YPL2 – Issue 13 (December 2012)

Editors: Natalie Fecher, Dima Al-Malahmeh, Mariam Dar, Ella Jeffries, Ania Kubisz, Man Ki Theodora Lee, Kaj Nyman and Rebecca Woods
Why Don’t Deaf Readers Garden-Path?
Linda Coulter and Helen Goodluck
Abstract Download PDF

Self-paced reading was used in three experiments to examine online sentence processing in sixteen deaf adults. Experiment 1 examined processing of simple declarative sentences. Experiment 2 examined processing of temporarily ambiguous sentences with filler-gap dependences, and experiment 3 explored sentences with a temporary ambiguity between direct object and subject analysis of an NP. The results from each experiment show that the deaf participants did not process the experimental items in the same way as the hearing participants. In experiment 1 we found no syntactic 'wrap up' effect (slowing of reaction times at the end of a sentence) for the deaf; in experiment 2, we found no tendency for the deaf to erroneously anticipate the position of a wh-word; and in experiment 3, we found that deaf readers do not follow a strategy of attempting to incorporate the most recently input material into the parse of an existing phrase. We suggest that lack of access to prosodic information in the language input during acquisition (because of auditory deprivation) results in atypical development of parsing strategies.

Reference Sample Size and the Computation of Numerical Likelihood Ratios Using Articulation Rate
Vincent Hughes, Ashley Brereton and Erica Gold
Abstract Download PDF

This paper explores the effects of variability in the amount of reference data used in quantifying the strength of speech evidence using numerical likelihood ratios (LRs). Monte Carlo simulations (MCS) are performed to generate synthetic data from a sample of existing raw local articulation rate (AR) data. LRs are computed as the number of reference speakers (up to 1000), and the number of tokens per reference speaker (up to 200) is systematically increased. The distributions of same-speaker and different-speaker LRs and system performance (log LR cost (Cllr) and equal error rate (EER)) are assessed as a function of the size of the reference data. Results reveal that LRs based on AR are relatively robust to small reference samples, but that system calibration plays an important role in determining the sensitivity of the LRs to sample size.

Late Talking Toddlers: Relating Early Phonological Development to Later Language Advance
Marilyn M. Vihman, Tamar Keren-Portnoy, Christopher Whitaker, Amy Bidgood and Michelle McGillion
Abstract Download PDF

Background. Expressive Late talkers are identified as children with an unusually small productive vocabulary for their age, in the absence of any other known neurological, sensory or cognitive deficit. Their lexical delay has been found to be associated with phonetic delay.

Aims. The two primary goals of this study are (1) to provide intensive analyses of phonetic and phonological characteristics of late talkers (LTs) at the end of the single- word period as a basis for comparing their speech with that of typically developing children (TDs), not at the same age but at the same developmental point; (2) to compare the relative phonological, lexical, morphological and syntactic advance of the same two groups 14 months later, based on analysis of spontaneous language use, and to relate this advance to phonetic and phonological resources at the earlier measurement point.

Methods and procedures. Time 1 analyses included assessment of volubility, size of consonant inventory, percent consonants correct and extent of consonant variegation and of the use of selected prosodic patterns or 'templates'. Time 2 analyses assessed advance in phonology (percent consonants correct), lexicon (diversity of verb and function word types), morphology (provision of obligatory morphemes) and syntax (MLU, IPSyn).

Outcomes and results. Although three groups differing in age at achieving Time 1 lexical criterion were identified (TDs, LTs and 'transitional' LTs or TLTs), there is little evidence of group differences in other measures of linguistic advance at either sampling point, when the groups are compared at the same lexical level. Exploratory statistical analyses using Canonical Correlations revealed that a combination of high age at Time 1, small consonant inventory and low phonetic variegation are strong predictors of low accuracy in consonant use and relatively poor lexicon, morphology and syntax at Time 2, while dependence on a limited set of phonological patterns at Time 1 was significantly correlated more specifically with slower morphological advance at Time 2.

Conclusions and implications. The study found that, once the groups are equated for lexical level, the linguistic skills of LTs as a group are not distinguishable from those of TDs in either the early period of phonological development or the year following the end of the single-word period. Nevertheless, based on the relation of Time 1 to Time 2 measures within individuals, the study also demonstrates that phonetic and phonological knowledge and skills constitute a key foundation for later linguistic advance, as regards grammar as well as phonology.

Object Preposing in Late Archaic Chinese
Aiqing Wang
Abstract Download PDF

This paper investigates the preverbal positioning of objects in Late Archaic Chinese (5th-3rdc BC; "LAC"). As an SVO language, LAC permits DP objects to front into preverbal positions in a medial domain below TP and above vP. Based on the relative ordering of preposed non-wh-constituents and negation, two positions can be found: a high position and a low position. The high position for the fronting of non-wh-objects displays topic properties, while the low position displays focus features. Nominal and pronominal objects in LAC may occur in either position; all preposed constituents occupy a specifier node of functional projections (Paul 2002, 2005), followed by an optional fronting marker as the head of the relevant functional categories. Within the medial domain, head-like elements are always in a fixed relative order: negatives precede modals of ability, and follow other modals. In terms of wh-DPs, D-linked which-phrases appear in the topic position, whereas non-D-linked wh-phrases are permitted exclusively in an extra (focus) position between the topic position and negation, triggered by the Intervention Effect of negation (Kim 2002). This paper also explores the underlying structure of wh-P.

Inference of Threat from Neutrally-Worded Utterances in Familiar and Unfamiliar Languages
Dominic Watt, Sarah Kelly and Carmen Llamas
Abstract Download PDF

Although verbal threats are a very common kind of language crime, the ways in which listeners interpret ostensibly 'neutral' utterances as threats are currently poorly understood. We present the results of an experiment in which monolingual English-speaking listeners were exposed to the same innocuously-worded phrase spoken in a 'neutral' or a 'threatening' way. They heard translations of the phrase in four unknown foreign languages as well as the original form in English. The listeners were asked to rate the utterances with respect to two perceived properties: (a) how threatening they thought the utterances sounded, and (b) how much 'intent' to carry out a harmful act the listeners inferred from the talker's speech. As predicted, the listeners assigned higher threat and intent ratings to the 'threatening' utterances in both English and the foreign languages than they did to the neutral ones. However, the listeners' ratings were considerably higher for both threatening and neutral utterances spoken in English than they were for the foreign language utterances. In the English condition there was also a much larger difference between the neutral and the threat utterances with respect to the overall perceived threat and intent ratings than there was for the foreign language utterances. This suggests that correctly interpreting the threatening utterances as threats is dependent upon familiarity with the language in which they were spoken. A gender effect was also found, whereby male listeners assigned higher threat and intent ratings than did women. It is suggested that men and women may respond differently to speech cues associated with threatening behaviour.