Neurophysiological signatures of prediction in language: A critical review of anticipatory negativities.

Recent event-related potential (ERP) studies in language comprehension converge in finding anticipatory negativities preceding words or word segments that can be pre-activated based on either sentence contexts or phonological cues. We review these findings from different paradigms in the light of evidence from other cognitive domains in which slow negative potentials have long been associated with anticipatory processes and discuss their potential underlying mechanisms. We propose that this family of anticipatory negativities captures common mechanisms associated with the pre-activation of linguistic information both within words and within sentences. Future studies could utilize these anticipatory negativities in combination with other, well-established ERPs, to simultaneously track prediction-related processes emerging at different time intervals (before and after the perception of pre-activated input) and with distinct time courses (shorter-lived and longer-lived cognitive operations).

Swift Prosodic Modulation of Lexical Access: Brain Potentials From Three North Germanic Language Varieties.

PURPOSE: According to most models of spoken word recognition, listeners probabilistically activate a set of lexical candidates, which is incrementally updated as the speech signal unfolds. Speech carries segmental (speech sound) as well as suprasegmental (prosodic) information. The role of the latter in spoken word recognition is less clear. We investigated how suprasegments (tone and voice quality) in three North Germanic language varieties affected lexical access by scrutinizing temporally fine-grained neurophysiological effects of lexical uncertainty and information gain. METHOD: Three event-related potential (ERP) studies were reanalyzed. In all varieties investigated, suprasegments are associated with specific word endings. Swedish has two lexical "word accents" realized as pitch falls with different timings across dialects. In Danish, the distinction is in voice quality. We combined pronunciation lexica and frequency lists to calculate estimates of lexical uncertainty about an unfolding word and information gain upon hearing a suprasegmental cue and the segment upon which it manifests. We used single-trial mixed-effects regression models run every 4 ms. RESULTS: Only lexical uncertainty showed solid results: a frontal effect at 150-400 ms after suprasegmental cue onset and a later posterior effect after 200 ms. While a model including only segmental information mostly performed better, it was outperformed by the suprasegmental model at 200-330 ms at frontal sites. CONCLUSIONS: The study points to suprasegmental cues contributing to lexical access over and beyond segments after around 200 ms in the North Germanic varieties investigated. Furthermore, the findings indicate that a previously reported "pre-activation negativity" predominantly reflects forward-looking processing. SUPPLEMENTAL MATERIAL: https://doi.org/10.23641/asha.25016486.

The predictive function of Swedish word accents.

Swedish lexical word accents have been repeatedly said to have a low functional load. Even so, the language has kept these tones ever since they emerged probably over a thousand years ago. This article proposes that the primary function of word accents is for listeners to be able to predict upcoming morphological structures and narrow down the lexical competition rather than being lexically distinctive. Psycho- and neurophysiological evidence for the predictive function of word accents is discussed. A novel analysis displays that word accents have a facilitative role in word processing. Specifically, a correlation is revealed between how much incorrect word accents hinder listeners' processing and how much they reduce response times when correct. Finally, a dual-route model of the predictive use of word accents with distinct neural substrates is put forth.

Native language experience shapes pre-attentive foreign tone processing and guides rapid memory trace build-up: An ERP study.

Language experience, particularly from our native language (L1), shapes our perception of other languages around us. The present study examined how L1 experience moulds the initial processing of foreign (L2) tone during acquisition. In particular, we investigated whether learners were able to rapidly forge new neural memory traces for novel tonal words, which was tracked by recording learners' ERP responses during two word acquisition sessions. We manipulated the degree of L1-L2 familiarity by comparing learners with a nontonal L1 (German) and a tonal L1 (Swedish) and by using tones that were similar (fall) or dissimilar (high, low, rise) to those occurring in Swedish. Our results indicate that a rapid, pre-attentive memory trace build-up for tone manifests in an early ERP component at ~50 ms but only at particularly high levels of L1-L2 similarity. Specifically, early processing was facilitated for an L2 tone that had a familiar pitch shape (fall) and word-level function (inflection). This underlines the importance of these L1 properties for the early processing of L2 tone. In comparison, a later anterior negativity related to the processing of the tones' grammatical content was unaffected by native language experience but was instead influenced by lexicality, pitch prominence, entrenchment, and successful learning. Behaviorally, learning effects emerged for all learners and tone types, regardless of L1-L2 familiarity or pitch prominence. Together, the findings suggest that while L1-based facilitation effects occur, they mainly affect early processing stages and do not necessarily result in more successful L2 acquisition at behavioral level.

Cortical and white matter correlates of language-learning aptitudes.

People learn new languages with varying degrees of success but what are the neuroanatomical correlates of the difference in language-learning aptitude? In this study, we set out to investigate how differences in cortical morphology and white matter microstructure correlate with aptitudes for vocabulary learning, phonetic memory, and grammatical inferencing as measured by the first-language neutral LLAMA test battery. We used ultra-high field (7T) magnetic resonance imaging to estimate the cortical thickness and surface area from sub-millimeter resolved image volumes. Further, diffusion kurtosis imaging was used to map diffusion properties related to the tissue microstructure from known language-related white matter tracts. We found a correlation between cortical surface area in the left posterior-inferior precuneus and vocabulary learning aptitude, possibly indicating a greater predisposition for storing word-figure associations. Moreover, we report negative correlations between scores for phonetic memory and axial kurtosis in left arcuate fasciculus as well as mean kurtosis, axial kurtosis, and radial kurtosis of the left superior longitudinal fasciculus III, which are tracts connecting cortical areas important for phonological working memory.

Cortical thickness and surface area of left anterior temporal areas affects processing of phonological cues to morphosyntax.

Lack of methods to experimentally assess the perceptual processing of sound features and allow one to measure differences in phonological proficiency has been a limitation for speech processing studies in native speakers. Tonal features associated with Swedish word-stems, word accents, which cue grammatical suffixes, constitute, however, such sound features that can be exploited to generate measures of reliance on morphosyntactically relevant phonological information during word processing. Specifically, there is a natural variance between native speakers in response time (RT) difference between phonologically valid and invalid word accent-suffix combinations that can be used to quantify perceptual phonological proficiency. This study uses ultra-high field magnetic resonance imaging (MRI) to investigate word accents as phonological cues to morphosyntactic meaning. The study adds to the understanding of the neural basis for both morphosyntactically relevant phonological cues by reporting correlations between differences in listeners' RT for validly and invalidly cued suffixes and cortical thickness in left anterior and middle temporal gyrus, and the left anterior superior temporal sulcus as well as cortical surface area in the left middle and inferior temporal gyri. The cortical areas studied are known constituents of the ventral speech processing stream, necessary for word and phrase recognition.

Different neural mechanisms for rapid acquisition of words with grammatical tone in learners from tonal and non-tonal backgrounds: ERP evidence.

Initial second language acquisition proceeds surprisingly quickly. Foreign words can sometimes be used within minutes after the first exposure. Yet, it is unclear whether such rapid learning also takes place for more complex, multi-layered properties like words with complex morphosyntax and/or tonal features, and whether it is influenced by transfer from the learners' native language. To address these questions, we recorded tonal and non-tonal learners' brain responses while they acquired novel tonal words with grammatical gender and number on two consecutive days. Comparing the novel words to repeated but non-taught pseudoword controls, we found that tonal learners demonstrated a full range of early and late event-related potentials in novel tonal word processing: an early word recognition component (~50 ms), an early left anterior negativity (ELAN), a left anterior negativity (LAN), and a P600. Non-tonal learners exhibited mainly late processing when accessing the meaning of the tonal words: a P600, as well as a LAN after an overnight consolidation. Yet, this group displayed correlations between pitch perception abilities and ELAN, and between acquisition accuracy and LAN, suggesting that certain features may lead to facilitated processing of tonal words in non-tonal learners. Furthermore, the two groups displayed indistinguishable performance at the behavioural level, clearly suggesting that the same learning outcome may be achieved through at least partially different neural mechanisms. Overall, the results suggest that it is possible to rapidly acquire words with grammatical tone and that transfer plays an important role even in very early second language acquisition.

Cortical thickness of Broca's area and right homologue is related to grammar learning aptitude and pitch discrimination proficiency.

Aptitude for and proficiency in acquiring new languages varies in the human population but their neural bases are largely unknown. We investigated the influence of cortical thickness on language learning predictors measured by the LLAMA tests and a pitch-change discrimination test. The LLAMA tests are first language-independent assessments of language learning aptitude for vocabulary, phonetic working memory, sound-symbol correspondence (not used in this study), and grammatical inferencing. Pitch perception proficiency is known to predict aptitude for learning new phonology. Results show a correlation between scores in a grammatical meaning-inferencing aptitude test and cortical thickness of Broca's area (r(30) = 0.65, p = 0.0202) and other frontal areas (r(30) = 0.66, p = 0.0137). Further, a correlation was found between proficiency in discriminating pitch-change direction and cortical thickness of the right Broca homologue (r(30) = 0.57, p = 0.0006). However, no correlations were found for aptitude for vocabulary learning or phonetic working memory. Results contribute to locating cortical regions important for language-learning aptitude.

Rapid syntactic pre-activation in Broca's area: Concurrent electrophysiological and haemodynamic recordings.

Listeners are constantly trying to predict what the speaker will say next. We concurrently measured the electrophysiological and haemodynamic correlates of syntactic pre-activation, investigating when and where the brain processes speech melody cues to upcoming word order structure. Pre-activation of syntactic structure was reflected in a left-lateralised pre-activation negativity (PrAN), which was subserved by Broca's area in the left inferior frontal gyrus, as well as the contiguous left anterior insula.

Cortical thickness of planum temporale and pars opercularis in native language tone processing.

The present study investigated the relationship between linguistic tone processing and cortical thickness of bilateral planum temporale (PT) and pars opercularis of the inferior frontal gyrus (IFGpo). Swedish tones on word stems function as cues to upcoming endings. Correlating structural brain imaging data with participants' response time patterns for suffixes, we found that thicker cortex in the left PT was associated with greater reliance on tones to anticipate upcoming inflections on real words. On inflected pseudoword stems, however, the cortical thickness of left IFGpo was associated with tone-suffix processing. Thus cortical thickness of the left PT might play a role in processing tones as part of stored representations for familiar speech segments, most likely when inflected forms are accessed as whole words. In the absence of stored representations, listeners might need to rely on morphosyntactic rules specifying tone-suffix associations, potentially facilitated by greater cortical thickness of left IFGpo.

Forehearing words: Pre-activation of word endings at word onset.

Occurring at rates up to 6-7 syllables per second, speech perception and understanding involves rapid identification of speech sounds and pre-activation of morphemes and words. Using event-related potentials (ERPs) and functional magnetic resonance imaging (fMRI), we investigated the time-course and neural sources of pre-activation of word endings as participants heard the beginning of unfolding words. ERPs showed a pre-activation negativity (PrAN) for word beginnings (first two segmental phonemes) with few possible completions. PrAN increased gradually as the number of possible completions of word onsets decreased and the lexical frequency of the completions increased. The early brain potential effect for few possible word completions was associated with a blood-oxygen-level-dependent (BOLD) contrast increase in Broca's area (pars opercularis of the left inferior frontal gyrus) and angular gyrus of the left parietal lobe. We suggest early involvement of the left prefrontal cortex in inhibiting irrelevant left parietal activation during lexical selection. The results further our understanding of the importance of Broca's area in rapid online pre-activation of words.

Tone-grammar association within words: Concurrent ERP and fMRI show rapid neural pre-activation and involvement of left inferior frontal gyrus in pseudoword processing.

Using a concurrent ERP/fMRI paradigm, we investigated how listeners take advantage of morphologically relevant tonal information at the beginning of words to predict and pre-activate likely word endings. More predictive, low tone word stems gave rise to a 'pre-activation negativity' (PrAN) in the ERPs, a brain potential which has previously been found to increase along with the degree of predictive certainty as regards how a word is going to end. It is suggested that more predictive, low tone stems lead to rapid access to word endings with processing subserved by the left primary auditory cortex as well as the supramarginal gyrus, while high tone stems - which are less predictive - decrease predictive certainty, leading to increased competition between activated word endings, which needs to be resolved by the left inferior frontal gyrus.

Stem Tones Pre-activate Suffixes in the Brain.

Results from the present event-related potentials (ERP) study show that tones on Swedish word stems can rapidly pre-activate upcoming suffixes, even when the word stem does not carry any lexical meaning. Results also show that listeners are able to rapidly restore suffixes which are replaced with a cough. Accuracy in restoring suffixes correlated positively with the amplitude of an anterior negative ERP elicited by stem tones. This effect is proposed to reflect suffix pre-activation. Suffixes that were cued by an incorrect tone elicited a left-anterior negativity and a P600, suggesting that the correct processing of the suffix is crucially tied to the activation of the preceding validly associated tone.

Pre-Activation Negativity (PrAN) in Brain Potentials to Unfolding Words.

We describe an event-related potential (ERP) effect termed the "pre-activation negativity" (PrAN), which is proposed to index the degree of pre-activation of upcoming word-internal morphemes in speech processing. Using lexical competition measures based on word-initial speech fragments (WIFs), as well as statistical analyses of ERP data from three experiments, it is shown that the PrAN is sensitive to lexical competition and that it reflects the degree of predictive certainty: the negativity is larger when there are fewer upcoming lexical competitors.

Time-Driven Effects on Processing Relative Clauses.

The present response time study investigated how a hypothesized time-based working memory constraint of 2-3 s affects the resolution of grammatical and semantic dependencies. Congruent and incongruent object relative (OR) and subject relative sentences were read at different presentation rates so that the distance between dependent words was either shorter or longer than 2-3 s. Incongruent OR sentences showed an effect of presentation rate. Experiment 1 focused on grammatical dependencies. Processing of adjectives with agreement features mismatching those of the preceding dependent word showed rapid agreement resolution at a time-interval below 2 s. Dependency intervals over 3 s reflected a different, more time-consuming process possibly due to extended search in sentence semantic representations as the grammatical form of the first word in the dependency fades away. In experiment 2, focusing on semantic dependencies, incongruent OR sentences displayed a different pattern: a gradual increase in processing time as a function of distance between dependent words. Thus, the 2-3 s long time-window seems to constrain the maintenance of grammatical forms in working memory.

Word tones cueing morphosyntactic structure: Neuroanatomical substrates and activation time-course assessed by EEG and fMRI.

Previous studies distinguish between right hemisphere-dominant processing of prosodic/tonal information and left-hemispheric modulation of grammatical information as well as lexical tones. Swedish word accents offer a prime testing ground to better understand this division. Although similar to lexical tones, word accents are determined by words' morphosyntactic structure, which enables listeners to use the tone at the beginning of a word to predict its grammatical ending. We recorded electrophysiological and hemodynamic brain responses to words where stem tones matched or mismatched inflectional suffixes. Tones produced brain potential effects after 136 ms, correlating with subject variability in average BOLD in left primary auditory cortex, superior temporal gyrus, and inferior frontal gyrus. Invalidly cued suffixes activated the left inferior parietal lobe, arguably reflecting increased processing cost of their meaning. Thus, interaction of word accent tones with grammatical morphology yielded a rapid neural response correlating in subject variability with activations in predominantly left-hemispheric brain areas.

Word-stem tones cue suffixes in the brain.

High and low tones on Swedish word stems are associated with different classes of suffixes. We tested the electrophysiological effects of high and low stem tones as well as tonally cued and uncued suffixes. Two different tasks were used involving either choosing the suffix-dependent meaning of the words, or pressing a button when the word ended. To determine whether effects were in fact due to association of tones with lexical material, delexicalized stimuli were also used. High tones in lexical items produced an increase in the P2 component in both tasks, interpreted as showing passive anticipatory attention allocated to the associated upcoming suffix. This effect was absent for delexicalized forms, where instead an N1 increase was found for high tones, indicating that the high pitch was unexpected in the absence of lexical material, and did not lead to anticipatory attention. A P600 effect was found for uncued high-associated suffixes in the semantic task, which was also where the largest increase was found in reaction times. This suggests that the tonal cues were most important when participants were required to process the meaning of the words.

Time-driven effects on processing grammatical agreement.

"Agreement" is a grammatical relation between words; e.g., the verbal suffix -s reflects agreement with a singular subject (He run-s). Previous studies with time intervals under 2.5 s between disagreeing words have found a left-lateralized negative brain potential, arguably reflecting detection of the morphosyntactic violation. We tested the neurophysiological effects of number agreement between the first and last word in sentences at temporal distances between 1.75 and 3.25 s. Distances were varied by visually presenting sentences word by word at different rates. For distances under 2.5 s, a left-lateralized negativity was observed. At a 3.25-s interval, an anterior, slightly right-lateralized negativity was found. At an intermediate distance of 2.75 s, the difference between disagreement and agreement at left electrodes correlated with participants' working memory span. Results indicate that different brain processes occur when agreement involves agreement domains approaching and exceeding 3 s than when the agreement dependency involves shorter temporal intervals.

Time-driven effects on parsing during reading.

The phonological trace of perceived words starts fading away in short-term memory after a few seconds. Spoken utterances are usually 2-3s long, possibly to allow the listener to parse the words into coherent prosodic phrases while they still have a clear representation. Results from this brain potential study suggest that even during silent reading, words are organized into 2-3s long 'implicit' prosodic phrases. Participants read the same sentences word by word at different presentation rates. Clause-final words occurring at multiples of 2-3s from sentence onset yielded increased positivity, irrespective of presentation rate. The effect was interpreted as a closure positive shift (CPS), reflecting insertion of implicit prosodic phrase boundaries every 2-3s. Additionally, in participants with low working memory span, clauses over 3s long produced a negativity, possibly indicating increased working memory load.

Atypical associations to abstract words in Broca's aphasia.

INTRODUCTION: Left frontal brain lesions are known to give rise to aphasia and impaired word associations. These associations have previously been difficult to analyze. We used a semantic space method to investigate associations to cue words. The degree of abstractness of the generated words and semantic similarity to the cue words were measured. METHOD: Three subjects diagnosed with Broca's aphasia and twelve control subjects associated freely to cue words. Results were evaluated with latent semantic analysis (LSA) applied to the Swedish Parole corpus. RESULTS: The aphasic subjects could be clearly distinguished from controls by a lower degree of abstractness in the words they generated. The aphasic group's associations showed a negative correlation between semantic similarity to cue word and abstractness of cue word. CONCLUSIONS: By developing novel semantic measures, we showed that Broca's aphasic subjects' word production was characterized by a low degree of abstractness and low degree of coherence in associations to abstract cue words. The results support models where meanings of concrete words are represented in neural networks involving perceptual and motor areas, whereas the meaning of abstract words is more dependent on connections to other word forms in the left frontal region. Semantic spaces can be used in future developments of evaluative tools for both diagnosis and research purposes.