Explore the vast range of titles available.

Surprisal theory posits that less-predictable words should take more time to process, with word predictability quantified as surprisal, i.e., negative log probability in context. While evidence supporting the predictions of surprisal theory has been replicated widely, much of it has focused on a very narrow slice of data: native English speakers reading English texts. Indeed, no comprehensive multilingual analysis exists. We address this gap in the current literature by investigating the relationship between surprisal and reading times in eleven different languages, distributed across five language families. Deriving estimates from language models trained on monolingual and multilingual corpora, we test three predictions associated with surprisal theory: (i) whether surprisal is predictive of reading times, (ii) whether expected surprisal, i.e., contextual entropy, is predictive of reading times, and (iii) whether the linking function between surprisal and reading times is linear. We find that all three predictions are borne out crosslinguistically. By focusing on a more diverse set of languages, we argue that these results offer the most robust link to date between information theory and incremental language processing across languages.

Over the past two decades, numerous studies have demonstrated how less-predictable (i.e., higher surprisal) words take more time to read. In general, these studies have implicitly assumed the reading process is purely : Readers observe a new word and allocate time to process it as required. We argue that prior results are also compatible with a reading process that is at least partially : Readers could make predictions about a future word and allocate time to process it based on their expectation. In this work, we operationalize this anticipation as a word’s contextual entropy. We assess the effect of anticipation on reading by comparing how well surprisal and contextual entropy predict reading times on four naturalistic reading datasets: two self-paced and two eye-tracking. Experimentally, across datasets and analyses, we find substantial evidence for effects of contextual entropy over surprisal on a word’s reading time (RT): In fact, entropy is sometimes better than surprisal in predicting a word’s RT. Spillover effects, however, are generally not captured by entropy, but only by surprisal. Further, we hypothesize four cognitive mechanisms through which contextual entropy could impact RTs—three of which we are able to design experiments to analyze. Overall, our results support a view of reading that is not just responsive, but also anticipatory.

Gender stereotypes influence subjective beliefs about the world, and this is reflected in our use of language. But do gender biases in language transparently reflect subjective beliefs? Or is the process of translating thought to language itself biased? During the 2016 United States (N = 24,863) and 2017 United Kingdom (N = 2,609) electoral campaigns, we compared participants’ beliefs about the gender of the next head of government with their use and interpretation of pronouns referring to the next head of government. In the United States, even when the female candidate was expected to win, she pronouns were rarely produced and induced substantial comprehension disruption. In the United Kingdom, where the incumbent female candidate was heavily favored, she pronouns were preferred in production but yielded no comprehension advantage. These and other findings suggest that the language system itself is a source of implicit biases above and beyond previously known biases, such as those measured by the Implicit Association Test.

IntroductionEvidence-based low-intensity psychological interventions such as Problem Management Plus (PM+) have the potential to expand treatment access for depression and anxiety, yet these interventions are not yet effectively implemented in rural, public health systems in resource-limited settings. In 2017, Partners In Health adapted PM+ for delivery by primary care nurses in rural Rwanda and began integrating PM+ into health centres in collaboration with the Rwandan Ministry of Health, using established implementation strategies for mental health integration into primary care (Mentoring and Enhanced Supervision at Health Centers for Mental Health (MESH MH)). A gap in the evidence regarding whether low-intensity psychological interventions can be successfully integrated into real-world primary care settings and improve outcomes for common mental disorders remains. In this study, we will rigorously evaluate the delivery of PM+ by primary care nurses, supported by MESH MH, as it is scaled across one rural district in Rwanda.Methods and analysisWe will conduct a hybrid type 1 effectiveness-implementation study to test the clinical outcomes of routinely delivered PM+ and to describe the implementation of PM+ at health centres. To study the clinical effectiveness of PM+, we will use a pragmatic, randomised multiple baseline design to determine whether participants experience improvement in depression symptoms (measured by the Patient Health Questionnaire-9) and functioning (measured by the WHO-Disability Assessment Scale Brief 2.0) after receiving PM+. We will employ quantitative and qualitative methods to describe and evaluate PM+ implementation outcomes using the Reach, Effectiveness, Adoption, Implementation and Maintenance framework, using routinely collected programme data and semistructured interviews.Ethics and disseminationThis evaluation was approved by the Rwanda National Ethics Committee (Protocol #196/RNEC/2019) and deemed exempt by the Harvard University Institutional Review Board. The results from this evaluation will be useful for health systems planners and policy-makers working to translate the evidence base for low-intensity psychological interventions into practice.

Jeff Elman (1/22/1948–6/28/2018) was a major and much beloved figure in cognitive science, best known for his work on the TRACE model of speech perception, simple recurrent network models of the temporal dynamics of language processing, and his coauthored monograph, Rethinking Innateness. Beyond his individual and collaborative research, he is widely recognized for his lasting contributions to building our scientific community. Here we celebrate his contributions by briefly recounting his life’s work and sharing commentaries and reminiscences from a number of his closest colleagues over the years.

Children’s early speech often bears little resemblance to that of adults, and yet parents and other caregivers are able to interpret that speech and react accordingly. Here we investigate how adult listeners’ inferences reflect sophisticated beliefs about what children are trying to communicate, as well as how children are likely to pronounce words. Using a Bayesian framework for modelling spoken word recognition, we find that computational models can replicate adult interpretations of children’s speech only when they include strong, context-specific prior expectations about the messages that children will want to communicate. This points to a critical role of adult cognitive processes in supporting early communication and reveals how children can actively prompt adults to take actions on their behalf even when they have only a nascent understanding of the adult language. We discuss the wide-ranging implications of the powerful listening capabilities of adults for theories of first language acquisition. The authors use a computational model of word recognition to show that adults’ interpretation of young children’s speech depends heavily on beliefs about what children are likely to say.

What have language models (LMs) learned about grammar? This question remains hotly debated, with major ramifications for linguistic theory. However, since probability and grammaticality are distinct notions in linguistics, it is not obvious what string probabilities can reveal about an LM’s underlying grammatical knowledge. We present a theoretical analysis of the relationship between grammar, meaning, and string probability, based on simple assumptions about the generative process of corpus data. Our framework makes three predictions, which we validate empirically using 280K sentence pairs in English and Chinese: (1) correlation between the probability of strings within minimal pairs, i.e., string pairs with minimal semantic differences; (2) correlation between models’ and humans’ deltas within minimal pairs; and (3) poor separation in probability space between unpaired grammatical and ungrammatical strings. Our analyses give theoretical grounding for using probability to learn about LMs’ structural knowledge, and suggest directions for future work in LM grammatical evaluation.

There is an increasing demand for the evaluation of expenditure and regulatory measures undertaken by the European Union in order to improve accountability and achieve “value for money” objectives. At the most general level, the task of organizing evaluation systems for these programs falls to the European Commission. Historically, the commission has focused on developing policies rather than monitoring or delivering them. With the maturing of certain policy areas, the commission’s role is shifting in the direction of review and evaluation. From a systemic point of view, the management of EU policy presents particularly severe challenges in the area of evaluation. There are multiple actors located at local, national, and supranational levels; divergent administrative cultures and practices; variable quality of information, records, and capabilities; an attenuated system of reporting; and unclear lines of accountability. Joint funding of some programs creates additional problems by entangling program impacts and the audit purposes and management of national and EU institutions, respectively. The commission has still to come to a view on the parameters defining commonality and diversity in evaluation. In order to improve the situation, the commission has taken a number of initiatives, among them the setting up of an expert working group on the evaluation process. This article reports, in general terms, the findings of that group for 1994–95.

Recent work has found that contemporary language models such as transformers can become so good at next-word prediction that the probabilities they calculate become worse for predicting reading time. In this paper, we propose that this can be explained by reading time being sensitive to simple n-gram statistics rather than the more complex statistics learned by state-of-the-art transformer language models. We demonstrate that the neural language models whose predictions are most correlated with n-gram probability are also those that calculate probabilities that are the most correlated with eye-tracking-based metrics of reading time on naturalistic text.