Spoken Word Identification Research Articles

Spread the Word: Enhancing Replicability of Speech Research Through Stimulus Sharing.

The ongoing replication crisis within and beyond psychology has revealed the numerous ways in which flexibility in the research process can affect study outcomes. In speech research, examples of these "researcher degrees of freedom" include the particular syllables, words, or sentences presented; the talkers who produce the stimuli and the instructions given to them; the population tested; whether and how stimuli are matched on amplitude; the type of masking noise used and its presentation level; and many others. In this research note, we argue that even seemingly minor methodological choices have the potential to affect study outcomes. To that end, we present a reanalysis of six existing data sets on spoken word identification in noise to assess how differences in talkers, stimulus processing, masking type, and listeners affect identification accuracy. Our reanalysis revealed relatively low correlations among word identification rates across studies. The data suggest that some of the seemingly innocuous methodological details that differ across studies-details that cannot possibly be reported in text given the idiosyncrasies inherent to speech-introduce unknown variability that may affect replicability of our findings. We therefore argue that publicly sharing stimuli is a crucial step toward improved replicability in speech research. https://doi.org/10.23641/asha.21985907.

Spoken Word Recognition in Listeners with Mild Dementia Symptoms.

Difficulty understanding speech is a common complaint of older adults. In quiet, speech perception is often assumed to be relatively automatic. However, higher-level cognitive processes play a key role in successful communication in noise. Limited cognitive resources in adults with dementia may therefore hamper word recognition. The goal of this study was to determine the impact of mild dementia on spoken word recognition in quiet and noise. Participants were 53-86 years with (n = 16) or without (n = 32) dementia symptoms as classified by the Clinical Dementia Rating scale. Participants performed a word identification task with two levels of word difficulty (few and many similar sounding words) in quiet and in noise at two signal-to-noise ratios, +6 and +3 dB. Our hypothesis was that listeners with mild dementia symptoms would have more difficulty with speech perception in noise under conditions that tax cognitive resources. Listeners with mild dementia symptoms had poorer task accuracy in both quiet and noise, which held after accounting for differences in age and hearing level. Notably, even in quiet, adults with dementia symptoms correctly identified words only about 80% of the time. However, word difficulty was not a factor in task performance for either group. These results affirm the difficulty that listeners with mild dementia may have with spoken word recognition, both in quiet and in background noise, consistent with a role of cognitive resources in spoken word identification.

Orthographic influence on spoken word identification: Behavioral and fMRI evidence

The current study investigated behavioral and neuroimaging evidence for orthographic influences on auditory word identification. To assess such influences, the proportion of similar sounding words (i.e. phonological neighbors) that were also spelled similarly (i.e., orthographic neighbors) was computed for each auditorily presented word as the Orthographic-to-Phonological Overlap Ratio (OPOR). Speech intelligibility was manipulated by presenting monosyllabic words in multi-talker babble at two signal-to-noise ratios: + 3 and + 10 dB SNR. Identification rates were lower for high overlap words in the challenging + 3 dB SNR condition. In addition, BOLD contrast increased with OPOR at the more difficult SNR, and decreased with OPOR under more favorable SNR conditions. Both voxel-based and region of interest analyses demonstrated robust effects of OPOR in several cingulo-opercular regions. However, contrary to prior theoretical accounts, no task-related activity was observed in posterior regions associated with phonological or orthographic processing. We suggest that, when processing is difficult, orthographic-to-phonological feature overlap increases the availability of competing responses, which then requires additional support from domain general performance systems in order to produce a single response.

Effects of Auditory and Visual Priming on the Identification of Spoken Words.

This study examined the effects of preceding contextual stimuli, either auditory or visual, on the identification of spoken target words. Fifty-one participants (29% males, 71% females; mean age = 24.5 years, SD = 8.5) were divided into three groups: no context, auditory context, and visual context. All target stimuli were spoken words masked with white noise. The relationships between the context and target stimuli were as follows: identical word, similar word, and unrelated word. Participants presented with context experienced a sequence of six context stimuli in the form of either spoken words or photographs. Auditory and visual context conditions produced similar results, but the auditory context aided word identification more than the visual context in the similar word relationship. We discuss these results in the light of top-down processing, motor theory, and the phonological system of language.

The Impact of Orthographic Consistency on German Spoken Word Identification

An auditory lexical decision experiment was conducted to find out whether sound-to-spelling consistency has an impact on German spoken word processing, and whether such an impact is different at different stages of reading development. Four groups of readers (school children in the second, third and fifth grades, and university students) participated in a lexical decision experiment. They were presented with spoken monosyllabic words that had either orthographically consistent rimes (i.e., there is only one spelling for the phonological rime) or orthographically inconsistent rimes (i.e., there are words with the same phonological rime but a different rime spelling). Results reveal that sound-to-spelling consistency influences spoken word processing in early and advanced stages of reading development in German. This finding indicates that word spelling knowledge gets intertwined with knowledge about phonological word forms early during reading acquisition.

The Hardest Butter to Button: Immediate Context Effects in Spoken Word Identification

According to some theories, the context in which a spoken word is heard has no impact on the earliest stages of word identification. This view has been challenged by recent studies indicating an interactive effect of context and acoustic similarity on language-mediated eye movements. However, an alternative explanation for these results is that participants looked less at acoustically similar objects in constraining contexts simply because they were looking more at other objects that were cued by the context. The current study addressed this concern whilst providing a much finer grained analysis of the temporal evolution of context effects. Thirty-two adults listened to sentences while viewing a computer display showing four objects. As expected, shortly after the onset of a target word (e.g., "button") in a neutral context, participants saccaded preferentially towards a cohort competitor of the word (e.g., butter). This effect was significantly reduced when the preceding verb made the competitor an unlikely referent (e.g., "Sam fastened the button"), even though there were no other contextually congruent objects in the display. Moreover, the time-course of these two effects was identical to within approximately 30 ms, indicating that certain forms of contextual information can have a near-immediate effect on word identification.

Achieving constancy in spoken word identification: Time course of talker normalization

This event-related potential (ERP) study examines the time course of context-dependent talker normalization in spoken word identification. We found three ERP components, the N1 (100–220ms), the N400 (250–500ms) and the Late Positive Component (500–800ms), which are conjectured to involve (a) auditory processing, (b) talker normalization and lexical retrieval, and (c) decisional process/lexical selection respectively. Talker normalization likely occurs in the time window of the N400 and overlaps with the lexical retrieval process. Compared with the nonspeech context, the speech contexts, no matter whether they have semantic content or not, enable listeners to tune to a talker’s pitch range. In this way, speech contexts induce more efficient talker normalization during the activation of potential lexical candidates and lead to more accurate selection of the intended word in spoken word identification.

Let's All Speak Together! Exploring the Masking Effects of Various Languages on Spoken Word Identification in Multi-Linguistic Babble

This study aimed to characterize the linguistic interference that occurs during speech-in-speech comprehension by combining offline and online measures, which included an intelligibility task (at a −5 dB Signal-to-Noise Ratio) and 2 lexical decision tasks (at a −5 dB and 0 dB SNR) that were performed with French spoken target words. In these 3 experiments we always compared the masking effects of speech backgrounds (i.e., 4-talker babble) that were produced in the same language as the target language (i.e., French) or in unknown foreign languages (i.e., Irish and Italian) to the masking effects of corresponding non-speech backgrounds (i.e., speech-derived fluctuating noise). The fluctuating noise contained similar spectro-temporal information as babble but lacked linguistic information. At −5 dB SNR, both tasks revealed significantly divergent results between the unknown languages (i.e., Irish and Italian) with Italian and French hindering French target word identification to a similar extent, whereas Irish led to significantly better performances on these tasks. By comparing the performances obtained with speech and fluctuating noise backgrounds, we were able to evaluate the effect of each language. The intelligibility task showed a significant difference between babble and fluctuating noise for French, Irish and Italian, suggesting acoustic and linguistic effects for each language. However, the lexical decision task, which reduces the effect of post-lexical interference, appeared to be more accurate, as it only revealed a linguistic effect for French. Thus, although French and Italian had equivalent masking effects on French word identification, the nature of their interference was different. This finding suggests that the differences observed between the masking effects of Italian and Irish can be explained at an acoustic level but not at a linguistic level.

The influence of talker and foreign-accent variability on spoken word identification

In spoken word identification and memory tasks, stimulus variability from numerous sources impairs performance. In the current study, the influence of foreign-accent variability on spoken word identification was evaluated in two experiments. Experiment 1 used a between-subjects design to test word identification in noise in single-talker and two multiple-talker conditions: multiple talkers with the same accent and multiple talkers with different accents. Identification performance was highest in the single-talker condition, but there was no difference between the single-accent and multiple-accent conditions. Experiment 2 further explored word recognition for multiple talkers in single-accent versus multiple-accent conditions using a mixed design. A detriment to word recognition was observed in the multiple-accent condition compared to the single-accent condition, but the effect differed across the language backgrounds tested. These results demonstrate that the processing of foreign-accent variation may influence word recognition in ways similar to other sources of variability (e.g., speaking rate or style) in that the inclusion of multiple foreign accents can result in a small but significant performance decrement beyond the multiple-talker effect.

The influence of talker and accent variability on spoken word identification and discrimination

In spoken word identification and memory tasks, stimulus variability from numerous sources impairs performance. The phonetic-relevance hypothesis (Sommers etal., 2006) proposes that only acoustic properties that influence phonetic perception (e.g., speaking rate) cause decrements in perception. In the current study, the influence of foreign-accent variability on identification and discrimination of spoken words was evaluated. In experiment 1, word identification in noise was tested in single-talker and two multiple-talker conditions: multiple talkers with the same accent or multiple talkers with different accents. In experiment 2, participants discriminated minimal pairs from a single talker, different talkers with the same accent, and talkers with different accents. Identification and discrimination performance was the highest in the single-talker conditions, but there was no difference between the single-accent and multiple-accent conditions. However, reaction time in the discrimination task was the highest in the multiple accent condition with no difference between the single-talker and single-accent conditions. Thus, the presence of multiple accents does not decrease accuracy beyond the multiple-talker effect, but processing time increases with the addition of multiple accents. These results provide partial support for the phonetic-relevance hypothesis. [Work supported by NIDCD R21DC010027 and Indiana University.]

An Assessment of Behavioral Dynamic Information Processing Measures in Audiovisual Speech Perception

Research has shown that visual speech perception can assist accuracy in identification of spoken words. However, little is known about the dynamics of the processing mechanisms involved in audiovisual integration. In particular, architecture and capacity, measured using response time methodologies, have not been investigated. An issue related to architecture concerns whether the auditory and visual sources of the speech signal are integrated “early” or “late.” We propose that “early” integration most naturally corresponds to coactive processing whereas “late” integration corresponds to separate decisions parallel processing. We implemented the double factorial paradigm in two studies. First, we carried out a pilot study using a two-alternative forced-choice discrimination task to assess architecture, decision rule, and provide a preliminary assessment of capacity (integration efficiency). Next, Experiment 1 was designed to specifically assess audiovisual integration efficiency in an ecologically valid way by including lower auditory S/N ratios and a larger response set size. Results from the pilot study support a separate decisions parallel, late integration model. Results from both studies showed that capacity was severely limited for high auditory signal-to-noise ratios. However, Experiment 1 demonstrated that capacity improved as the auditory signal became more degraded. This evidence strongly suggests that integration efficiency is vitally affected by the S/N ratio.

Learning Representations of Wordforms With Recurrent Networks: Comment on Sibley, Kello, Plaut, & Elman (2008)

Sibley etal. (2008) report a recurrent neural network model designed to learn wordform representations suitable for written and spoken word identification. The authors claim that their sequence encoder network overcomes a key limitation associated with models that code letters by position (e.g., CAT might be coded as C-in-position-1, A-in-position-2, T-in-position-3). The problem with coding letters by position (slot-coding) is that it is difficult to generalize knowledge across positions; for example, the overlap between CAT and TOMCAT is lost. Although we agree this is a critical problem with many slot-coding schemes, we question whether the sequence encoder model addresses this limitation, and we highlight another deficiency of the model. We conclude that alternative theories are more promising.

The activation of embedded words in spoken word identification is robust but constrained: Evidence from the picture-word interference paradigm.

Three picture-word interference (PWI) experiments assessed the extent to which embedded subset words are activated during the identification of spoken superset words (e.g., bone in trombone). Participants named aloud pictures (e.g., brain) while spoken distractors were presented. In the critical condition, superset distractors contained a semantically related embedded word (e.g., charm, which contains arm). In Experiment 1, supersets and subsets differed by one phoneme (charm/arm) and interference effects were obtained when subsets were embedded at the beginning or end of the superset. In Experiment 2, the subsets and supersets differed by three or more phonemes. Interference was obtained for final embedded words aligned with a syllable boundary of the superset (acrobat/bat) but not otherwise (pioneer/ear). In Experiment 3, the size of these PWI effects was compared to the effects obtained with the subset words presented in isolation. The implication of these findings for theories of speech perception and production are discussed. We also consider the possible advantages of the PWI task over cross-modal priming and "visual-world" procedures when studying these issues.

The Psychology of Literacy: New Developments

In this paper we begin by addressing the concept of literacy, which for us is “the ensemble of representations and processes that an individual acquires as an obligatory and direct consequence of learning to read and write”, and distinguish it from other conceptions that can be found in the literature. In the remaining of the paper we discuss some of the most recent developments in the psychology of literacy: (i) the occurrence of language differences in the progressive mastery of the orthographic code; (ii) the neural basis of written word identification; and (iii) the consequences of literacy on spoken word identification.

Stimulus variability and the phonetic relevance hypothesis: Effects of variability in speaking style, fundamental frequency, and speaking rate on spoken word identification

Three experiments were conducted to examine the effects of trial-to-trial variations in speaking style, fundamental frequency, and speaking rate on identification of spoken words. In addition, the experiments investigated whether any effects of stimulus variability would be modulated by phonetic confusability (i.e., lexical difficulty). In Experiment 1, trial-to-trial variations in speaking style reduced the overall identification performance compared with conditions containing no speaking-style variability. In addition, the effects of variability were greater for phonetically confusable words than for phonetically distinct words. In Experiment 2, variations in fundamental frequency were found to have no significant effects on spoken word identification and did not interact with lexical difficulty. In Experiment 3, two different methods for varying speaking rate were found to have equivalent negative effects on spoken word recognition and similar interactions with lexical difficulty. Overall, the findings are consistent with a phonetic-relevance hypothesis, in which accommodating sources of acoustic-phonetic variability that affect phonetically relevant properties of speech signals can impair spoken word identification. In contrast, variability in parameters of the speech signal that do not affect phonetically relevant properties are not expected to affect overall identification performance. Implications of these findings for the nature and development of lexical representations are discussed.

Variability in bimodal spoken language processing by native and nonnative speakers of English: A closer look at effects of speech style

Experiments using the gating paradigm investigated the influence of speech style (unscripted vs. scripted), visual cues from a talker’s face (auditory-visual vs. auditory-only presentations), word length (one vs. two syllables) and initial consonant (IC) visual category in spoken word identification by native- (NSs) and nonnative speakers (NNSs) of English. Two talkers were videotaped in separate conversations with the author on various topics (unscripted speech) with the camera focused on the talker’s face. They later recorded selected sentences as scripted speech. In Experiment 1, target words were excised from sentences, gated, and presented to NSs and NNSs. Results for both groups showed significantly earlier identification with visual cues and for bisyllabic vs. monosyllabic words, and significant interaction of speech style and IC category across talkers. For Talker A, identification of unscripted monosyllabic words in some IC categories was earlier than the scripted versions. For this talker, words beginning with / ▪/ were identified later than others; for Talker B, they were the earliest to be identified. For NNSs, the AV advantage was accentuated for words beginning with / ▪, w, ▪/ in unscripted speech by Talker A, and for / ▪/- and / l/-initial words in both speech styles by Talker B. Experiment 2 presented the preceding sentence context with the gated word to NSs. Results revealed earlier identification in AV presentation and for unscripted vs. scripted words by Talker A. With context, word length was not significant. Findings highlight the priming role of visual cues, and the talker- and context-dependent nature of bimodal spoken language processing, but do not support a strict conversational-clear speech dichotomy.

Do words go to sleep? Exploring consolidation of spoken forms through direct and indirect measures

We address the notion of integration of new memory representations and the potential dependence of this phenomenon on sleep, in light of recent findings on the lexicalization of spoken words. A distinction is introduced between measures tapping directly into the strength of the newly acquired knowledge and indirect measures assessing the influence of this knowledge on spoken word identification.

Implicit word activation during pre‐recognition processing influences correct recognition and estimates of presentation frequency

False recognition of new test words is higher for experimental lures (e.g., universal) with initial phonemes identical to studied words (e.g., university) than for control lures. A proposed mechanism to explain this phenomenon involves implicit activation of potential solution words during the brief period of uncertainty immediately following onset of a spoken study word. Two experiments examined whether the presumed pre‐recognition processing during the stimulus discovery phase of spoken word identification increased familiarity of a studied word, thereby increasing correct recognitions and estimates of presentation frequency. Critical test words were presented a single time during study in Experiment 1, and their phonologically related words were presented one, two, or three times. Correct recognition and frequency estimates of targets were enhanced by multiple presentations of associates sharing initial phonemes. Experiment 2 provided a replication with five repetitions of phonological associates during study and two study presentations of critical test words. The results of these two experiments confirmed a necessary theoretical consequence of the implicit activation mechanism that has been invoked to explain the effects of phonological similarity on false recognition.

Musical Features That Facilitate Melody Identification: How Do You Know It's “Your” Song When They Finally Play It?

What information do listeners use to “Name that Tune”? This question was investigated in a two-phase experiment. In Phase 1, the participants heard familiar melodies that were played on a note-by-note basis until they were identified. In Phase 2, each note of the melody was analyzed along a variety of musical dimensions. Multiple regression analyses determined which musical characteristics predicted identification performance. Identification was most strongly associated with notes located at phrase boundaries, notes that completed alternating sequences of rising and falling pitches, and metrically accented notes. As well, identification peaked after listeners heard moderate amounts of information (i.e., 5–7 notes). The data suggest that melody identification is a holistic, all-or-none process and that parallels can be drawn between melody and spoken word identification. Implications for current theories, future research, and the relationship between music perception and melody identification are discussed.

Lexical competition and spoken word identification in schizophrenia

Schizophrenia (SZ) is a disorder associated with dysfunction of the neural substrate involved in lexical processing, the superior temporal lobe (STL). We examined whether the detrimental effects of lexical competition on word identification are exaggerated in SZ. The results showed that SZ patients were less able than controls to identify words coming from high lexical competitor environments (i.e., high density and high frequency lexical neighborhoods). They did not differ from controls, however, in identifying words coming from low lexical competitor environments (i.e., low density and low frequency lexical neighborhoods). Further, when SZ patients misidentified high competitor words, the erroneously reported words were of a significantly higher frequency than those reported by controls when they misidentified high competitor words (i.e., the HF misidentification bias). Identification accuracy for high competitor words negatively correlated with amount of thought disorder, and the HF misidentification bias positively correlated with the lifetime frequency and severity of auditory hallucinations, the latter of which serving as presumed independent measures of STL dysfunction in SZ.