Multi-Level Linguistic Alignment in a Dynamic Collaborative Problem-Solving Task.

Cocreating meaning in collaboration is challenging. Success is often determined by people's abilities to coordinate their language to converge upon shared mental representations. Here we explore one set of low-level linguistic behaviors, linguistic alignment, that both emerges from, and facilitates, outcomes of high-level convergence. Linguistic alignment captures the ways people reuse, that is, "align to," the lexical, syntactic, and semantic forms of others' utterances. Our focus is on the temporal change of multi-level linguistic alignment, as well as how alignment is related to communicative outcomes within a unique collaborative problem-solving paradigm. The primary task, situated within a virtual educational video game, requires creative thinking between three people where the paths for possible solutions are highly variable. We find that over time interactions are marked by decreasing lexical and syntactic alignment, with a trade-off of increasing semantic alignment. However, greater semantic alignment does not translate into better team performance. Overall, these findings provide greater clarity on the role of linguistic coordination within complex and dynamic collaborative problem-solving tasks.

Precision communication: Physicians' linguistic adaptation to patients' health literacy.

Little quantitative research has explored which clinician skills and behaviors facilitate communication. Mutual understanding is especially challenging when patients have limited health literacy (HL). Two strategies hypothesized to improve communication include matching the complexity of language to patients' HL ("universal tailoring"); or always using simple language ("universal precautions"). Through computational linguistic analysis of 237,126 email exchanges between dyads of 1094 physicians and 4331 English-speaking patients, we assessed matching (concordance/discordance) between physicians' linguistic complexity and patients' HL, and classified physicians' communication strategies. Among low HL patients, discordance was associated with poor understanding (P = 0.046). Physicians' "universal tailoring" strategy was associated with better understanding for all patients (P = 0.01), while "universal precautions" was not. There was an interaction between concordance and communication strategy (P = 0.021): The combination of dyadic concordance and "universal tailoring" eliminated HL-related disparities. Physicians' ability to adapt communication to match their patients' HL promotes shared understanding and equity. The 'Precision Medicine' construct should be expanded to include the domain of 'Precision Communication.'

Couples' co-regulation dynamics as a function of perceived partner dyadic coping.

BACKGROUND AND OBJECTIVES: Perceptions of partners' dyadic coping (DC) behaviors are associated with lower stress and higher relationship well-being. Albeit well-documented, these associations have predominately relied on cross-sectional data, overlooking temporal associations during conversations of mutual stress. Based on the systemic transactional model of DC [Bodenmann, G. (2005). Dyadic coping and its significance for marital functioning. In T. Revenson, K. Kayser, & G. Bodenmann (Eds.), Couples coping with stress: Emerging perspectives on dyadic coping (pp. 33-50). American Psychological Association.], we hypothesized that co-regulatory dynamics would be displayed for couples who generally perceive high positive DC, whereas co-dysregulatory dynamics would be displayed for couples who generally perceive high negative DC. DESIGN AND METHODS: Using video-prompted second-by-second recall of stress experience from 42 different-gender romantic couples, this study examined whether couples' co-regulation dynamics were moderated by perceived partner DC behaviors, measured at baseline. RESULTS: On average, partners' stress ratings were coupled; females tended to coregulate males' stress and both partners' stress dampened over time. Perceived negative DC moderated the coregulation of stress, suggesting that females unidirectionally coregulated males' stress when (1) negative DC was low in both partners and (2) when females reported lower negative DC than males. However, coregulation did not occur when (1) negative DC was high in both partners and (2) females reported higher negative DC than males. CONCLUSIONS: Implications for utilizing methods sensitive to temporal interpersonal emotion dynamics are presented.

Challenges and solutions to employing natural language processing and machine learning to measure patients' health literacy and physician writing complexity: The ECLIPPSE study.

OBJECTIVE: In the National Library of Medicine funded ECLIPPSE Project (Employing Computational Linguistics to Improve Patient-Provider Secure Emails exchange), we attempted to create novel, valid, and scalable measures of both patients' health literacy (HL) and physicians' linguistic complexity by employing natural language processing (NLP) techniques and machine learning (ML). We applied these techniques to > 400,000 patients' and physicians' secure messages (SMs) exchanged via an electronic patient portal, developing and validating an automated patient literacy profile (LP) and physician complexity profile (CP). Herein, we describe the challenges faced and the solutions implemented during this innovative endeavor. MATERIALS AND METHODS: To describe challenges and solutions, we used two data sources: study documents and interviews with study investigators. Over the five years of the project, the team tracked their research process using a combination of Google Docs tools and an online team organization, tracking, and management tool (Asana). In year 5, the team convened a number of times to discuss, categorize, and code primary challenges and solutions. RESULTS: We identified 23 challenges and associated approaches that emerged from three overarching process domains: (1) Data Mining related to the SM corpus; (2) Analyses using NLP indices on the SM corpus; and (3) Interdisciplinary Collaboration. With respect to Data Mining, problems included cleaning SMs to enable analyses, removing hidden caregiver proxies (e.g., other family members) and Spanish language SMs, and culling SMs to ensure that only patients' primary care physicians were included. With respect to Analyses, critical decisions needed to be made as to which computational linguistic indices and ML approaches should be selected; how to enable the NLP-based linguistic indices tools to run smoothly and to extract meaningful data from a large corpus of medical text; and how to best assess content and predictive validities of both the LP and the CP. With respect to the Interdisciplinary Collaboration, because the research required engagement between clinicians, health services researchers, biomedical informaticians, linguists, and cognitive scientists, continual effort was needed to identify and reconcile differences in scientific terminologies and resolve confusion; arrive at common understanding of tasks that needed to be completed and priorities therein; reach compromises regarding what represents "meaningful findings" in health services vs. cognitive science research; and address constraints regarding potential transportability of the final LP and CP to different health care settings. DISCUSSION: Our study represents a process evaluation of an innovative research initiative to harness "big linguistic data" to estimate patient HL and physician linguistic complexity. Any of the challenges we identified, if left unaddressed, would have either rendered impossible the effort to generate LPs and CPs, or invalidated analytic results related to the LPs and CPs. Investigators undertaking similar research in HL or using computational linguistic methods to assess patient-clinician exchange will face similar challenges and may find our solutions helpful when designing and executing their health communications research.

ALIGN: Analyzing linguistic interactions with generalizable techNiques-A Python library.

Linguistic alignment (LA) is the tendency during a conversation to reuse each other's linguistic expressions, including lexical, conceptual, or syntactic structures. LA is often argued to be a crucial driver in reciprocal understanding and interpersonal rapport, though its precise dynamics and effects are still controversial. One barrier to more systematic investigation of these effects lies in the diversity in the methods employed to analyze LA, which makes it difficult to integrate and compare results of individual studies. To overcome this issue, we have developed ALIGN (Analyzing Linguistic Interactions with Generalizable techNiques), an open-source Python package to measure LA in conversation (https://pypi.python.org/pypi/align) along with in-depth open-source tutorials hosted on ALIGN's GitHub repository (https://github.com/nickduran/align-linguistic-alignment). Here, we first describe the challenges in the study of LA and outline how ALIGN can address them. We then demonstrate how our analytical protocol can be applied to theory-driven questions using a complex corpus of dialogue (the Devil's Advocate corpus; Duran & Fusaroli, 2017). We close by identifying further challenges and point to future developments of the field. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

Examining the Effects of Couples' Real-Time Stress and Coping Processes on Interaction Quality: Language Use as a Mediator.

Stress in romantic relationships is an all-too-common phenomenon that has detrimental effects on relationship well-being. Specifically, stress can lead to negative interactions between partners and ultimately decrease relationship functioning. The systemic-transactional model of dyadic coping posits that by effectively communicating stress and coping with one's romantic partner, couples can mitigate the deleterious effects of stress. Specifically, partners can engage in positive dyadic coping, which may foster couples' sense of "we-ness," strengthen their emotional connection, and facilitate their understanding of each other's stressful experiences. However, these associations have not yet been examined during partners' real-time stress conversations. When assessing dyadic coping, a particular aspect of interest is partners' language use (i.e., pronouns, emotion words, and cognition words), as it may reflect the types of support they communicate to one another. Using real-time interaction data from 41 heterosexual couples, this study examined how couples' stress and coping processes affect perceived interaction quality following discussions of stress. Specifically, language use (i.e., pronouns, emotion words, and cognition words) was assessed as a mediator on the association between observed stress communication and perceived interaction quality. Overall, results supported our hypotheses; when one partner communicated stress, the other partner responded with language use indicative of different types of dyadic coping (i.e., more you-talk and use of emotion words, less we-talk, I-talk, and use of cognition words), which were in turn associated with interaction quality in mixed directions. Implications of these findings for romantic couples are discussed.

Conversing with a devil's advocate: Interpersonal coordination in deception and disagreement.

This study investigates the presence of dynamical patterns of interpersonal coordination in extended deceptive conversations across multimodal channels of behavior. Using a novel "devil's advocate" paradigm, we experimentally elicited deception and truth across topics in which conversational partners either agreed or disagreed, and where one partner was surreptitiously asked to argue an opinion opposite of what he or she really believed. We focus on interpersonal coordination as an emergent behavioral signal that captures interdependencies between conversational partners, both as the coupling of head movements over the span of milliseconds, measured via a windowed lagged cross correlation (WLCC) technique, and more global temporal dependencies across speech rate, using cross recurrence quantification analysis (CRQA). Moreover, we considered how interpersonal coordination might be shaped by strategic, adaptive conversational goals associated with deception. We found that deceptive conversations displayed more structured speech rate and higher head movement coordination, the latter with a peak in deceptive disagreement conversations. Together the results allow us to posit an adaptive account, whereby interpersonal coordination is not beholden to any single functional explanation, but can strategically adapt to diverse conversational demands.

When expectancies collide: Action dynamics reveal the interaction between stimulus plausibility and congruency.

The cognitive architecture routinely relies on expectancy mechanisms to process the plausibility of stimuli and establish their sequential congruency. In two computer mouse-tracking experiments, we use a cross-modal verification task to uncover the interaction between plausibility and congruency by examining their temporal signatures of activation competition as expressed in a computer- mouse movement decision response. In this task, participants verified the content congruency of sentence and scene pairs that varied in plausibility. The order of presentation (sentence-scene, scene-sentence) was varied between participants to uncover any differential processing. Our results show that implausible but congruent stimuli triggered less accurate and slower responses than implausible and incongruent stimuli, and were associated with more complex angular mouse trajectories independent of the order of presentation. This study provides novel evidence of a disassociation between the temporal signatures of plausibility and congruency detection on decision responses.

Self-serving dishonest decisions can show facilitated cognitive dynamics.

We use a novel task to test two competing hypotheses concerning the cognitive processes involved in dishonesty. Many existing accounts of deception imply that in order to act dishonestly one has to use cognitive control to overcome a bias toward the truth, which results in more time and effort. A recent hypothesis suggests that lying in order to serve self-interest may be a rapid, even automatic tendency taking less time than refraining from lying. In the current study, we track the action dynamics of potentially dishonest decisions to investigate the underlying cognitive processes. Participants are asked to privately predict the outcome of a virtual coin flip, report their accuracy and receive bonus credit for accurate predictions. The movements of the computer cursor toward the target answer are recorded and used to characterize the dynamics of decisions. Our results suggest that when a self-serving condition holds, decisions that have a high probability of being dishonest take less time and experience less hesitation.

A mass assembly of associative mechanisms: a dynamical systems account of natural social interaction.

The target article offers a negative, eliminativist thesis, dissolving the specialness of mirroring processes into a solution of associative mechanisms. We support the authors' project enthusiastically. What they are currently missing, we argue, is a positive, generative thesis about associative learning mechanisms and how they might give way to the complex, multimodal coordination that naturally arises in social interaction.

Exploring the movement dynamics of deception.

BOTH THE SCIENCE AND THE EVERYDAY PRACTICE OF DETECTING A LIE REST ON THE SAME ASSUMPTION: hidden cognitive states that the liar would like to remain hidden nevertheless influence observable behavior. This assumption has good evidence. The insights of professional interrogators, anecdotal evidence, and body language textbooks have all built up a sizeable catalog of non-verbal cues that have been claimed to distinguish deceptive and truthful behavior. Typically, these cues are discrete, individual behaviors-a hand touching a mouth, the rise of a brow-that distinguish lies from truths solely in terms of their frequency or duration. Research to date has failed to establish any of these non-verbal cues as a reliable marker of deception. Here we argue that perhaps this is because simple tallies of behavior can miss out on the rich but subtle organization of behavior as it unfolds over time. Research in cognitive science from a dynamical systems perspective has shown that behavior is structured across multiple timescales, with more or less regularity and structure. Using tools that are sensitive to these dynamics, we analyzed body motion data from an experiment that put participants in a realistic situation of choosing, or not, to lie to an experimenter. Our analyses indicate that when being deceptive, continuous fluctuations of movement in the upper face, and somewhat in the arms, are characterized by dynamical properties of less stability, but greater complexity. For the upper face, these distinctions are present despite no apparent differences in the overall amount of movement between deception and truth. We suggest that these unique dynamical signatures of motion are indicative of both the cognitive demands inherent to deception and the need to respond adaptively in a social context.

Prediction during statistical learning, and implications for the implicit/explicit divide.

Accounts of statistical learning, both implicit and explicit, often invoke predictive processes as central to learning, yet practically all experiments employ non-predictive measures during training. We argue that the common theoretical assumption of anticipation and prediction needs clearer, more direct evidence for it during learning. We offer a novel experimental context to explore prediction, and report results from a simple sequential learning task designed to promote predictive behaviors in participants as they responded to a short sequence of simple stimulus events. Predictive tendencies in participants were measured using their computer mouse, the trajectories of which served as a means of tapping into predictive behavior while participants were exposed to very short and simple sequences of events. A total of 143 participants were randomly assigned to stimulus sequences along a continuum of regularity. Analysis of computer-mouse trajectories revealed that (a) participants almost always anticipate events in some manner, (b) participants exhibit two stable patterns of behavior, either reacting to vs. predicting future events, (c) the extent to which participants predict relates to performance on a recall test, and (d) explicit reports of perceiving patterns in the brief sequence correlates with extent of prediction. We end with a discussion of implicit and explicit statistical learning and of the role prediction may play in both kinds of learning.

Listeners invest in an assumed other's perspective despite cognitive cost.

We explored perspective-taking behavior in a visuospatial mental rotation task that requires listeners to adopt an egocentric or "other-centric" frame of reference. In the current task, objects could be interpreted relative to the point-of-view of the listener (egocentric) or of a simulated partner (other-centric). Across three studies, we evaluated participants' willingness to consider and act on partner-specific information, showing that a partner's perceived ability to contribute to collaborative mutual understanding modulated participants' perspective-taking behavior, either by increasing other-centric (Study 2) or egocentric (Study 3) responding. Moreover, we show that a large proportion of participants resolved referential ambiguity in terms of their partner's perspective, even when it was more cognitively difficult to do so (as tracked by online movement measures), and when the presence of a social partner had to be assumed (Studies 1 and 2). In addition, participants continued to consider their partner's perspective during trials where visual perspectives were shared. Our results show that participants will thoroughly invest in either an other-centric or egocentric mode of responding, and that perspective-taking strategies are not always dictated by minimizing processing demands, but by more potent (albeit subtle) factors in the social context.

The cognitive dynamics of negated sentence verification.

We explored the influence of negation on cognitive dynamics, measured using mouse-movement trajectories, to test the classic notion that negation acts as an operator on linguistic processing. In three experiments, participants verified the truth or falsity of simple statements, and we tracked the computer-mouse trajectories of their responses. Sentences expressing these facts sometimes contained a negation. Such negated statements could be true (e.g., "elephants are not small") or false (e.g., "elephants are not large"). In the first experiment, as predicted by the classic notion of negation, we found that negation caused more discreteness in the mouse trajectory of a response. The second experiment induced a simple context for these statements, yet negation still increased discreteness in trajectories. A third experiment enhanced the pragmatic context of sentences, and the discreteness was substantially diminished, with one primary measure no longer significantly showing increased discreteness at all. Traditional linguistic theories predict rapid shifts in cognitive dynamics occur due to the nature of negation: It is an operator that reverses the truth or falsity of an interpretation. We argue that these results support both propositional and contextual accounts of negation present in the literature, suggesting that contextual factors are crucial for determining the kind of cognitive dynamics displayed. We conclude by drawing broader lessons about theories of cognition from the case of negation.

The bit in the middle and why it's important: a computational analysis of the linguistic features of body paragraphs.

This study examines the relationship between the linguistic characteristics of body paragraphs of student essays and the total number of paragraphs in the essays. Results indicate a significant relationship between the total number of paragraphs and a variety of linguistic characteristics known to affect student essay scores. These linguistic characteristics (e.g., semantic overlap, syntactic complexity) contribute to two underlying factors (i.e., textual cohesion and difficulty) that are used as dependent variables in mixed-effect models. Results suggest that student essays with 5-8 paragraphs tend to be more linguistically consistent than student essays with 3, 4, and 9 paragraphs. Essays with totals of 5-8 paragraphs, considered by many educators to contain an optimal number of paragraphs, may include functionally and structurally similar paragraphs. These findings could aid writing researchers and educators in obtaining a clearer view of the relationship between the total number of paragraphs comprising an essay and the linguistic characteristics that affect essay evaluation. Consequently, writing interventions may become better equipped to pinpoint student difficulties and facilitate student writing skills by providing more detailed and informed feedback.

The action dynamics of overcoming the truth.

A convincing deceiver must act in discordance with their knowledge of the truth. To do so requires the deceiver to resolve competition between what is known to be true and what is intended to be false. We investigated the temporal signature of this competition by examining the action dynamics of arm movement while participants responded falsely or truthfully to autobiographical information. The participants answered no or yes by navigating a Nintendo Wii Remote to no and yes regions on a large projector screen. Trajectory analyses of the fine-grained arm movements show increased complexity in false responding relative to truthful responding, with the greatest difference in false yes answers. The dynamic motor movements also reveal greater strength of competition during the act of false responding, thereby extending traditional response time measures that capture latent competition alone. These results suggest that deceptive processes may be detectable when action is allowed to covary with thought. Supplemental figures and a list of the sentence stimuli may be downloaded from http://pbr.psychonomic-journals.org/content/supplemental.

Identifying topic sentencehood.

Four experiments were conducted to assess two models of topic sentencehood identification: the derived model and the free model. According to the derived model, topic sentences are identified in the context of the paragraph and in terms of how well each sentence in the paragraph captures the paragraph's theme. In contrast, according to the free model, topic sentences can be identified on the basis of sentential features without reference to other sentences in the paragraph (i.e., without context). The results of the experiments suggest that human raters can identify topic sentences both with and without the context of the other sentences in the paragraph. Another goal of this study was to develop computational measures that approximated each of these models. When computational versions were assessed, the results for the free model were promising; however, the derived model results were poor. These results collectively imply that humans' identification of topic sentences in context may rely more heavily on sentential features than on the relationships between sentences in a paragraph.

Using temporal cohesion to predict temporal coherence in narrative and expository texts.

We investigated the linguistic features of temporal cohesion that distinguish variations in temporal coherence. In an analysis of 150 texts, experts rated temporal coherence on three continuous scale measures designed to capture unique representations of time. Coh-Metrix, a computational tool that assesses textual cohesion, correctly predicted the human ratings with five features of temporal cohesion. The correlations between predicted and actual scores were all statistically significant. In a complementary study, we explored the importance of temporal cohesion in characterizing genre. A discriminant function analysis, using Coh-Metrix temporal indices, successfully distinguished the genres of science, history, and narrative texts. The results suggested that history texts are more similar to narrative texts than to science texts in terms of temporal cohesion.