The electrophysiology of lexical prediction of emoji and text.

As emoji often appear naturally alongside text in utterances, they provide a way to study how prediction unfolds in multimodal sentences in direct comparison to unimodal sentences. In this experiment, participants (N = 40) read sentences in which the sentence-final noun appeared in either word form or emoji form, a between-subjects manipulation. The experiment featured both high constraint sentences and low constraint sentences to examine how the lexical processing of emoji interacts with prediction processes in sentence comprehension. Two well-established ERP components linked to lexical processing and prediction - the N400 and the Late Frontal Positivity - are investigated for sentence-final words and emoji to assess whether, to what extent, and in what linguistic contexts emoji are processed like words. Results indicate that the expected effects, namely an N400 effect to an implausible lexical item compared to a plausible one and an LFP effect to an unexpected lexical item compared to an expected one, emerged for both words and emoji. This paper discusses the similarities and differences between the stimulus types and constraint conditions, contextualized within theories of linguistic prediction, ERP components, and a multimodal lexicon.

Improved estimation of the relationship between fetal growth and late stillbirth in the United States, 2014-15.

In the United States the rate of stillbirth after 28 weeks' gestation (late stillbirth) is 2.7/1000 births. Fetuses that are small for gestational age (SGA) or large for gestational age (LGA) are at increased risk of stillbirth. SGA and LGA are often categorized as growth or birthweight ≤ 10th and ≥ 90th centile, respectively; however, these cut-offs are arbitrary. We sought to characterize the relationship between birthweight and stillbirth risk in greater detail. Data on singleton births between 28- and 44-weeks' gestation from 2014 to 2015 were extracted from the US Centers for Disease Control and Prevention live birth and fetal death files. Growth was assessed using customized birthweight centiles (Gestation Related Optimal Weight; GROW). The analyses included logistic regression using SGA/LGA categories and a generalized additive model (GAM) using birthweight centile as a continuous exposure. Although the SGA and LGA categories identified infants at risk of stillbirth, categorical models provided poor fits to the data within the high-risk bins, and in particular markedly underestimated the risk for the extreme centiles. For example, for fetuses in the lowest GROW centile, the observed rate was 39.8/1000 births compared with a predicted rate of 11.7/1000 from the category-based analysis. In contrast, the model-predicted risk from the GAM tracked closely with the observed risk, with the GAM providing an accurate characterization of stillbirth risk across the entire birthweight continuum. This study provides stillbirth risk estimates for each GROW centile, which clinicians can use in conjunction with other clinical details to guide obstetric management.

Risk factors for late (28+ weeks' gestation) stillbirth in the United States, 2014-2015.

BACKGROUND: In the United States (US) late stillbirth (at 28 weeks or more of gestation) occurs in 3/1000 births. AIM: We examined risk factors for late stillbirth with the specific goal of identifying modifiable factors that contribute substantially to stillbirth burden. SETTING: All singleton births in the US for 2014-2015. METHODS: We used a retrospective population-based design to assess the effects of multiple factors on the risk of late stillbirth in the US. Data were drawn from the US Centers for Disease Control and Prevention live birth and fetal death data files. RESULTS: There were 6,732,157 live and 18,334 stillbirths available for analysis (late stillbirth rate = 2.72/1000 births). The importance of sociodemographic determinants was shown by higher risks for Black and Native Hawaiian and Other Pacific Islander mothers compared with White mothers, mothers with low educational attainment, and older mothers. Among modifiable risk factors, delayed/absent prenatal care, diabetes, hypertension, and maternal smoking were associated with increased risk, though they accounted for only 3-6% of stillbirths each. Two factors accounted for the largest proportion of late stillbirths: high maternal body mass index (BMI; 15%) and infants who were small for gestational age (38%). Participation in the supplemental nutrition for women, infants and children program was associated with a 28% reduction in overall stillbirth burden. CONCLUSIONS: This study provides population-based evidence for stillbirth risk in the US. A high proportion of late stillbirths was associated with high maternal BMI and small for gestational age, whereas participation in supplemental nutrition programs was associated with a large reduction in stillbirth burden. Addressing obesity and fetal growth restriction, as well as broadening participation in nutritional supplementation programs could reduce late stillbirths.

Nature dependent tourism - Combining big data and local knowledge.

The ability to quantify nature's value for tourism has significant implications for natural resource management and sustainable development policy. This is especially true in the Eastern Caribbean, where many countries are embracing the concept of the Blue Economy. The utilization of user-generated content (UGC) to understand tourist activities and preferences, including the use of artificial intelligence and machine learning approaches, remains at the early stages of development and application. This work describes a new effort which has modelled and mapped multiple nature dependent sectors of the tourism industry across five small island nations. It makes broad use of UGC, while acknowledging the challenges and strengthening the approach with substantive input, correction, and modification from local experts. Our approach to measuring the nature-dependency of tourism is practical and scalable, producing data, maps and statistics of sufficient detail and veracity to support sustainable resource management, marine spatial planning, and the wider promotion of the Blue Economy framework.

Misretrieval but not misrepresentation: A feature misbinding account of post-interpretive effects in number attraction.

Attraction effects in comprehension have reliably shown a grammaticality asymmetry in which mismatching plural attractors confer facilitatory interference for ungrammatical verbs, but no processing cost for grammatical verbs. While this has favoured cue-based retrieval accounts of attraction phenomena in comprehension, Patson and Husband offered offline evidence suggesting that comprehenders systematically misrepresent number information in attraction phrases, leaving open the possibility for faulty noun phrase (NP) representations later in processing. The current study employs two self-paced reading discourse experiments to test for number attraction misrepresentations in real time. Specifically, the attraction phrases occurred as embedded direct object phrases, allowing for a direct test of the role of attractor noun number in head noun number misrepresentation (i.e., no number cue from verb). Although no online evidence for misrepresentation was found, a third single-sentence rapid serial visual presentation experiment showed error rates to offline probes corroborating the post-interpretive findings from Patson and Husband, suggesting that a search in memory for associative features may not employ the same processes as the formation of dependencies in discourse comprehension. The findings are discussed in the framework of feature misbinding in memory in line with recent post-interpretive accounts of offline comprehension errors.

Semantic constraint, reading control, and the granularity of form-based expectations during semantic processing: Evidence from ERPs.

We investigated the role that semantic constraint and participant control over stimulus presentation have on early stages of visual word recognition. Namely, we tested how the presence of a highly constraining sentential context influences the expectations that readers have during incremental sentence processing. Further, we tested whether allowing participants to self-pace the experiment affected early sensory perceptions of written stimuli. Event-related potentials (ERPs) were recorded in three experiments. Participants read sentences containing a target word from one of four conditions: 1) the target, spelled as expected; 2) the target with two internal characters transposed; 3) a nonword one vowel different from a target; or 4) an illegal consonant string. In Experiment 1, sentences were minimally constraining up to the target word (average cloze at target word: 0.01); in Experiments 2 and 3, sentences were highly constraining (average cloze at target word: 0.93). In both Experiments 1 and 2, sentences were presented using rapid-serial-visual presentation (RSVP). In Experiment 3, participants saw the same sentences used in Experiment 2 but were allowed to self-pace the presentation of each word in every trial. In Experiments 1 and 2, results showed early neural sensitivity to nonsensical consonant strings only, and only when they appeared within high constraint. In Experiment 3, results showed graded N170 effects to all target words containing unexpected visual information. P600 modulations were observed in all three experiments, indexing the difficulty of processing unexpected orthography, particularly in downstream, integrative processing. Results support a nuanced view of early visual processing, namely one arguing that visual processing is more fine-grained the more control participants have over how they read.

Reassessing the electrophysiological evidence for categorical perception of Mandarin lexical tone: ERP evidence from native and naïve non-native Mandarin listeners.

Some studies have argued that native speakers of tonal languages have been shown to perceive lexical tone continua in a more categorical manner than speakers of non-tonal languages. Among these, Zhang and colleagues (NeuroReport 23 (1): 35-9) conducted an event-related potential (ERP) study using an oddball paradigm showing that native Mandarin speakers exhibit different sensitivity to deviant tones that cross category boundaries compared to deviants that belong to the same category as the standard. Other recent ERP findings examining consonant voicing categories question whether perception is truly categorical. The current study investigated these discrepant findings by replicating and extending the Zhang et al. study. Native Mandarin speakers and naïve English speakers performed an auditory oddball detection test while ERPs were recorded. Naïve English speakers were included to test for language experience effects. We found that Mandarin speakers and English speakers demonstrated qualitatively similar responses, in that both groups showed a larger N2 to the across-category deviant and a larger P3 to the within-category deviant. The N2/P3 pattern also did not differ in scalp topography for the within- versus across-category deviants, as was reported by Zhang et al. Cross-language differences surfaced in behavioral results, where Mandarin speakers showed better discrimination for the across-category deviant, but English speakers showed better discrimination for within-category deviants, though all results were near-ceiling. Our results therefore support models suggesting that listeners remain sensitive to gradient acoustic differences in speech even when they have learned phonological categories along an acoustic dimension.

Robust neurocognitive individual differences in grammatical agreement processing: A latent variable approach.

Many neurocognitive accounts of language processing presume that neural responses detected in grand mean analyses of cortical electrophysiological activity reflect the normative brain response in the population under investigation. However, emerging work now shows that individuals' brain responses can vary systematically in both the size and type of effect elicited. The present research therefore examined individual differences in neural activity elicited by grammatical agreement anomalies during language comprehension in a large cohort of highly literate, monolingual English speakers (N = 114), a population generally assumed to be relatively homogenous in terms of linguistic knowledge and processing. Results showed systematic variability in event-related brain potentials (ERPs) elicited by subject-verb agreement anomalies, with brain responses varying on a continuum between N400 and P600 dominant responses. Similar variation was found both when agreement was realized via inflectional morphology or via lexical alternations. Individuals' brain response type correlated strongly across these two conditions. Similar variation was also found for ERPs elicited during rapid serial visual presentation and when self-paced ERPs were recorded. Multilevel latent variable regression showed that variation in brain response amplitude and type was not related to individual differences in language experience or verbal working memory capacity, despite high statistical power. These findings indicate that descriptions of processing dynamics predicated solely on grand mean analyses of central tendency can fail to provide an accurate, generalizable account of how processing unfolds in many or most individual members of the population studied. Furthermore, these findings show that systematic individual variation in engagement of neural system supporting grammatical processing is found even in language users at the highest end of the proficiency spectrum and in grammatically simple sentences. This study therefore has implications for studies of language processing in atypical populations.

A strong wink between verbal and emoji-based irony: How the brain processes ironic emojis during language comprehension.

Emojis are ideograms that are becoming ubiquitous in digital communication. However, no research has yet investigated how humans process semantic and pragmatic content of emojis in real time. We investigated neural responses to irony-producing emojis, the question being whether emoji-generated irony is processed similarly to word-generated irony. Previous ERP studies have routinely found P600 effects to verbal irony. Our research sought to identify whether the same neural responses could also be elicited by emoji-induced irony. In three experiments, participants read sentences that ended in either a congruent, incongruent, or ironic (wink) emoji. Results across all three experiments demonstrated clear P600 effects, the amplitudes of which were correlated with participants' tendency to treat the emoji as a marker of irony, as indicated by behavioral comprehension question responses. These ironic wink emojis also elicited a strong P200 effect, also found in studies of verbal irony processing. Moreover, unexpected emojis (both mismatch and ironic emoji) also elicited late frontal positivities, which have been implicated processing unpredicted words in context. These results are the first to identify how linguistically-relevant ideograms are processed in real-time at the neural level, and specifically draw parallels between the processing of word- and emoji-induced irony.

How right is left? Handedness modulates neural responses during morphosyntactic processing.

Most neurocognitive models of language processing generally assume population-wide homogeneity in the neural mechanisms used during language comprehension, yet individual differences are known to influence these neural mechanisms. In this study, we focus on handedness as an individual difference hypothesized to affect language comprehension. Left-handers and right-handers with a left-handed blood relative, or familial sinistrals, are hypothesized to process language differently than right-handers with no left-handed relatives (Hancock and Bever, 2013; Ullman, 2004). Yet, left-handers are often excluded from neurocognitive language research, and familial sinistrality in right-handers is often not taken into account. In the current study we used event-related potentials to test morphosyntactic processing in three groups that differed in their handedness profiles: left-handers (LH), right-handers with a left-handed blood relative (RH FS+), and right-handers with no reported left-handed blood relative (RH FS-; both right-handed groups were previously tested by Tanner and Van Hell, 2014). Results indicated that the RH FS- group showed only P600 responses during morphosyntactic processing whereas the LH and RH FS+ groups showed biphasic N400-P600 patterns. N400s in LH and RH FS+ groups are consistent with theories that associate left-handedness (self or familial) with increased reliance on lexical/semantic mechanisms during language processing. Inspection of individual-level results illustrated that variability in RH FS- individuals' morphosyntactic processing was remarkably low: most individuals were P600-dominant. In contrast, LH and RH FS+ individuals showed marked variability in brain responses, which was similar for both groups: half of individuals were N400-dominant and half were P600-dominant. Our findings have implications for neurocognitive models of language that have been largely formulated around data from only right-handers without accounting for familial sinistrality or including left-handers, and moreover highlight that there is systematic - and often ignored - variability in language processing outcomes in neurologically healthy populations.

Dissociating retrieval interference and reanalysis in the P600 during sentence comprehension.

We investigated the relative independence of two key processes in language comprehension, as reflected in the P600 ERP component. Numerous studies have linked the P600 to sentence- or message-level reanalysis; however, much research has shown that skilled, cue-based memory retrieval operations are also important to successful language processing. Our goal was to identify whether these cue-based retrieval operations are part of the reanalysis processes indexed by the P600. To this end, participants read sentences that were either grammatical or ungrammatical via subject-verb agreement violations, and in which there was either no possibility for retrieval interference or there was an attractor noun interfering with the computation of subject-verb agreement (e.g., "The slogan on the political poster(s) was/were …"). A stimulus onset asynchrony manipulation (fast, medium, or slow presentation rate) was designed to modulate participants' ability to engage in reanalysis processes. Results showed a reliable attraction interference effect, indexed by reduced behavioral sensitivity to ungrammaticalities and P600 amplitudes when there was an opportunity for retrieval interference, as well as an effect of presentation rate, with reduced behavioral sensitivity and smaller P600 effects at faster presentation rates. Importantly, there was no interaction between the two, suggesting that retrieval interference and sentence-level reanalysis processes indexed by the P600 can be neurocognitively distinct processes.

"Going to town": Large-scale norming and statistical analysis of 870 American English idioms.

An idiom is classically defined as a formulaic sequence whose meaning is comprised of more than the sum of its parts. For this reason, idioms pose a unique problem for models of sentence processing, as researchers must take into account how idioms vary and along what dimensions, as these factors can modulate the ease with which an idiomatic interpretation can be activated. In order to help ensure external validity and comparability across studies, idiom research benefits from the availability of publicly available resources reporting ratings from a large number of native speakers. Resources such as the one outlined in the current paper facilitate opportunities for consensus across studies on idiom processing and help to further our goals as a research community. To this end, descriptive norms were obtained for 870 American English idioms from 2,100 participants along five dimensions: familiarity, meaningfulness, literal plausibility, global decomposability, and predictability. Idiom familiarity and meaningfulness strongly correlated with one another, whereas familiarity and meaningfulness were positively correlated with both global decomposability and predictability. Correlations with previous norming studies are also discussed.

Effects of Grammaticality and Morphological Complexity on the P600 Event-Related Potential Component.

We investigated interactions between morphological complexity and grammaticality on electrophysiological markers of grammatical processing during reading. Our goal was to determine whether morphological complexity and stimulus grammaticality have independent or additive effects on the P600 event-related potential component. Participants read sentences that were either well-formed or grammatically ill-formed, in which the critical word was either morphologically simple or complex. Results revealed no effects of complexity for well-formed stimuli, but the P600 amplitude was significantly larger for morphologically complex ungrammatical stimuli than for morphologically simple ungrammatical stimuli. These findings suggest that some previous work may have inadequately characterized factors related to reanalysis during morphosyntactic processing. Our results show that morphological complexity by itself does not elicit P600 effects. However, in ungrammatical circumstances, overt morphology provides a more robust and reliable cue to morphosyntactic relationships than null affixation.

Cues, quantification, and agreement in language comprehension.

We investigated factors that affect the comprehension of subject-verb agreement in English, using quantification as a window into the relationship between morphosyntactic processes in language production and comprehension. Event-related brain potentials (ERPs) were recorded while participants read sentences with grammatical and ungrammatical verbs, in which the plurality of the subject noun phrase was either doubly marked (via overt plural quantification and morphological marking on the noun) or singly marked (via only plural morphology on the noun). Both acceptability judgments and the ERP data showed heightened sensitivity to agreement violations when quantification provided an additional cue to the grammatical number of the subject noun phrase, over and above plural morphology. This is consistent with models of grammatical comprehension that emphasize feature prediction in tandem with cue-based memory retrieval. Our results additionally contrast with those of prior studies that showed no effects of plural quantification on agreement in language production. These findings therefore highlight some nontrivial divergences in the cues and mechanisms supporting morphosyntactic processing in language production and comprehension.

How inappropriate high-pass filters can produce artifactual effects and incorrect conclusions in ERP studies of language and cognition.

Although it is widely known that high-pass filters can reduce the amplitude of slow ERP components, these filters can also introduce artifactual peaks that lead to incorrect conclusions. To demonstrate this and provide evidence about optimal filter settings, we recorded ERPs in a typical language processing paradigm involving syntactic and semantic violations. Unfiltered results showed standard N400 and P600 effects in the semantic and syntactic violation conditions, respectively. However, high-pass filters with cutoffs at 0.3 Hz and above produced artifactual effects of opposite polarity before the true effect. That is, excessive high-pass filtering introduced a significant N400 effect preceding the P600 in the syntactic condition, and a significant P2 effect preceding the N400 in the semantic condition. Thus, inappropriate use of high-pass filters can lead to false conclusions about which components are influenced by a given manipulation. The present results also lead to practical recommendations for high-pass filter settings that maximize statistical power while minimizing filtering artifacts.

The time-course of feature interference in agreement comprehension: Multiple mechanisms and asymmetrical attraction.

Attraction interference in language comprehension and production may be as a result of common or different processes. In the present paper, we investigate attraction interference during language comprehension, focusing on the contexts in which interference arises and the time-course of these effects. Using evidence from event-related brain potentials (ERPs) and sentence judgment times, we show that agreement attraction in comprehension is best explained as morphosyntactic interference during memory retrieval. This stands in contrast to attraction as a message-level process involving the representation of the subject NP's number features, which is a strong contributor to attraction in production. We thus argue that the cognitive antecedents of agreement attraction in comprehension are non-identical with those of attraction in production, and moreover, that attraction in comprehension is primarily a consequence of similarity-based interference in cue-based memory retrieval processes. We suggest that mechanisms responsible for attraction during language comprehension are a subset of those involved in language production.

ERPs reveal individual differences in morphosyntactic processing.

We investigated individual differences in the neural substrates of morphosyntactic processing among monolingual English speakers using event-related potentials (ERPs). Although grand-mean analysis showed a biphasic LAN-P600 pattern to grammatical violations, analysis of individuals׳ ERP responses showed that brain responses varied systematically along a continuum between negativity- and positivity-dominant ERP responses across individuals. Moreover, the left hemisphere topography of the negativity resulted from component overlap between a centro-parietal N400 in some individuals and a right hemisphere-dominant P600 in others. Our results show that biphasic ERP waveforms do not always reflect separable processing stages within individuals, and moreover, that the LAN can be a variant of the N400. These results show that there are multiple neurocognitive routes to successful grammatical comprehension in language users across the proficiency spectrum. Our results underscore that understanding and quantifying individual differences can provide an important source of evidence about language processing in the general population.