Cross-Linguistic Trends in Speech Errors: An Analysis of Sub-Lexical Errors in Cantonese

Though past research on the sound structure of speech errors has contributed greatly to our understanding of phonological encoding, most of this research comes from a small set of majority languages with similar linguistic structures. To increase the linguistic diversity of relevant evidence, a large collection of speech errors was investigated in Cantonese, an under-studied language with unique phonological structures. In particular, the Cantonese data were examined for nine psycholinguistic effects commonly used as a lens on word-form encoding. Detailed quantitative analysis found that Cantonese has eight of these effects, providing broader cross-linguistic support for models based on these patterns. Yet Cantonese also exhibited differences with well-known Indo-European languages by having a higher rate of errors involving whole syllables and sub-constituents inside the syllable rime. These differences can be accounted for by recognizing the primacy of the syllable and mora in encoding Cantonese words, following proposals that have been made for Mandarin Chinese and Japanese.

Cascading activation in phonological planning and articulation: Evidence from spontaneous speech errors

Speaking involves both retrieving the sounds of a word (phonological planning) and realizing these selected sounds in fluid speech (articulation). Recent phonetic research on speech errors has argued that multiple candidate sounds in phonological planning can influence articulation because the pronunciation of mis-selected error sounds is slightly skewed towards unselected target sounds. Yet research to date has only examined these phonetic distortions in experimentally-elicited errors, leaving doubt as to whether they reflect tendencies in spontaneous speech. Here, we analyzed the pronunciation of speech errors of English-speaking adults in natural conversations relative to matched correct words by the same speakers, and found the conjectured phonetic distortions. Comparison of these data with a larger set of experimentally-elicited errors failed to reveal significant differences between the two types of errors. These findings provide ecologically-valid data supporting models that allow for information about multiple planning representations to simultaneously influence speech articulation.

Sequential adaptation effects reveal proactive control in processing spoken sentences: Evidence from event-related potentials

How domain-general cognitive control is engaged in language processing remains debated. We address how linguistic processes are monitored and regulated by analyzing the effects of previous-trial sentence correctness on the P600 component of the event-related potential (ERP) in the current-trial. In data from a previous experiment about processing spoken sentences, P600 amplitudes to both correct and incorrect words in current sentences were smaller after incorrect as compared to correct previous sentences. Therefore, the detection of speech errors may initiate sustained proactive control over the monitoring demands for upcoming sentences. No sequential adaptation was found in the difference between P600 amplitudes to incorrect and correct current conditions. We propose that the P600 reflects the reactive reanalysis of speech processing and/or the resolution of linguistic conflicts, but is also sensitive to proactive speech monitoring, an important aspect of cognitive control.

Quantifying the Interplay of Semantics and Phonology During Failures of Word Retrieval by People With Aphasia Using a Multiplex Lexical Network

Investigating instances where lexical selection fails can lead to deeper insights into the cognitive machinery and architecture supporting successful word retrieval and speech production. In this paper, we used a multiplex lexical network approach that combines semantic and phonological similarities among words to model the structure of the mental lexicon. Network measures at different levels of analysis (degree, network distance, and closeness centrality) were used to investigate the influence of network structure on picture naming accuracy and errors by people with Anomic, Broca's, Conduction, and Wernicke's aphasia. Our results reveal that word retrieval is influenced by the multiplex lexical network structure in at least two ways-(a) the accuracy of production and error type on incorrect productions were influenced by the degree and closeness centrality of the target word, and (b) error type also varied in terms of network distance between the target word and produced error word. Taken together, the analyses demonstrate that network science techniques, particularly the use of the multiplex lexical network to simultaneously represent semantic and phonological relationships among words, reveal how the structure of the mental lexicon influences language processes beyond traditionally examined psycholinguistic variables. We propose a framework for how the multiplex lexical network approach allows for understanding the influence of mental lexicon structure on word retrieval processes, with an eye toward a better understanding of the nature of clinical impairments, like aphasia.

Tone slips in Cantonese: Evidence for early phonological encoding

This article examines speech errors in Cantonese with the aim of fleshing out a larger speech production architecture for encoding phonological tone. A corpus was created by extracting 2462 speech errors, including 668 tone errors, from audio recordings of natural conversations. The structure of these errors was then investigated in order to distinguish two contemporary approaches to tone in speech production. In the tonal frames account, tone is encoded like metrical stress, represented in abstract structural frames for a word. Because tone cannot be mis-selected in tonal frames, tone errors are expected to be rare and non-contextual, as observed with stress. An alternative is that tone is actively selected in phonological encoding like phonological segments. This approach predicts that tone errors will be relatively common and exhibit the contextual patterns observed with segments, like perseveration and anticipation. In our corpus, tone errors are the second most common type of error, and the majority of errors exhibit contextual patterns that parallel segmental errors. Building on prior research, a two-stage model of phonological tone encoding is proposed, following the patterns seen in tone errors: Tone is phonologically selected concurrently with segments, but then sequentially assigned after segments to a syllable.

Investigating Perceptual Biases, Data Reliability, and Data Discovery in a Methodology for Collecting Speech Errors From Audio Recordings

This work describes a methodology of collecting speech errors from audio recordings and investigates how some of its assumptions affect data quality and composition. Speech errors of all types (sound, lexical, syntactic, etc.) were collected by eight data collectors from audio recordings of unscripted English speech. Analysis of these errors showed that: (i) different listeners find different errors in the same audio recordings, but (ii) the frequencies of error patterns are similar across listeners; (iii) errors collected "online" using on the spot observational techniques are more likely to be affected by perceptual biases than "offline" errors collected from audio recordings; and (iv) datasets built from audio recordings can be explored and extended in a number of ways that traditional corpus studies cannot be.

The Similarities Between the Target and the Intruder in Naturally Occurring Person Naming Errors: A Comparison Between Repeated and Single Naming Confusions

This study investigated the phenomenon of personal name confusion, i.e. calling a familiar person by someone else's name. Two types of name confusion were considered: single confusions (i.e. confusions that appeared only once) and repeated confusions (i.e. confusions that appeared repeatedly). The main purpose of the present study was to compare these two types of personal name confusion and to identify the similarities and differences between them. Participants were asked to fill in two questionnaires (one for each type of confusion) in order to collect information about the properties of the confusions. Results indicated that single and repeated confusions shared some similarities (the similarity of the gender and age of the bearers of the confused names, the phonological similarity between the confused names, the positive or negative valence of the relationship between the participant and the bearers of the names, the frequency of encountering the bearers of the names, and the low level of stress on the part of the participant when the confusions were made). However, some differences were also found between single and repeated confusions (the context of encountering, the length of time that the participant had known the two bearers of the names, the presence of the inverse confusion, the facial resemblance between the two bearers, the kind of relationship shared by the participant and the two bearers, and the state of tiredness on the part of the participant when the confusions were made). In addition, regression analysis indicated that the facial resemblance between the target person and the intruder, the phonological similarity of the names and the difference in age between the two bearer of the names were significant predictors of the frequency of the repeated confusions.

The prosodic domain of phonological encoding: Evidence from speech errors

Phonological encoding of segments is thought to occur within a prosodically-defined frame, but it is not clear which of the constituent/s within the prosodic hierarchy (syllables, phonological words, intonational phrases and utterances) serve/s as the domain of phonological encoding. This experiment investigated whether segmental speech errors elicited in tongue-twisters were influenced by position within prosodic constituents above the level of the phonological word. Forty-four participants produced six repetitions each of 40 two-intonational phrase tongue-twisters with error-prone word-initial "target" segments in phrase-initial and phrase-final words. If the domain of phonological encoding is the intonational phrase, we hypothesised that segments within a current intonational phrase would interact in more errors than would segments across intonational phrase boundaries. Participants made more anticipatory than perseveratory errors on target segments in phrase-initial words as predicted. They also made more perseveratory than anticipatory errors on targets in phrase-final words, but only in utterance-final phrases. These results suggest that the intonational phrase is one domain of phonological encoding, and that segments for upcoming phrases are activated while current phrases are being articulated.

Bilingual language intrusions and other speech errors in Alzheimer's disease

The current study investigated how Alzheimer's disease (AD) affects production of speech errors in reading-aloud. Twelve Spanish-English bilinguals with AD and 19 matched controls read-aloud 8 paragraphs in four conditions (a) English-only, (b) Spanish-only, (c) English-mixed (mostly English with 6 Spanish words), and (d) Spanish-mixed (mostly Spanish with 6 English words). Reading elicited language intrusions (e.g., saying la instead of the), and several types of within-language errors (e.g., saying their instead of the). Patients produced more intrusions (and self-corrected less often) than controls, particularly when reading non-dominant language paragraphs with switches into the dominant language. Patients also produced more within-language errors than controls, but differences between groups for these were not consistently larger with dominant versus non-dominant language targets. These results illustrate the potential utility of speech errors for diagnosis of AD, suggest a variety of linguistic and executive control impairments in AD, and reveal multiple cognitive mechanisms needed to mix languages fluently. The observed pattern of deficits, and unique sensitivity of intrusions to AD in bilinguals, suggests intact ability to select a default language with contextual support, to rapidly translate and switch languages in production of connected speech, but impaired ability to monitor language membership while regulating inhibitory control.

The Frame Constraint on Experimentally Elicited Speech Errors in Japanese

The so-called syllable position effect in speech errors has been interpreted as reflecting constraints posed by the frame structure of a given language, which is separately operating from linguistic content during speech production. The effect refers to the phenomenon that when a speech error occurs, replaced and replacing sounds tend to be in the same position within a syllable or word. Most of the evidence for the effect comes from analyses of naturally occurring speech errors in Indo-European languages, and there are few studies examining the effect in experimentally elicited speech errors and in other languages. This study examined whether experimentally elicited sound errors in Japanese exhibits the syllable position effect. In Japanese, the sub-syllabic unit known as "mora" is considered to be a basic sound unit in production. Results showed that the syllable position effect occurred in mora errors, suggesting that the frame constrains the ordering of sounds during speech production.

Automatic analysis of slips of the tongue: Insights into the cognitive architecture of speech production

Traces of the cognitive mechanisms underlying speaking can be found within subtle variations in how we pronounce sounds. While speech errors have traditionally been seen as categorical substitutions of one sound for another, acoustic/articulatory analyses show they partially reflect the intended sound. When "pig" is mispronounced as "big," the resulting /b/ sound differs from correct productions of "big," moving towards intended "pig"-revealing the role of graded sound representations in speech production. Investigating the origins of such phenomena requires detailed estimation of speech sound distributions; this has been hampered by reliance on subjective, labor-intensive manual annotation. Computational methods can address these issues by providing for objective, automatic measurements. We develop a novel high-precision computational approach, based on a set of machine learning algorithms, for measurement of elicited speech. The algorithms are trained on existing manually labeled data to detect and locate linguistically relevant acoustic properties with high accuracy. Our approach is robust, is designed to handle mis-productions, and overall matches the performance of expert coders. It allows us to analyze a very large dataset of speech errors (containing far more errors than the total in the existing literature), illuminating properties of speech sound distributions previously impossible to reliably observe. We argue that this provides novel evidence that two sources both contribute to deviations in speech errors: planning processes specifying the targets of articulation and articulatory processes specifying the motor movements that execute this plan. These findings illustrate how a much richer picture of speech provides an opportunity to gain novel insights into language processing.

Did I say dog or cat? A study of semantic error detection and correction in children

Although naturalistic studies of spontaneous speech suggest that young children can monitor their speech, the mechanisms for detection and correction of speech errors in children are not well understood. In particular, there is little research on monitoring semantic errors in this population. This study provides a systematic investigation of detection and correction of semantic errors in children between the ages of 5 and 8years as they produced sentences to describe simple visual events involving nine highly familiar animals (the moving animals task). Results showed that older children made fewer errors and corrected a larger proportion of the errors that they made than younger children. We then tested the prediction of a production-based account of error monitoring that the strength of the language production system, and specifically its semantic-lexical component, should be correlated with the ability to detect and repair semantic errors. Strength of semantic-lexical mapping, as well as lexical-phonological mapping, was estimated individually for children by fitting their error patterns, obtained from an independent picture-naming task, to a computational model of language production. Children's picture-naming performance was predictive of their ability to monitor their semantic errors above and beyond age. This relationship was specific to the strength of the semantic-lexical part of the system, as predicted by the production-based monitor.

Perfectionism and stuttering: Findings of an online survey

PURPOSE

Using a multi-dimensional measure of perfectionism: the Frost Multi-dimensional Perfectionism Scale (FMPS: Frost, Marten, Lahart, & Rosenblate, 1990), this study investigates: (a) whether adults who stutter (AWS) display more perfectionistic attitudes and beliefs than those who do not stutter, and (b) whether, in AWS, more perfectionistic attitudes and beliefs are associated with greater self-reported difficulty communicating verbally and speaking fluently.

METHOD

In the first analysis, FMPS responses from 81 AWS and 81 matched, normally-fluent controls were analyzed using logistic regression to investigate the relative contributions of four FMPS perfectionism-subscale self-ratings to the likelihood of being in the AWS group. In the subsequent analyses, data from the 81 AWS were analyzed using linear multiple regression to determine which FMPS subscale self-ratings best predicted their Communication-Difficulty and Fluency-Difficulty scores.

RESULTS

Both the likelihood of being a member of the AWS group, and also the magnitude of the AWS group's Communication-Difficulty and Fluency-Difficulty scores, were positively part-correlated to respondents' Concern over Mistakes-Doubts about Actions (CMD) subscale self-ratings but negatively part-correlated to their Personal Standards (PS) subscale self-ratings.

CONCLUSIONS

The FMPS profiles of respondents who stutter suggest that, as a group, they are not abnormally perfectionistic overall, but may be (or perceive themselves to be) abnormally error-prone. Also, AWS who are more concerned about their errors and uncertain of their actions experience more difficulty communicating verbally and speaking fluently.

EDUCATIONAL OBJECTIVES

After reading this article, participants will be able to: (a) describe the findings of previous research investigating the role of perfectionism in stuttering and psychopathologies; (b) discuss why a multidimensional assessment of perfectionism is important in relation to stuttering; (c) discuss ways in which data from perfectionism assessments can contribute to the planning of therapy for adults who stutter.

The effects of anodal stimulation of the left prefrontal cortex on sentence production

BACKGROUND

Most studies in which Anodal Transcranial Direct Current Stimulation (A-tDCS) has been used to improve language production have focused on single words. Yet sentence production requires more than lexical retrieval. For example, successful suppression of the past and careful planning of the future are two critical requirements for producing a correct sentence. Can A-tDCS improves those, and by extension, production at the sentence level?

OBJECTIVE/HYPOTHESIS

Given that many aspects of sentence production beyond word retrieval require frontally-mediated operations, we hypothesized that A-tDCS to the left prefrontal cortex should benefit various operation involved in producing sentences, two of which, suppression of the past and planning of the future, were targeted in this study.

METHODS

We used a paradigm that elicited construction of sentences through event description, but was structured enough to allow for between-subject comparison, clear error identification, and implementation of experimental manipulations to probe certain aspects of production.

RESULTS

We showed that A-tDCS to the left PFC reliably decreased the number of incomplete and errorful sentences. When the origin of this improvement was probed, we found that A-tDCS significantly decreased errors due to premature commitment to the future word (insufficient internal monitoring), and had a marginal effect on errors of perseverations (insufficient suppression of the past).

CONCLUSION

We conclude that A-tDCS is a promising tool for improving production at the sentence level, and that improvement can be expected when internal monitoring and control over verbal responses is impaired, or for certain cases of perseveratory errors.