Internet-based language production research with overt articulation: Proof of concept, challenges, and practical advice

Effects of irrelevant unintelligible and intelligible background speech on spoken language production

Earlier work has explored spoken word production during irrelevant background speech such as intelligible and unintelligible word lists. The present study compared how different types of irrelevant background speech (word lists vs. sentences) influenced spoken word production relative to a quiet control condition, and whether the influence depended on the intelligibility of the background speech. Experiment 1 presented native Dutch speakers with Chinese word lists and sentences. Experiment 2 presented a similar group with Dutch word lists and sentences. In both experiments, the lexical selection demands in speech production were manipulated by varying name agreement (high vs. low) of the to-be-named pictures. Results showed that background speech, regardless of its intelligibility, disrupted spoken word production relative to a quiet condition, but no effects of word lists versus sentences in either language were found. Moreover, the disruption by intelligible background speech compared with the quiet condition was eliminated when planning low name agreement pictures. These findings suggest that any speech, even unintelligible speech, interferes with production, which implies that the disruption of spoken word production is mainly phonological in nature. The disruption by intelligible background speech can be reduced or eliminated via top-down attentional engagement.

The role of answer content and length when preparing answers to questions

Research suggests that interlocutors manage the timing demands of conversation by preparing what they want to say early. In three experiments, we used a verbal question-answering task to investigate what aspects of their response speakers prepare early. In all three experiments, participants answered more quickly when the critical content (here, barks) necessary for answer preparation occurred early (e.g., Which animal barks and is also a common household pet?) rather than late (e.g., Which animal is a common household pet and also barks?). In the individual experiments, we found no convincing evidence that participants were slower to produce longer answers, consisting of multiple words, than shorter answers, consisting of a single word. There was also no interaction between these two factors. A combined analysis of the first two experiments confirmed this lack of interaction, and demonstrated that participants were faster to answer questions when the critical content was available early rather than late and when the answer was short rather than long. These findings provide tentative evidence for an account in which interlocutors prepare the content of their answer as soon as they can, but sometimes do not prepare its length (and thus form) until they are ready to speak.

Response planning during question-answering: does deciding what to say involve deciding how to say it?

To answer a question, speakers must determine their response and formulate it in words. But do they decide on a response before formulation, or do they formulate different potential answers before selecting one? We addressed this issue in a verbal question-answering experiment. Participants answered questions more quickly when they had one potential answer (e.g., Which tourist attraction in Paris is very tall?) than when they had multiple potential answers (e.g., What is the name of a Shakespeare play?). Participants also answered more quickly when the set of potential answers were on average short rather than long, regardless of whether there was only one or multiple potential answers. Thus, participants were not affected by the linguistic complexity of unselected but plausible answers. These findings suggest that participants select a single answer before formulation.

Self-initiation enhances perceptual processing of auditory stimuli in an online study

Understanding how the brain incorporates sensory and motor information will enable better theory building on human perception and behavior. In this study, we aimed to estimate the influence of predictive mechanisms on the magnitude and variability of sensory attenuation in two online samples. After the presentation of a visual cue stimulus, participants (Experiment 1: N = 224, Experiment 2: N = 84) compared the loudness of two consecutive tones in a two-alternative forced-choice task. In Experiment 1, the first tone was either self-initiated or not; in Experiment 2, the second tone was either self-initiated or not (active and passive condition, respectively). We further manipulated identity prediction (i.e., the congruence of pre-learned cue-sound combinations; congruent vs. incongruent), and the duration of the onset delay (to account for effects of attentional differences between the passive and active condition, 50 ms vs. 0 ms). We critically discuss our results within the framework of both classical (i.e., motor-based forward models) and contemporary approaches (i.e., predictive processing framework). Contrary to our preregistered hypothesis, we observed enhanced perceptual processing, instead of attenuation, for self-initiated auditory sensory input. Further, our results reveal an effect of fixed sound delays on the processing of motor and non-motor-based predictive information, and may point to according shifts in attention, leading to a perceptual bias. These results might best be captured by a hybrid explanatory model, combining predictions based on self-initiated motor action with a global predictive mechanism.

Phonetic accommodation in non-native directed speech supports L2 word learning and pronunciation

This study assessed whether Non-native Directed Speech (NNDS) facilitates second language (L2) learning, specifically L2 word learning and production. Spanish participants (N = 50) learned novel English words, presented either in NNDS or Native-Directed Speech (NDS), in two tasks: Recognition and Production. Recognition involved matching novel objects to their labels produced in NNDS or NDS. Production required participants to pronounce these objects' labels. The novel words contained English vowel contrasts, which approximated Spanish vowel categories more (/i-ɪ/) or less (/ʌ-æ/). Participants in the NNDS group exhibited faster recognition of novel words, improved learning, and produced the /i-ɪ/ contrast with greater distinctiveness in comparison to the NDS group. Participants' ability to discriminate the target vowel contrasts was also assessed before and after the tasks, with no improvement detected in the two groups. These findings support the didactic assumption of NNDS, indicating the relevance of the phonetic adaptations in this register for successful L2 acquisition.

Word frequency has similar effects in picture naming and gender decision: A failure to replicate Jescheniak and Levelt (1994)

Word frequency plays a key role in theories of lexical access, which assume that the word frequency effect (WFE, faster access to high-frequency than low-frequency words) occurs as a result of differences in the representation and processing of the words. In a seminal paper, Jescheniak and Levelt (1994) proposed that the WFE arises during the retrieval of word forms, rather than the retrieval of their syntactic representations (their lemmas) or articulatory commands. An important part of Jescheniak and Levelt's argument was that they found a stable WFE in a picture naming task, which requires complete lexical access, but not in a gender decision task, which only requires access to the words' lemmas and not their word forms. We report two attempts to replicate this pattern, one with new materials, and one with Jescheniak and Levelt's orginal pictures. In both studies we found a strong WFE when the pictures were shown for the first time, but much weaker effects on their second and third presentation. Importantly these patterns were seen in both the picture naming and the gender decision tasks, suggesting that either word frequency does not exclusively affect word form retrieval, or that the gender decision task does not exclusively tap lemma access.

Déjà-lu: When Orthographic Representations are Generated in the Absence of Orthography

When acquiring novel spoken words, English-speaking children generate preliminary orthographic representations even before seeing the words' spellings (Wegener et al., 2018). Interestingly, these orthographic skeletons are generated even when novel words' spellings are uncertain, at least in transparent languages like Spanish (Jevtović et al., 2022). Here we investigate whether this process depends on the orthographic rules of the language, and specifically, whether orthographic skeletons for words with uncertain spellings are generated even when the probability of generating an incorrect representation is high. Forty-six French adults first acquired novel words via aural instruction and were then presented with words' spellings in a self-paced reading task. Importantly, novel words were presented in their unique (consistent words) or one of their two possible spellings (preferred and unpreferred inconsistent words). A significant reading advantage observed for aurally acquired words indicates that participants indeed generated orthographic representations before encountering novel words' spellings. However, while no differences in reading times were found for aurally acquired words with unique and those presented in their preferred spellings, unpreferred spellings yielded significantly longer reading times. This shows that orthographic skeletons for words with multiple spellings were generated even in a language in which the risk of generating an incorrect representation is high. This finding raises a possibility that generating orthographic skeletons during spoken word learning may be beneficial. In line with this conclusion is the finding showing that - in interaction with good phonological short-term memory capacity - generating orthographic skeletons is linked to better word recall.

Web-based language production experiments: Semantic interference assessment is robust for spoken and typed response modalities

For experimental research on language production, temporal precision and high quality of the recorded audio files are imperative. These requirements are a considerable challenge if language production is to be investigated online. However, online research has huge potential in terms of efficiency, ecological validity and diversity of study populations in psycholinguistic and related research, also beyond the current situation. Here, we supply confirmatory evidence that language production can be investigated online and that reaction time (RT) distributions and error rates are similar in written naming responses (using the keyboard) and typical overt spoken responses. To assess semantic interference effects in both modalities, we performed two pre-registered experiments (n = 30 each) in online settings using the participants' web browsers. A cumulative semantic interference (CSI) paradigm was employed that required naming several exemplars of semantic categories within a seemingly unrelated sequence of objects. RT is expected to increase linearly for each additional exemplar of a category. In Experiment 1, CSI effects in naming times described in lab-based studies were replicated. In Experiment 2, the responses were typed on participants' computer keyboards, and the first correct key press was used for RT analysis. This novel response assessment yielded a qualitatively similar, very robust CSI effect. Besides technical ease of application, collecting typewritten responses and automatic data preprocessing substantially reduce the work load for language production research. Results of both experiments open new perspectives for research on RT effects in language experiments across a wide range of contexts. JavaScript- and R-based implementations for data collection and processing are available for download.

Semantic Cues Modulate Children’s and Adults’ Processing of Audio-Visual Face Mask Speech

During the COVID-19 pandemic, questions have been raised about the impact of face masks on communication in classroom settings. However, it is unclear to what extent visual obstruction of the speaker’s mouth or changes to the acoustic signal lead to speech processing difficulties, and whether these effects can be mitigated by semantic predictability, i.e., the availability of contextual information. The present study investigated the acoustic and visual effects of face masks on speech intelligibility and processing speed under varying semantic predictability. Twenty-six children (aged 8-12) and twenty-six adults performed an internet-based cued shadowing task, in which they had to repeat aloud the last word of sentences presented in audio-visual format. The results showed that children and adults made more mistakes and responded more slowly when listening to face mask speech compared to speech produced without a face mask. Adults were only significantly affected by face mask speech when both the acoustic and the visual signal were degraded. While acoustic mask effects were similar for children, removal of visual speech cues through the face mask affected children to a lesser degree. However, high semantic predictability reduced audio-visual mask effects, leading to full compensation of the acoustically degraded mask speech in the adult group. Even though children did not fully compensate for face mask speech with high semantic predictability, overall, they still profited from semantic cues in all conditions. Therefore, in classroom settings, strategies that increase contextual information such as building on students’ prior knowledge, using keywords, and providing visual aids, are likely to help overcome any adverse face mask effects.