This cross-linguistic study examined inflectional morphology and word order in Moroccan Arabic (MA) and English-speaking persons with agrammatic aphasia (PWAA). MA has rich verbal morphology and flexible word order, whereas English has limited morphology and rigid order, providing a strong test of accounts of agrammatism. The closed-class deficit hypothesis predicts selective impairment of inflections with preserved word order, while the syntactic deficit hypothesis (SDH) attributes the disorder to a syntactic impairment affecting both domains. Speech from nine MA speakers (four PWAA, five typical participants (TP)) and ten English speakers (five PWAA, five TP) was analyzed. In both languages, PWAA showed deficits in morphology and word order, supporting the SDH. Severity patterns differed: MA-speaking PWAA trended toward greater morphological impairment, whereas English-speaking PWAA showed greater word order disruption. MA-speaking PWAA also deviated from TP's canonical VSO pattern, suggesting compensatory subject-initial strategies. Findings support a core syntactic deficit modulated by language typology.

Previous research showed that people enumerate objects faster and more accurately when they form clusters, a phenomenon called "groupitizing". While mainly studied visually, its dependence on vision is unclear. Congenitally blind (CB) individuals provide a critical test: if vision is essential, CB people should lack groupitizing; if not, they may apply it across modalities, potentially outperforming sighted participants. We compared CB and sighted adults on an auditory groupitizing task, based on the estimation of 5-12 pure tones presented either randomly or grouped by temporal proximity. Both groups showed lower errors and higher precision for grouped sequences, confirming that groupitizing can emerge without visual experience. Importantly, for larger numerosities, sighted individuals' grouping benefit decreased, whereas CB participants maintained robust advantages across all set sizes. These findings suggest that groupitizing relies on amodal perceptual mechanisms and that congenital blindness may enhance auditory enumeration strategies.

This study explores bimodal cross-language interactions in the context of sign language vocabulary learning. Specifically, whether such interactions occur during oral language production, and whether they extend to an oral language not directly used in the training. Across three sessions, hearing Catalan-Spanish non-signers were trained on Catalan Sign Language (LSC) signs through an associative learning task (LSC-written Catalan). Participants subsequently performed an LSC-to-Catalan prime translation task with primes written in Catalan or Spanish. The primes were either phonologically related or unrelated to the target signs via their LSC translations. Behaviourally, LSC phonologically related word primes elicited faster translations, regardless of the prime language. Conversely, the N400 ERP component showed prime language-dependent effects. N400 differences were initially limited to Catalan, the training language, and extended to Spanish by the third session. These findings highlight the dynamic interplay between oral and sign languages during early sign language learning.

This study investigates cross-linguistic asymmetries in noun and verb production in a Hindi-English bilingual with Broca's aphasia (RZ), focusing on the influence of task demands (narrative vs. noun naming, verb naming, repetition), morphological richness, and code-switching (frequency and type). RZ exhibited features of agrammatism in both languages, with more pronounced deficits in English. RZ showed grammatical class asymmetries in noun-verb production across tasks. He produced more verbs than nouns in Hindi in the narrative task, likely due to its rich morphology, while showing comparable noun-verb production in naming. Verb retrieval remained consistently impaired in English across tasks. RZ frequently but rigidly code-switched, mainly inserting English nouns within Hindi matrix structure and used bilingual compound verbs, suggesting a strategy to compensate for lexical deficits in Hindi and morphosyntactic challenges in English. These findings underscore the importance of language typology and task demands in shaping aphasic symptomatology in bilinguals.

Hemianopic patients with visual field defects due to cerebral damage sometimes demonstrate residual visual capacities in their contralesional field, known as blindsight. Its associated subjective experience remains poorly understood. We developed a novel task for patients combining forced-choice detection and discrimination paradigms with a confidence scale to assess individual objective sensitivity and metacognitive sensitivity and efficiency, addressing the unique challenges of probing subjective experience in hemianopia, where nature is rarely evident. Four patients participated, revealing distinct perceptual and metacognitive profiles. One patient exhibited remarkable contralesional objective capacities and some preserved metacognition despite rarely reporting perception, suggesting subjective content may not reflect the stimulus itself. Concomitantly, preliminary results reveal ipsilesional metacognitive impairments in three patients despite optimal objective performance. Overall, we present the first application of state-of-the-art metacognitive measures to hemianopia, bridging perceptual and metacognitive measures to explore subjective experience, offering new insights into blindsight and the dissociation between consciousness and metacognitive processes.

The cognitive underpinnings of stuttering are likely heterogeneous. Yet existing research typically focuses on testing only one or a few hypotheses and on group-level differences. Here, 12 adults who stutter (AWS) completed a comprehensive battery assessing linguistic, auditory, somatosensory, rhythmic and motor abilities supporting fluent speech. Each AWS's performance was compared with that of 20 control participants to establish individual cognitive profiles. The AWS showed heterogeneous profiles; however, all exhibited some form of orosensory tactile weakness, which, for some, was the only weakness observed. These findings demonstrate the value of individualized cognitive profiling in stuttering and suggest a stronger link between stuttering and orosensory tactile processing than previously recognized.

Rapid turn-taking in conversation suggests that speakers plan part of their turn in advance, but evidence for this is scarce. Using context-driven picture naming, we examined whether (a) speakers preplan lexical-semantic and phonological information at the word level in constraining sentential contexts, and (b) phonological preplanning encompasses the whole word. Analysis of naming response times (RTs) showed that constraining contexts enable preplanning of both lexical-semantic and phonological representations (Experiment 1). Using a picture-word interference version of the same task (Experiment 2), we found that speakers preplan the phonological form of the whole word, however, only the subset of constraining trials with the shortest RTs indicated preplanning. The results confirm previous findings from turn-taking, which suggested that speakers can complete later stages of lexical access in advance, but also highlight that the presence of preplanning varies from trial to trial.

This study investigates the predictive role of fine motor skills (FMS) and executive functions (EFs) in adult handwriting. While the initial stages of handwriting acquisition are assumed to rely upon executive functions and fine motor skills, the two components are no longer viewed as needed in automatized, expert handwriting. Thirty-three adults were assessed on dexterity, manual praxis, working memory, inhibition and flexibility to predict handwriting speed, legibility and fluency. The results showed that FMS significantly contributed to all aspects of handwriting performance, while flexibility predicted writing speed. These findings highlight that, even at a high level of expertise and automatization, handwriting remains a skill whose performance depends on executive and fine motor control capacities. They support a predictive coding model where internal models guide movement execution and monitoring. The study questions how motor and executive impairments may disrupt handwriting performance, highlighting the need for further research into this complex motor behaviour.

This study applies implicit structural priming as a novel treatment for sentence production in persons with aphasia (PWA), investigating the learning mechanism(s) that drive robust and enduring recovery. Sixteen PWA and 16 controls completed baseline, three training sessions, and 1-day and 1-week post-testing. Each participant received both alternating and single structure prime training conditions to test error-based versus repeated activation-based learning. Both groups showed significantly improved production and maintenance of trained and untrained target sentences in both training conditions. While controls showed greater gains following alternating prime structure training, single prime structure training resulted in greater improvements for PWA. These results suggest that structural priming is an effective training for aphasia. Additionally, to the extent that the different priming conditions reflected different mechanisms underlying the learning and access of impaired structure, increased base-level activation of target syntactic structure supports learning of grammatical encoding in aphasia more effectively than processing prime sentences with competing syntactic structures.

Speech monitoring abilities vary among individuals with stroke-induced aphasia, with brain lesion location as a potential factor. Left posterior temporal cortex (pLTC) regions are thought to be central to lexical access. We tested whether pLTC lesions affect the medial frontal action monitoring system, as indexed by the Error-Related Negativity (ERN), which has been implicated in inner speech monitoring. Electroencephalography was recorded during picture naming in 11 individuals with pLTC lesions (4 from each of two institutions included in EEG analyses), 7 with lesions sparing the pLTC (6 included), and 20 matched controls (14 included). Individuals with pLTC lesions were slower and less accurate than other groups. Individuals with lesions sparing the pLTC showed the expected ERN; individuals with pLTC lesions did not. Therefore, the medial frontal monitoring mechanism may be compromised when regions central to lexical access are damaged, indicating that pLTC-medial frontal interactions may support inner speech monitoring.

We investigated visual working memory (VWM) for faces and two novel non-face pattern types (Blobs and Mondrians) in individuals with developmental prosopagnosia (DP) and age-matched controls. Participants completed both simultaneous and sequential encoding tasks, judging whether a probe item matched one shown at encoding. DPs showed a consistent face disadvantage across both encoding types, while controls showed a face advantage, but only during simultaneous encoding. Compared to controls, DPs had impaired face VWM in both tasks but performed equivalently for abstract shapes and patterns. Face VWM impairments in DP were not exacerbated by increased memory load or updating demands, suggesting these deficits stem from face perception difficulties that affect encoding rather than general VWM mechanisms. Our group-based analyses were supplemented by individual case statistics. Overall, our findings indicate that DPs do not exhibit general VWM deficits, but rather specific difficulties with face processing across formats.

Primary progressive aphasia (PPA) is a neurodegenerative syndrome characterized by prominent language symptoms. Distinguishing between PPA variants, particularly non-fluent and logopenic variants, remains challenging. Language production is a crucial aspect of diagnosing PPA, with confrontation naming tests being commonly used. However, there are limitations to the use of confrontation naming alone and it is still unclear how confrontation naming relates to (semi-)spontaneous language production. Additionally, most studies have focused on English-speaking patients. This bias hinders a thorough understanding of PPA, as symptoms may vary across languages. In a pre-registered study, 49 Dutch-speaking individuals with PPA and 21 controls completed confrontation naming and a picture description task, from which we derived nine linguistic variables. The percentage of nouns was the only linguistic variable differentiating between the non-fluent and logopenic variants, highlighting the challenge of distinguishing these variants, also in a language other than English. We found a moderate correlation between confrontation naming and the frequency of nouns produced semi-spontaneously for the logopenic variant only. Together, these findings underscore the relevance of semi-spontaneous language production as a complement to confrontation naming for a more complete understanding of production abilities in PPA.

Although it is generally assumed that face recognition relies on holistic processing, whether face recognition deficits observed in Developmental Prosopagnosics (DPs) can be explained by impaired holistic processing is currently under debate. The mixed findings from past studies could be the consequence of DP's heterogeneous deficit nature and the use of different measures of holistic processing-the inversion, part-whole, and composite tasks-which showed a poor association among each other. The present study aimed to gain further insight into the role of holistic processing in DPs. Groups of DPs and neurotypicals completed three tests measuring holistic face processing and non-face objects (i.e., Navon task). At a group level, DPs showed (1) diminished, but not absent, inversion and part-whole effects, (2) comparable magnitudes of the composite face effect and (3) global precedence effect in the Navon task. However, single-case analyses showed that these holistic processing deficits in DPs are .

We critically examine the procedural deficit hypothesis (PDH) that proposes that a deficit in procedural (as opposed to declarative) learning underlies dyslexia and other developmental disorders. We first note that the existence of dissociated learning disorders (and multiple forms for each disorder) appears incompatible with a general deficit account. Moreover, the PDH formulation appears generally underspecified in terms of predictions to be tested. A particular focus is on the conceptualization of automatization. However, there are alternative views of automaticity, and comparing these different views helps frame the body of findings on the PDH. The insufficient PDH specification led to tasks touching on different skills and selecting target groups based on general diagnostic categories. Accordingly, several recent reviews and meta-analyses reported mixed patterns of findings and reached contradictory conclusions on the PDH. We propose avenues for future research to effectively examine the role of PDH in learning and other developmental disorders.

Many aspects of human performance require producing sequences of items in serial order. The current study takes a multiple-case approach to investigate whether the system responsible for serial order is shared across cognitive domains, focusing on working memory (WM) and word production. Serial order performance in three individuals with post-stroke language and verbal WM disorders (hereafter persons with aphasia, PWAs) were assessed using recognition and recall tasks for verbal and visuospatial WM, as well as error analyses in spoken and written production tasks to assess whether there was a tendency to produce the correct phonemes/letters in the wrong order. One PWA exhibited domain-specific serial order deficits in verbal and visuospatial WM. The PWA with verbal serial order WM deficit made more serial order errors than expected by chance in both repetition and writing-to-dictation tasks, whereas the other two PWAs showed no serial order deficits in verbal WM and production tasks. These findings suggest separable serial order systems for verbal and visuospatial WM and a shared system for serial order processing in verbal WM and word production. Implications for the domain-generality of WM, its connection to language production, and serial order processing across cognitive functionssc are discussed.

This narrative review explores the phenomenon of inner speech - mental speech without visible articulation - and its implications for cognitive science and clinical practice. Despite its importance, the many neural mechanisms underlying inner speech remain unclear. We propose classifying inner speech into monologic, dialogal, elicited, and spontaneous forms, and discuss related phenomenological and neural correlates theories. A literature review on PubMed (1990-2024) identified 83 studies. Dialogal forms recruit Theory of Mind networks, compared to monologic forms. Task-elicited inner speech activates the left inferior frontal gyrus more strongly, while spontaneous inner speech engages Heschl's gyrus, suggesting auditory involvement. Evidence regarding aphasia suggests inner speech may be partially preserved even when overt speech is impaired, offering a potential route for rehabilitation. Future research should also address the emotional aspects of inner speech, its role in psychopathology, and its developmental trajectory. Such studies may improve interventions for disorders related to dysfunctional inner speech. ACC: anterior cingulate cortex; ALE: activation likelihood estimation; AVH: auditory verbal hallucination; BMI: brain-machine interface; CD: corollary discharge; ConDialInt: consciousness-dialogue-intentionality; DES: descriptive experience sampling; DTI: diffusion tensor imaging; dPMC: dorsal premotor cortex; dmPFC: dorsomedial prefrontal cortex; IFG: inferior frontal gyrus; M1: primary motor cortex; MedFG: medial frontal gyrus; MFG: middle frontal gyrus; MTG: middle temporal gyrus; MRI: magnetic resonance imaging; preSMA: presupplementary motor area; PrG: precentral gyrus; SMA: supplementary motor area; SMG: supramarginal gyrus; SPC: superior parietal cortex; SPL: superior parietal lobule; STG: superior temporal gyrus; STS: superior temporal sulcus; TVA: temporal vocal areas; ToM: theory of mind; vmPFC: ventromedial prefrontal cortex.

Lexical retrieval is commonly studied in the context of single words, even though words are usually produced within sentences. We present a framework for investigating the interplay between lexical retrieval, argument structure, and morphology. We propose a model for the retrieval of alternating-verbs, which, in Hebrew, are morphologically marked based on argument structure. We tested 23 Hebrew-speakers with aphasia, first identifying their functional locus of impairment within a lexical retrieval model for single words, and then administering a test battery to assess their production of alternating verbs within sentences. We found that the conceptual system, the semantic lexicon, the syntactic lexicon, the phonological output lexicon, and the phonological output buffer, each plays a unique role in retrieving morphologically-complex verbs, yielding a different error pattern when impaired. These error patterns are predicted by the proposed model for retrieval of alternating verbs with their argument structure and morphology.

Historically, memory and language have been seen as separate cognitive functions and studied in isolation. To date, it remains an open question to what extent these cognitive domains are related. Here, we present the interdisciplinary discussions from the 42nd European Workshop on Cognitive Neuropsychology around the central question of how we should see the relationship between the domains of language and memory. We discuss relevant empirical evidence from the fields of cognitive psychology, cognitive neuropsychology, and cognitive neuroscience and take a philosophical perspective on this central question, considering issues such as how to weigh different types of evidence and how to conceptualize the relationship between language and memory. We conclude that elucidating questions about the nature of the relationship between language and memory requires not only more empirical data, but also parallel conceptual development.

Aphantasia is a form of neurodivergence characterized by an absence of voluntary mental imagery. This absence affects not only basic cognition but also the processing of complex semantic content such as object colour, which typically relies on both visual and verbal representations. According to the Dual Coding Theory (DCT), combining these representations enhances semantic processing. However, this advantage has not been fully investigated under cross-modal conditions. To address this, we tested 24 individuals with congenital aphantasia and 22 controls on object colour decision and retrieval tasks using picture and word stimuli. Unlike controls, individuals with aphantasia showed no benefit from picture-based learning when retrieving colour via words. While their accuracy was unimpaired, their response efficiency was reduced. These findings support DCT and demonstrate the importance of visual imagery in facilitating cross-modal retrieval of object colour under verbal conditions.

Dyslexia is a multifaceted condition with diverse manifestations, yet assessment tools too often target limited subtypes, creating diagnostic gaps. This study examines the progression of dyslexia-related reading errors across primary school in typically developing readers, using the Mariette, a French nonsense-text reading screener. Analysis of 812 French children (grades 1-5) revealed systematic decreases in error rates with age, following distinct developmental trajectories. Regularizations of irregular words, misapplication of contextual rules and misreading of digraphs predominated in early grades, while voicing errors nearly disappeared by Grade 2. Clinical testing of the Mariette with 18 struggling readers identified specific reading errors overlooked by standard dyslexia assessments. These findings demonstrate the value of precise error analysis for understanding developmental reading patterns and tailoring targeted educational interventions. By comparing typical and clinical populations, this research advances our understanding of dyslexia's cognitive mechanisms while advocating for more comprehensive diagnostic approaches.

Previous studies have demonstrated that classifiers associated with nouns are activated during lexical access. Shape classifiers, a specific type, incorporate visual shape information. This study examined how visual shape information in classifiers is processed during the production of classifier-noun phrases by native Chinese speakers. Participants performed a picture-naming task using the blocked cyclic naming paradigm, where classifier congruency and shape similarity were manipulated. Behavioural results revealed a classifier congruency effect, with slower reaction times for classifier-incongruent conditions, and a shape interference effect, where classifiers with similar shapes slowed responses. EEG analysis showed that classifier-incongruent conditions elicited more positive voltage amplitudes than congruent ones, while shape-dissimilar conditions produced more negative amplitudes compared to shape-similar conditions after 300 ms post-stimulus. These findings indicated that classifiers were activated when producing noun phrases in a blocked cyclic naming paradigm. Moreover, visual shape information embedded in classifiers was processed during the production of classifier-noun phrases.

Speech errors in typical and atypical populations provide an important empirical foundation for theories of speech production and their neural bases. Studies of consonant speech errors have provided important evidence for cascading activation, c.f. categorical selection, across linguistic representations. Specifically, consonant errors show a phonetic trace effect whereby the phonetic details of the errorful consonant retain an influence of the intended sound. In this paper, we extend an experimental paradigm for speech error elicitation to vowels. Results show that vowel errors are similar to consonants in that errors differ systematically from their canonical (non-error) counterparts on the dimensions of primary control in articulation, as indexed by formant measures. We also found that vowel errors were not consistently different from canonical productions on the dimension of duration. We discuss the implications of these results for models that implement cascading activation to relate speech output to higher levels of language production.

Several studies have observed double dissociations in the production of nouns and verbs in persons with aphasia. However, whether or not these dissociations point to grammatical class being a principle of organization in the brain remains contested. Cross-linguistic considerations are important for drawing conclusions in this regard. As such, this study provides the first exploration of Gulf Arabic, a non-concatenative language with complex morphology. Utilizing Bayesian approach, the study tested for dissociations in 8 Gulf Arabic-speaking persons with aphasia (PWA) performance on object and action naming tasks. A double dissociation of nouns and verbs was found across the group, with 4 individuals exhibiting preserved action naming and impaired object naming, while 1 demonstrated the opposite pattern. Further error analysis and theoretical discussion are provided, considering existing explanations for dissociation phenomena in light of our novel findings within the understudied domain of Gulf Arabic aphasia.

This study examined how semantic characteristics of verbal cues impact naming, focusing on cue distinctiveness, defined as the degree to which a semantic cue is uniquely associated with a target. Using a novel naming paradigm, we presented word pairs representing semantic features to elicit naming responses. These verbal cues were categorized as distinctive (e.g., "moo-milk" for COW) or shared (e.g., "feathers-pink" for FLAMINGO), and targets were classified as living or nonliving. Distinctive cues significantly improved naming accuracy and speed, with a greater benefit for living items. A follow-up analysis examined graded effects of distinctiveness while accounting for associative strength. Both cue distinctiveness and associative strength facilitated naming speed overall. However, when distinctiveness was residualized against associative strength, the effect remained stronger for living items, although other semantic factors are also discussed. This study introduces a novel paradigm for evaluating semantic factors on naming, with potential future application to clinical populations.

Coma and disorders of consciousness (DoC) are common manifestations of acute severe brain injuries. Research into their neuroanatomical basis can be traced from Hippocrates to the present day. Lesions causing DoC have traditionally been conceptualized as decreasing "alertness" from damage to the ascending arousal system, and/or, reducing level of "awareness" due to structural or functional impairment of large-scale brain networks. Within this framework, pharmacological and neuromodulatory interventions to promote recovery from DoC have hitherto met with limited success. This is partly due to inter-individual heterogeneity of brain injury patterns, and an incomplete understanding of brain network properties that characterize consciousness. Advances in multiscale computational modelling of brain dynamics have opened a unique opportunity to explore the causal mechanisms of brain activity at the biophysical level. These models can provide a novel approach for selection and optimization of potential interventions by simulation of brain network dynamics individualized for each patient.

The role of stems and prefixes in complex nonword reading was investigated in unimpaired readers and five individuals with acquired dyslexia. All participants completed a reading aloud task (and the reading impaired individuals also completed a repetition task) with four different types of nonwords: prefix + stem (), non-prefix + stem (), prefix + non-stem (), non-prefix + non-stem (); and prefixed and non-prefixed filler words. The unimpaired readers responded fastest to nonwords containing two morphemes (prefix + stem), slower to nonwords with one morpheme (non-prefix + stem; prefix + non-stem), and slowest in the non-morphemic control condition (non-prefix + non-stem), providing evidence for the added benefit of prefixes and stems during reading. The five reading impaired individuals showed facilitatory morpheme effects across both tasks, but stem-effects were more robust than affix-effects. There was no difference between the prefixed and non-prefixed words in any of the data. The impact of morphological structure on nonword reading and repetition points to the important role of morphemes across different modalities..

The exploration of naming error patterns in aphasia provides insights into the cognitive processes underlying naming performance. We investigated how semantic and phonological abilities correlate and how they influence naming performance in aphasia. Data from 296 individuals with aphasia, drawn from the Moss Aphasia Psycholinguistics Project Database, were analyzed using a structural equation model. The model incorporated latent variables for semantics and phonology and manifest variables for naming accuracy and error patterns. There was a moderate positive correlation between semantics and phonology after controlling for overall aphasia severity. Both semantic and phonological abilities influenced naming accuracy. Semantic abilities negatively related to semantic, mixed, unrelated errors, and no responses. Interestingly, phonology positively affected semantic errors. Additionally, phonological abilities negatively related to each of phonological and neologism errors. These results highlight the role of semantic and phonological skills on naming performance in aphasia and reveal a relationship between these cognitive processes.

Blind participants tend to encode sensory details of encounters with stimuli they will later recall or recollect, whereas sighted individuals tend to abstract meaning from sensory information when encoding memories. Here we ask whether blind individuals' use of perceptual in addition to semantic encoding extends to a task-word learning-whose purpose amounts to semantically encoding the word and definition. After studying the definitions of spoken or written words, blind braille readers (  23) and sighted print readers ( = 20) re-encountered each word and indicated whether it was previously presented in the same modality. Analyses showed blind participants had better recall of modality even for words they had read (i.e., processed tactually), indicating their use of perceptual encoding may be automatically deployed in situations where it is unnecessary for the task. We recommend further research on individual differences in perceptual encoding within and across groups and its potential costs and benefits.

We present a case study detailing cognitive performance, functional neuroimaging, and effects of a hypothesis-driven treatment in a 10-year-old girl diagnosed with complete, isolated corpus callosum agenesis. Despite having average overall intellectual abilities, the girl exhibited profound surface dyslexia and dysgraphia. Spelling treatment significantly and persistently improved her spelling of trained irregular words, and this improvement generalized to reading accuracy and speed of trained words. Diffusion weighted imaging revealed strengthened intrahemispheric white matter connectivity of the left temporal cortex after treatment and identified interhemispheric connectivity between the occipital lobes, likely facilitated by a pathway crossing the midline via the posterior commissure. This case underlines the corpus callosum's critical role in lexical reading and writing. It demonstrates that spelling treatment may enhance interhemispheric connectivity in corpus callosum agenesis through alternative pathways, boosting the development of a more efficient functional organization of the visual word form area within the left temporo-occipital cortex.

Disorders of speech production, including stuttering, are relatively common and need to be recognized early in childhood. Despite calls for clear symptom definitions when studying stuttering in languages other than English, there is currently no standardized tool available for assessing stuttering in Arabic. The distinct differences between Arabic and English can offer insights into language-specific influences on the manifestation of disfluency. This study investigates the impact of Arabic phonological characteristics on the occurrence of stuttering in children who stutter (CWS). Speech samples were collected from 16 Arabic-speaking children. Findings indicate that words produced with stuttering tend to have greater phonological complexity than those produced fluently. A logistic regression revealed that word shape was the most influential factor in predicting stuttering. The study findings could inform the diagnosis of stuttering among Arabic-speaking populations. The findings are discussed in comparison to previous research, with particular attention to the morphophonological features of Arabic.