Underspecified Representations Research Articles

Positional faithfulness and nasal assimilation in English

This article analyzes the process of Nasal Assimilation in English.1 The approach to Nasal Assimilation in a standard rule-based framework can be conducted in two ways: by assuming an underlying alveolar nasal or by employing underspecification. The article contributes to the ongoing debate regarding underspecification in phonology and focuses on employing underspecified representations in Optimality Theory. First, it is argued that in such words as, for instance, somber, Nasal Assimilation is best analyzed in terms of positional faithfulness in the form of prevocalic faithfulness. Second, as the analyses show, positional faithfulness does not provide a workable scenario for all the data, and it is necessary to use underspecification to satisfactorily analyze English words which lack the context for positional faithfulness, for example, swamp. Nevertheless, subsequent evaluations demonstrate that in certain phrases, for instance, sing boys, employing underspecification is not sufficient either, and level distinction is necessary. Therefore, the article also offers an argument in favour of levels in Optimality Theory.

Handwriting processes when spelling morphologically complex words in children with and without Developmental Language Disorder.

Representations activated during handwriting production code information on morphological structure and reflect decomposition of the root and suffix. Children with Developmental Language Disorder (DLD) have significant difficulties in spelling morphologically complex words, but previous research has not sought evidence for a morphological decomposition effect via an examination of handwriting processes in this population. Thirty-three children aged 9-10 years with DLD, 33 children matched for chronological age (CA), and 33 younger children aged 7-8 years matched for oral language ability (LA) completed a dictated spelling task (21 words; 12 with inflectional suffixes, nine with derivational suffixes). The task was completed on paper with an inking pen linked to a graphics tablet running the handwriting software Eye and Pen. Pause analyses and letter duration analyses were conducted. The three groups showed similar handwriting processes, evidencing a morphological decomposition effect in a natural writing task. Pause durations observed at the root/suffix boundary were significantly longer than those occurring in the root. Letter durations were also significantly longer for the letter immediately prior to the boundary compared to the letter after it. Nevertheless, despite being commensurate to their LA matches for mean pause durations and letter durations, children with DLD were significantly poorer at spelling derivational morphemes. Handwriting processes did significantly predict spelling accuracy but to a much lesser extent compared to reading ability. It is suggested that derivational spelling difficulties in DLD may derive more from problems with underspecified orthographic representations as opposed to handwriting processing differences.

Contrasting orthographically similar words facilitates adult second language vocabulary learning

When learners acquire new words in a second language (L2), their lexical representations and links are initially imprecise. As new, similar words are learned, these representations must become more specific to avoid errors. This study investigated whether contrasting similar words triggers this sharpening process and facilitates learning. In a multiple-choice learning task, 114 adults acquired orthographically and semantically similar L2 (pseudo)words by either contrasting them or not. In Experiment 1, participants contrasted the L2 words, and in Experiment 2 they contrasted words in their first language. Only contrasting orthographically similar L2 words facilitated their acquisition. We conclude that contrasting underspecified representations serves as a learning mechanism that guides attention to relevant lexical information. As such, it enables learners to build more specific representations and is conducive to learning. Possibilities for further research and potential implications for L2 vocabulary instruction are discussed.

Verbal Working Memory Error Patterns and Speech-Language Outcomes in Youth With Cochlear Implants.

Verbal working memory (VWM) delays are commonly found in prelingually deaf youth with cochlear implants (CIs), albeit with considerable interindividual variability. However, little is known about the neurocognitive information-processing mechanisms underlying these delays and how these mechanisms relate to spoken language outcomes. The goal of this study was to use error analysis of the letter-number sequencing (LNS) task to test the hypothesis that VWM delays in CI users are due, in part, to fragile, underspecified phonological representations in short-term memory. Fifty-one CI users aged 7-22 years and 53 normal hearing (NH) peers completed a battery of speech, language, and neurocognitive tests. LNS raw scores and error profiles were compared between samples, and a hierarchical regression model was used to test for associations with measures of speech, language, and hearing. Youth with CIs scored lower on the LNS test than NH peers and committed a significantly higher number of errors involving phonological confusions (recalling an incorrect letter/digit in place of a phonologically similar one). More phonological errors were associated with poorer performance on measures of nonword repetition and following spoken directions but not with hearing quality. Study findings support the hypothesis that poorer VWM in deaf children with CIs is due, in part, to fragile, underspecified phonological representations in short-term/working memory, which underlie spoken language delays. Programs aimed at strengthening phonological representations may improve VWM and spoken language outcomes in CI users.

The integrity of underlying phonological representations of Greek-speaking children with speech sound disorders

ABSTRACT Underlying representations are important for the development of spoken and written language. English-speaking children with speech sound disorders (SSD) have been reported to show significant difficulty with tasks tapping phonological representations. The present study describes the development of a task for the assessment of phonological representations in Greek-speaking children. The purpose of the study is twofold: to investigate the development of phonological representation in typically developing (TD) Greek-speaking children and to explore the possibility that children with SSD may have insufficiently defined phonological representations. Greek-speaking preschool aged children (4;0–5;5 years) with SSD receiving intervention (N = 20) were compared to their typically developing peers (N = 80) on their ability to complete a task of real-word auditory discrimination with picture choice. Performance differences were assessed across segmental properties and variation in word structure. TD children aged 4;0–4;5 scored lower than older TD children. A marginally significant difference in performance was found between children with SSD and TD controls. Quantitative analysis of mistakes indicated that children with SSD and TD children aged 4;0–4;5 opt for the phonological distractor. Stimuli tapping on the specification of underlying representations at a segmental level were more taxing compared to stimuli tapping on the integrity of stored word structures. Task materials and procedures used in the present study provide preliminary normative data and may be used for the assessment of Greek-speaking children with SSD. Children with SSD are at risk of developing underspecified phonological representations and therefore should receive assessment of underlying representations to guide intervention.

Contrast and Conflict in Dutch Vowels.

The nature of phonological representations has been extensively studied in phonology and psycholinguistics. While full specification is still the norm in psycholinguistic research, underspecified representations may better account for perceptual asymmetries. In this paper, we report on a mismatch negativity (MMN) study with Dutch listeners who took part in a passive oddball paradigm to investigate when the brain notices the difference between expected and observed vowels. In particular, we tested neural discrimination (indicating perceptual discrimination) of the tense mid vowel pairs /o/-/ø/ (place contrast), /e/-/ø/ (labiality or rounding contrast), and /e/-/o/ (place and labiality contrast). Our results show (a) a perceptual asymmetry for place in the /o/-/ø/ contrast, supporting underspecification of [CORONAL] and replicating earlier results for German, and (b) a perceptual asymmetry for labiality for the /e/-/ø/ contrast, which was not reported in the German study. A labial deviant [ø] (standard /e/) yielded a larger MMN than a deviant [e] (standard /ø/). No asymmetry was found for the two-feature contrast. This study partly replicates a similar MMN study on German vowels, and partly presents new findings indicating cross-linguistic differences. Although the vowel inventory of Dutch and German is to a large extent comparable, their (morpho)phonological systems are different, which is reflected in processing.

Spelling errors reveal underlying sequential and spatial processing deficits in adults with dyslexia

ABSTRACT Recent studies showed that some adults with dyslexia have difficulty processing sequentially arranged information. In a companion study, this deficit manifested as low accuracy during a word pair comparison task involving same/different decisions when two words differed in their letter sequences. This sequential deficit was associated with left/right spatial letter confusion. In the present study, we found the same underlying difficulty with sequential and spatial letter processing during word spelling. Participants were the same 22 adults with dyslexia and 20 age- and gender-matched controls as in the companion study. In the spelling task, sequential error rates were higher in the dyslexia group, compared to the controls. Measures of accuracy of serial letter order during the spelling task and the word comparison task were correlated. Only three participants, each with dyslexia, produced left/right letter reversals during spelling. These were the same participants who produced left/right errors when naming single letters. They also had profound difficulty with sequential and left/right letter processing in the spelling and word comparison tasks, and they had the most severe spelling impairment. We conclude that this pervasive, persistent difficulty with sequential and spatial reversals contributes to a severe dyslexia subtype. In the dyslexia group as a whole, additional and separate sources of errors were underspecified word representations in long-term memory and homophone errors that likely represent language-based deficits in word knowledge. In the participants, these three factors (sequential/spatial letter confusion, underspecified word form representation, language-based deficits) occurred either as single factors or in combination with each other.

Suppression in interpreting adjective noun combinations and the nature of the lexicon

AbstractA common assumption about our internal lexicon is that the meaning of words is underspecified and this underspecified representation is filled in based on the context in which the word occurs. In this paper I would like to explore a different hypothesis, that words are stored with overspecified representations which are ‘trimmed down’ by the context. This view seems to be in line with a well-known mechanism from psycholinguistics: suppression. Many studies have shown that conceptual properties of a word are initially activated but subsequently suppressed when these properties are in conflict with the context the word occurs in. This mechanism has not been tested, however, in a context where compositional application takes place as between an adjective and a noun. In this study I will discuss two lexical decision experiments testing the interpretation of two types of adjective noun combinations. For both types of combinations it is expected that the representation of the noun undergoes a change due to the conflicting information provided by the adjective. It is hypothesized that the properties of the noun that are in conflict with the adjective are initially activated, but subsequently suppressed to form a coherent representation of the adjective noun combination. While the results provide evidence for the initial activation of the conceptual properties, no evidence for the subsequent suppression was found. The initial activation shows that also in the case of adjective noun combinations, conceptual features do not ‘wait for’ an initial well-formed semantic structure. The lack of evidence for suppression primarily suggests the need for further research. If future experiments confirm that suppression does not take place within the time frame tested in the experiments, we must conclude, based on experimental findings by Schumacher (2013), that conceptual specification takes place after the shift in reference of the noun.

Auditory gating in adults with dyslexia: An ERP account of diminished rapid neural adaptation

ObjectiveA recent functional magnetic resonance imaging (fMRI) study of adults with dyslexia showed a general deficit in suppressing responses to various types of repetitive stimuli. This diminished neural adaptation may interfere with implicit learning and forming stable word representations. With fMRI, spatial but not temporal characteristics of the adaptation response could be identified. We address this knowledge gap using event-related potentials. MethodsFourteen adults with dyslexia and 14 controls participated in an auditory gating paradigm using tone pairs. Response amplitudes and latencies for N1 and P2 were measured. Participants also compared word pairs consisting of identical or subtly different words, a task requiring stable word representations. ResultsOnly the controls showed a robust gating effect in an attenuated N1 response to the second tone relative to the first. The dyslexia group was less accurate than the controls in detecting word differences. The N1 gating magnitude was associated with this detection accuracy. ConclusionsNeural adaptation occurs by approximately 100 ms after stimulus presentation and is diminished in adults with dyslexia. This complements fMRI findings of relevant brain regions by implying a time window representing sensory and pre-attentive auditory processes. SignificanceThe association between gating magnitude and word discrimination contributes to a neurophysiological account of underspecified word representations.

Individual differences in subphonemic sensitivity and phonological skills.

Many studies have established a link between phonological abilities (indexed by phonological awareness and phonological memory tasks) and typical and atypical reading development. Individuals who perform poorly on phonological assessments have been mostly assumed to have underspecified (or "fuzzy") phonological representations, with typical phonemic categories, but with greater category overlap due to imprecise encoding. An alternative posits that poor readers have overspecified phonological representations, with speech sounds perceived allophonically (phonetically distinct variants of a single phonemic category). On both accounts, mismatch between phonological categories and orthography leads to reading difficulty. Here, we consider the implications of these accounts for online speech processing. We used eye tracking and an individual differences approach to assess sensitivity to subphonemic detail in a community sample of young adults with a wide range of reading-related skills. Subphonemic sensitivity inversely correlated with meta-phonological task performance, consistent with overspecification.

A expressão morfofonológica da concordância de número variável em português brasileiro: uma proposta para a aquisição e a produção infantil

Crianças brasileiras estão expostas a input de concordância de número variável (NARO, 1981; SCHERRE, 1991; SCHERRE; NARO, 1998). Explora-se, neste artigo, uma possível caracterização da representação e da produção da variação morfofonológica da concordância de número no português brasileiro. Apresentam-se os resultados de um experimento de produção eliciada conduzido com crianças em idade pré-escolar de escolas pública (EscPub) e particular (EscPriv) no Rio de Janeiro com vistas a verificar a preferência das crianças por formas redundantes (plural no artigo, no nome e no verbo) e formas não-redundantes (plural apenas no artigo), e variações entre as duas formas. EscPub não apresenta preferência em relação à concordância de número interna ao DP sujeito, enquanto EscPriv prefere a forma redundante do DP. Parte-se de um modelo procedimental da aquisição da linguagem, segundo o qual a criança extrai informação gramaticalmente relevante dos padrões morfofonológicos legíveis na interface fônica (CORRÊA, 2009, 2014). Diante de input variável, sugere-se que a representação gramatical da expressão morfofonológica de número plural estaria subespecificada, sendo, portanto, sujeita a fatores de frequência e informações de uso no momento da codificação morfofonológica na produção da linguagem (ADGER, 2007; LEVELT, 1999; YANG, 2002), o que acarretaria variação.

Aspect and modality in the interpretation of deverbal -er nominals in English

This article presents a corpus study of over 16,000 tokens of -er nominalizations on 62 verbal bases that were extracted from the Corpus of Contemporary American English and the British National Corpus. We show that an individual -er nominal can often be given a range of modal and aspectual readings and that a number of factors influence the availability of different readings for -er nominals, including verb type, syntactic context (verb tenses, adverbs), and encyclopedic information. On the basis of these data, we argue, contra Cohen (2016), that the core meaning of the affix -er (as in writer, printer, etc.) cannot be that of a dynamic modal. We show that neither Cohen’s (2016) analysis nor syntactic analyses such as that of Alexiadou and Schafer (2010) can account for the range of readings we find. We conclude by sketching one possible analysis in terms of the Lexical Semantic Framework of Lieber (2004, 2016) that postulates underspecified lexical representations of the -er nominals and resolution of underspecification in context.

Artist-for-work metonymy

Abstract There is an on-going debate about how the language system handles expressions that may refer to different word senses. Some theories propose derivational operations triggered by a type-mismatch; others assume underspecified lexical representations that engage in sense selection. Previous studies yielded mixed evidence. To further understand the diverse results and to tease apart type conflicts from sense selection, we conducted an ERP study on one type of metonymy using question-answer pairs. We capitalized on the fact that the metonymic relation could be anticipated by a wh-word, yielding a type clash between the wh-word and the predicate that cannot be resolved in the question (Whom did she read? vs. What did she read?) – in contrast to the answer (She read Brecht.). These conditions were contrasted with a non-metonymic reading. The results revealed a pronounced N400 at the verb of the whom-question in contrast to the questions without a type clash, and no reliable differences at the artist’s name in the answer. We therefore argue for an underspecification account for this metonymy-type which is not preceded by the detection of a type conflict. In comparison with other metonymy-types, this substantiates a classification of sense alternations into sense selection and sense creation.

A Notion of Semantic Coherence for Underspecified Semantic Representation

The general problem of finding satisfying solutions to constraint-based underspecified representations of quantifier scope is NP-complete. Existing frameworks, including Dominance Graphs, Minimal Recursion Semantics, and Hole Semantics, have struggled to balance expressivity and tractability in order to cover real natural language sentences with efficient algorithms. We address this trade-off with a general principle of coherence, which requires that every variable introduced in the domain of discourse must contribute to the overall semantics of the sentence. We show that every underspecified representation meeting this criterion can be efficiently processed, and that our set of representations subsumes all previously identified tractable sets.

Is vowel nasalisation phonological in English? A systematic review

Vowel nasalisation in American English has been the subject of a long discussion as to whether it is a phonological process or whether it should better be analysed as phonetic coarticulation. As a predictable allophonic process, vowel nasalisation also provides a testing ground for theories of lexical representation, since if it turns out that language users store this predictable information in long-term memory, there is no reason to assume any kind of phonological underspecification. In this article, experimental studies on this phenomenon are reviewed with these two questions in mind: is the phenomenon a phonological process at all and is this predictable information, i.e. vowel nasalisation, stored in the mental lexicon. The majority of studies provide evidence supporting the phonological view and underspecified lexical representations.

Early language experience and underspecified phonological representations

Taking as a premise that phonological working memory (PWM) influences later language development, in their keynote article, Pierce, Genesee, Delcenserie, and Morgan aim to specify the relations between early language input and the development of PWM in terms of separable influences of timing, quantity, and quality of early language input. We concur that prior work has established that PWM and language development have reciprocal influences on one another during development (e.g., Baddeley, Gathercole, & Papagno, 1998; Gathercole, 2006; Gathercole, Hitch, Service, & Martin, 1997; Metsala & Chisolm, 2010). The goal of the keynote article was to describe how early language experience may influence the development of PWM. Pierce et al. argue that this can be done by comparing the development of PWM across groups of children with differing language experiences during early childhood, specifically (a) delayed exposure to language, (b) impoverished language input, or (c) enriched language input. The authors suggest that this comparison may contribute to establishing that individual differences in PWM are due, in part, to early language experience. Sensitive periods for phonological development that are open roughly in the first year of life are discussed, and it is suggested that the quantity and quality of early language input shapes the quality of phonological representations. Efforts to specify mechanisms by which early language input may influence the development of PWM have both theoretical and, potentially, clinical importance. Considering this, Pierce et al.’s article, which aims to create a platform for future research in terms of the timing, quantity, and quality of early language input, is a valuable contribution.

About full and underspecified representations of alternations in the lexicon:evidence from sentence reading

About full and underspecified representations of alternations in the lexicon:evidence from sentence reading

A Diagnostic Marker to Discriminate Childhood Apraxia of Speech From Speech Delay: III. Theoretical Coherence of the Pause Marker with Speech Processing Deficits in Childhood Apraxia of Speech.

Previous articles in this supplement described rationale for and development of the pause marker (PM), a diagnostic marker of childhood apraxia of speech (CAS), and studies supporting its validity and reliability. The present article assesses the theoretical coherence of the PM with speech processing deficits in CAS. PM and other scores were obtained for 264 participants in 6 groups: CAS in idiopathic, neurogenetic, and complex neurodevelopmental disorders; adult-onset apraxia of speech (AAS) consequent to stroke and primary progressive apraxia of speech; and idiopathic speech delay. Participants with CAS and AAS had significantly lower scores than typically speaking reference participants and speech delay controls on measures posited to assess representational and transcoding processes. Representational deficits differed between CAS and AAS groups, with support for both underspecified linguistic representations and memory/access deficits in CAS, but for only the latter in AAS. CAS-AAS similarities in the age-sex standardized percentages of occurrence of the most frequent type of inappropriate pauses (abrupt) and significant differences in the standardized occurrence of appropriate pauses were consistent with speech processing findings. Results support the hypotheses of core representational and transcoding speech processing deficits in CAS and theoretical coherence of the PM's pause-speech elements with these deficits.

Good-enough linguistic representations and online cognitive equilibrium in language processing.

We review previous research showing that representations formed during language processing are sometimes just "good enough" for the task at hand and propose the "online cognitive equilibrium" hypothesis as the driving force behind the formation of good-enough representations in language processing. Based on this view, we assume that the language comprehension system by default prefers to achieve as early as possible and remain as long as possible in a state of cognitive equilibrium where linguistic representations are successfully incorporated with existing knowledge structures (i.e., schemata) so that a meaningful and coherent overall representation is formed, and uncertainty is resolved or at least minimized. We also argue that the online equilibrium hypothesis is consistent with current theories of language processing, which maintain that linguistic representations are formed through a complex interplay between simple heuristics and deep syntactic algorithms and also theories that hold that linguistic representations are often incomplete and lacking in detail. We also propose a model of language processing that makes use of both heuristic and algorithmic processing, is sensitive to online cognitive equilibrium, and, we argue, is capable of explaining the formation of underspecified representations. We review previous findings providing evidence for underspecification in relation to this hypothesis and the associated language processing model and argue that most of these findings are compatible with them.

Implementing semantic frames as typed feature structures with XMG

This work presents results on the integration of frame-based representations into the framework of eXtensible MetaGrammar (XMG). Originally XMG allowed for the description of tree-based syntactic structures and underspecified representations of predicate-logical formulae, but the representation of frames as a sort of typed feature structure, particularly type unification, was not supported. Therefore, we introduce an extension that is capable of handling frame representations directly by means of a novel <frame>-dimension. The aim is not only to make possible a straightforward specification of frame descriptions, but also to offer various ways to specify constraints on types, be it as a contiguous type hierarchy or a loose set of feature structure constraints. The presented extensions to XMG are fully operational in a new prototype.