoemann, H. W., & Andrews, C. E. (1975). Semantic organization in deaf and hearing subjects. Journal of Psycholinguistic Research, 4(1), 61–73. Unsworth, S. J., & Pexman, P. M. (2003). The impact of reader skill on phonological processing in visual word recognition. The quarterly Journal of Experimental Psychology, 56A(1), 63–81. Unsworth, S. J., Sears, C. R., & Pexman, P. M. (2005). Cultural influences on categorization processes. Journal of Cross-Cultural Differences, 36(6), 626–688. Van Loon-Vervoorn, W. A. (1985). Voorstelbaarheidswaarden vanNederlandse woorden [Imageability ratings of Dutch words]. Lisse, The Netherlands: Swets & Zeitlinger. Vellutino, F. R., Scanlon, D. M., DeSetto, L., & Pruzek, R. M. (1981). Developmental trends in the salience of meaning versus structural attributes of written words. Psychological Research, 43, 131–153. Vellutino, F. R., Fletcher, J. M., Snowling, M. J., & Scanlon, D. M. (2004). Specific reading disability (dyslexia): What have we learned in the past four decades? Journal of Child Psychology and Psychiatry, 45(1), 2–40. Vellutino, F. R., Scanlon, D. M., & Spearing, D. (1995). Semantic and phonological coding in poor and normal readers. Journal of Experimental Child Psychology, 59, 76–123. Vermeulen, A. M., van Bon, W., Schreuder, R., Knoors, H., & Snik, A. (2007). Reading comprehension of deaf children with cochlear implants. Journal of Deaf Studies and Deaf Education, 12(3), 283–302. Waters, G. S., & Doehring, D. G. (1990). Reading acquisition in congenitally deaf children who communicate orally: Insights from an analysis of component reading, language, and memory skills. In T. H. Carr & B. A. Levy (Eds.), Reading and its development (pp. 323–373). San Diego: Academic Press. Wauters, L. N., van Bon, W., & Tellings, A. (2006). Reading comprehension of Dutch deaf children. Reading and Writing, 19, 49–76.