Isabelle Darcy

Direct mapping of acoustics to phonology: On the lexical encoding of front rounded vowels in L1 English– L2 French acquisition

It is well known that adult US-English-speaking learners of French experience difficulties acquiring high /y/–/u/ and mid /œ/–/ɔ/ front vs. back rounded vowel contrasts in French. This study examines the acquisition of these French vowel contrasts at two levels: phonetic categorization and lexical representations. An ABX categorization task (for details, see Section IV) revealed that both advanced and intermediate learners categorized /œ/ vs. /ɔ/ and /y/ vs. /u/ differently from native speakers of French, although performance on the /y/–/u/ contrast was more accurate than on the /œ/–/ɔ/ contrast in all contexts. On a lexical decision task with repetition priming, advanced learners and native speakers produced no (spurious) response time (RT) facilitations for /y/–/u/ and /œ/–/ɔ/ minimal pairs; however, in intermediate learners, the decision for a word containing /y/ was speeded by hearing an otherwise identical word containing /u/ (and vice versa), suggesting that /u/ and /y/ are not distinguished in lexical representations. Thus, while it appears that advanced learners encoded the /y/–/u/ and /œ/–/ɔ/ contrasts in the phonological representations of lexical items, they gained no significant benefit on the categorization task. This dissociation between phonological representations and phonetic categorization challenges common assumptions about their relationship and supports a novel approach we label ‘direct mapping from acoustics to phonology’ (DMAP).

The role of the striatum in phonological processing: evidence from early stages of Huntington's disease.

The linguistic role of subcortical structures such as the striatum is still controversial. According to the claim that language processing is subdivided into a lexical memory store and a computational rule system (Pinker, 1999) several studies on word morphology (e.g., Ullman et al., 1997) and on syntax (e.g., Teichmann et al., 2005) have suggested that the striatum is specifically dedicated to the latter component. However, little is known about whether the striatum is involved in phonological operations and whether its role in linguistic rule application generalizes to phonological processing. We investigated this issue by assessing perceptual compensation for assimilation rules in a model of striatal disorders, namely in the early stages of Huntingtons disease (HD). In Experiment 1 we used a same-different task with isolated words to evaluate whether phoneme perception is intact in HD. In Experiment 2 a word detection task in phrasal contexts allowed for assessing both phoneme perception and perceptual compensation for the French regressive assimilation rule. Results showed that HD patients have normal performance with both phoneme perception in isolated words and regressive assimilation rules. However, in phrasal contexts they display reduced abilities of phoneme discrimination. These findings challenge the striatum-rule claim and suggest a more fine-grained function of striatal structures in linguistic rule processing. Alternative explanatory frameworks of the striatum-language link are discussed.

Exploring dyslexics’ phonological deficit II: Phonological grammar

Language learners have to acquire the phonological grammar of their native language, and different levels of representations on which the grammar operates. Developmental dyslexia is associated with a phonological deficit, which is commonly assumed to stem from degraded phonological representations. The present study investigates one aspect of the phonological grammar, phonological assimilation rules. Specifically, it examines whether dyslexic adults have acquired phonological rules in speech production normally, and whether they compensate for them in perception. Contrary to the ‘degraded phonological representations hypothesis’, these adults produced phonological assimilations, and perceptually compensated for assimilations to the same extent as control participants. This suggests that individuals with dyslexia have acquired the phonological rules of their native language normally, and implies that they must have well-specified phonological representations, at least with respect to the features tested here. Nevertheless, these dyslexic adults still exhibit the typical phonological deficit as measured by phonological awareness, verbal short-term memory and rapid automatic naming tasks. Thus, it is suggested that the explanation for their phonological deficit must lie elsewhere than in their phonological representations and grammar.

Modality-independent neural mechanisms for novel phonetic processing.

The present study investigates whether the inferior frontal gyrus is activated for phonetic segmentation of both speech and sign. Early adult second language learners of Spanish and American Sign Language at the very beginning of instruction were tested on their ability to classify lexical items in each language based on their phonetic categories (i.e., initial segments or location parameter, respectively). Conjunction analyses indicated that left-lateralized inferior frontal gyrus (IFG), superior parietal lobule (SPL), and precuneus were activated for both languages. Common activation in the left IFG suggests a modality-independent mechanism for phonetic segmentation. Additionally, common activation in parietal regions suggests spatial preprocessing of audiovisual and manuovisual information for subsequent frontal recoding and mapping. Taken together, we propose that this frontoparietal network is involved in domain-general segmentation of either acoustic or visual signal that is important to novel phonetic segmentation.

Lexical function of pitch in the first language shapes cross-linguistic perception of Thai tones

Abstract Determining the factors involved in the non-native perception of the pitch patterns of tones is complicated by the fact that all languages use pitch to various extents, whether linguistic (e.g., intonation) or non-linguistic (e.g., singing). Moreover, many languages use pitch to distinguish lexical items with varying degrees of functional load and differences in inventory of such pitch patterns. The current study attempts to understand what factors determine accurate naïve (= non-learner) perception of non-native tones, in order to establish the baseline for acquisition of a tonal L2. We examine the perception of Thai tones (i.e., three level tones, two contour tones) by speakers of languages on a spectrum of lexically contrastive pitch usage: Mandarin (lexical tone), Japanese (lexical pitch accent), English (lexical stress), and Korean (no lexically contrastive pitch). Results suggest that the importance of lexically contrastive pitch in the L1 influences non-native tone perception so that not all non-tonal language speakers possess the same level of tonal sensitivity, resulting in a hierarchy of perceptual accuracy. Referencing the Feature Hypothesis (McAllister et al. 2002), we propose the Functional Pitch Hypothesis to model our findings: the degree to which linguistic pitch differentiates lexical items in the L1 shapes the naïve perception of a non-native lexically contrastive pitch system, e.g., tones.

The Effect of Orthography on the Lexical Encoding of Palatalized Consonants in L2 Russian

The current study investigated the potential facilitative or inhibiting effects of orthography on the lexical encoding of palatalized consonants in L2 Russian. We hypothesized that learners with stable knowledge of orthographic and metalinguistic representations of palatalized consonants would display more accurate lexical encoding of the plain/palatalized contrast. The participants of the study were 40 American learners of Russian. Ten Russian native speakers served as a control group. The materials of the study comprised 20 real words, familiar to the participants, with target coronal consonants alternating in word-final and intervocalic positions. The participants performed three tasks: written picture naming, metalinguistic, and auditory word–picture matching. Results showed that learners were not entirely familiar with the grapheme–phoneme correspondences in L2 Russian. Even though they spelled almost all of these familiar Russian words accurately, they were able to identify the plain/palatalized status of the target consonants in these words with about 80% accuracy on a metalinguistic task. The effect of orthography on the lexical encoding was found to be dependent on the syllable position of the target consonants. In intervocalic position, learners erroneously relied on vowels following the target consonants rather than the consonants themselves to encode words with plain/palatalized consonants. In word-final position, although learners possessed the orthographic and metalinguistic knowledge of the difference in the palatalization status of the target consonants—and hence had established some aspects of the lexical representations for the words—those representations appeared to lack in phonological granularity and detail, perhaps due to the lack of perceptual salience.

Experience with computerized speech‐perception training (SPATS‐ESL) for speakers of other languages learning English.

SPATS‐ESL evolved as it was used by volunteers and as a supplement to classes in the Intensive English and English Enhancement Programs at Indiana University (Bloomington). These trials with eighty ESL‐learners representing 12 L1s resulted in the curriculum described in the companion poster. Before training, the ESL learners exhibited significant problems with perceiving spoken English. These were greatest for syllable nuclei, least for syllable onsets, and intermediate for syllable codas. Recognition of spoken sentences was also inferior to that of native speakers of English. Improvement on all tasks including sentence recognition was proportional to the total time spent with SPATS. Based on the data and on reasonable extrapolations therefrom, many ESL‐learners, after 20–30 hours of spaced practice with SPATS‐ESL, appear to be approaching the performance of native‐speakers of English on all SPATS‐ESL tasks. Responses to questionnaires and comments indicate that the participants believed that SPATS‐ESL tr...

The Beneficial Role of L1 Spoken Language Skills on Initial L2 Sign Language Learning: Cognitive and Linguistic Predictors of M2L2 Acquisition.

Understanding how language modality (i.e., signed vs. spoken) affects second language outcomes in hearing adults is important both theoretically and pedagogically, as it can determine the specificity of second language (L2) theory and inform how best to teach a language that uses a new modality. The present study investigated which cognitive-linguistic skills predict successful L2 sign language acquisition. A group ( n = 25) of adult hearing L2 learners of American Sign Language underwent a cognitive-linguistic test battery before and after one semester of sign language instruction. A number of cognitive-linguistic measures of verbal memory, phonetic categorization skills, and vocabulary knowledge were examined to determine whether they predicted proficiency in a multiple linear regression analysis. Results indicated that English vocabulary knowledge and phonetic categorization skills predicted both vocabulary growth and self-rated proficiency at the end of one semester of instruction. Memory skills did not significantly predict either proficiency measures. These results highlight how linguistic skills in the first language (L1) directly predict L2 learning outcomes regardless of differences in L1 and L2 language modalities.

When two become one—Orthography helps link two free variants to one lexical entry

L2 learners can become better at distinguishing an unfamiliar contrast by knowing the corresponding orthographic forms (e.g., Escudero et al., 2008). We ask whether learners could associate two free variants with the same lexical entry when the orthographic form was provided during learning. American learners learned an artificial language where [p]-[b] were in free variation (both were spelled as ) (test condition) while [t]-[d] were contrastive (control condition), or vice-versa ([t]-[d] in test, counterbalanced across subjects). Using a word-learning paradigm modified from Hayes-Harb et al. (2010), in the learning phase, participants heard novel words paired with pictures. One subgroup of learners saw the spellings as well (“Orth+”), while another did not (i.e., auditory only, “Orth−”). Then in a picture-auditory word matching task, the new form of the word was paired with the original picture. Orth + learners were expected to be more accurate at accepting the variant as the correct label for the original test item than Orth−. The results showed that Orth + learners detected and learned the [p]-[b] free variation significantly better than Orth− (p < 0.05), but not the [t]-[d] free variation. Thus, the benefit of orthography in speech learning could vary depending on the specific contrasts at hand.

The role of orthographic information in the learning of allophonic variation

Exposure to L2 orthography may facilitate learning a novel vocalic (e.g., Escudero et al., 2008) or tonal (Showalter and Hayes-Harb, 2013) L2 contrast. Yet it is unclear whether the benefit of orthographic information applies to the learning of L2 words involving allophonic variants. We investigated whether exposure to L2 orthography can help L2 learners establish a single lexical representation for words containing allophones. We used an invented language, with word-pairs of free variants (test condition) involving the vowel alternation [ɔ]-[u], both of which can be spelled as . In the control condition, vowel alternation [e]-[a] contrasted word meanings. In a word learning experiment, Mandarin and American English speakers were presented with words paired with pictures. In addition, one subgroup of participants saw the spellings when they heard the words, while another did not. Then, in a picture-auditory word matching task, participants who learned that the variants were allophonic were expected to ...