xv, 103 p. : ill. A print copy of this thesis is available through the UO Libraries. Search the library catalog for the location and call number. / While research has extensively documented the inter- and intra-personal consequences of perspective taking, less is known about the mechanisms that underlie this process. Recent research has explored self-other overlap as a mediator of perspective taking on various pro-social outcomes, such as helping and decreased stereotyping. Results have been mixed, perhaps due to the use of different methodologies and scales that actually measure different facets of self-other overlap. This dissertation investigates the structure of self-other overlap and examines how perspective taking may affect only certain facets of self-other overlap, as well as the direction in which this overlap occurs. To test the conceptual equivalence of different overlap measures, in Study 1, participants completed several previously used measures of overlap for two targets: their best friend and an acquaintance. Factor analyses revealed two distinct factors of self-other overlap-- perceived closeness and attribute overlap --although small variations emerged depending on target. These two factors had unique associations with several relationship quality and individual difference measures. Study 2 extended these results by manipulating perspective taking with a stranger. Results replicated the same factor structure from Study 1, and found that perspective taking had different effects on the two factors. Study 3 examined whether or not perspective taking affected the direction of self-other overlap by changing one's attitudes and beliefs to become more like the other person. Results supported a model in which perceived closeness predicted belief change toward the target person, even after accounting for other related consequences of perspective taking such as empathy and positive attitudes. Together, these results suggest that self-other overlap is a multi-dimensional construct associated with different psychological responses. These results are discussed in connection with the relationship between self-other overlap and perspective taking and how this may lead to "self-expansion." / Committee in charge: Sara Hodges, Chairperson, Psychology; Sanjay Srivastava, Member, Psychology; Marjorie Taylor, Member, Psychology; John Lysaker, Outside Member, Philosophy

Parkinson’s disease (PD) is a neurodegenerative condition which results in severe motor impairment. Deterioration in multiple domains of cognition is another hallmark of PD. Together, these motor and cognitive impairments impact substantially on language and communication. Co-speech gestures are a form of action and are also part of linguistic processes, yet have rarely been explored in PD. Gestures can provide imagistic depictions of concepts described by speech and contribute to communication in healthy individuals. They rely on visual, spatial, and motor simulations and imagery, which may be impaired in PD. It is therefore of clinical importance to evaluate how co-speech gestures might be impaired to understand the extent of communicative impairment in PD. PD can also provide a useful model to understanding the cognitive basis of co-speech gesture in healthy people. In Chapter 2, participants described isolated actions. Gesture rate did not differ between the two groups, however, the groups differed in terms of the visual perspective they adopted when depicting actions in gesture. Controls preferred a “character viewpoint” or first-person perspective where their hands represented the hands of the actor, whereas PD patients preferred an “observer viewpoint” or third-person perspective, where their hand represented a whole person. This finding was replicated and extended in Chapter 3 where low-motion and high-motion actions were described in a longer narrative task. PD patients produced fewer character viewpoint gestures when describing high-motion action events, suggesting a difficulty in simulating these events from a first-person perspective. In addition, PD patients had difficult depicting “manner” (how an action is performed) features in gesture during high but not low motion. Extending the findings of Chapter 2, whilst overall rate of gesture production was not affected, PD patients produced action gestures at a significantly lower rate than controls. Chapter 4 took a different focus by investigating gesture depictions of static spatial (rather than dynamic action) features via a house description task. Gesture rate did not differ, but the groups depicted different types of spatial properties to a different extent. Whilst both groups predominantly gestured about location and relative position information, PD patients gestured more about directions whereas controls gestures more about shape and size information. This suggests that different strategies were being employed by the two groups. Finally, testing young adults’ comprehension of these spatial gestures in Chapter 5 revealed that gestures did not significantly improve comprehension of either PD patients’ or controls’ spoken messages, though there may have been ceiling effects. However, both PD patients and controls were viewed as more competent when their messages were viewed with gestures. The findings suggest a selective action-gesture deficit in PD which complements work demonstrating action-verb impairments in these patients, and supports gesture production theories which hypothesise a role for motor simulations and imagery. Overall gesture rate appears to be largely unaffected. The effects of PD can be felt beyond changes to goal-directed action, in the realms of language and social behaviour, but gestures may be able to improve listeners’ social perceptions of PD patients.

Perspective taking is a pivotal behavioral repertoire essential for social functioning and is recognized as a hallmark deficit of the Autism Spectrum Disorder. Recent advancements in the Relational Frame Theory have led to the development of a perspective-taking training protocol shown to improve performance on Theory of Mind tasks in typically developing children; however, there has been little research on the generality of these findings in children with ASD. The impact of deictic responding on social interaction is also undetermined. The current study aimed to: a), evaluate the effectiveness of multiple exemplar training of deictic relations on perspective taking abilities in children with ASD, b), assess the impact of deictic relations on naturalistic Theory of Mind tasks, and c), assess generality of the deictic repertoire on pretend play. All three participants acquired deictic repertoires through double reversed complexity. Acquisition of the relational operants was variable and required many sessions for each participant. Two of three participants showed transfer of perspective taking to the Strange Stories test, all three participants showed overall improvement in performance on various Theory of Mind tasks; and lastly, participants showed mixed results on transfer to pretend play. Implications for using multiple exemplar training in supporting social understanding, prerequisite skills for deictic relational responding, and training strategies are discussed.

Abstract The study investigated how individuals evaluate others' knowledge when they work as a distributed team. Theoretically, the study was based on contemporary theory on collaborative learning and combined with the theories which explain how individuals evaluate others' perspectives in social learning situations. Interpersonal evaluation of knowledge was conceptualized as one mediating process which is needed between social and individual knowledge planes. This study builds on a design-based research approach. Three research experiments were conducted. In the study, a pedagogical model and a visualization tool were developed based on the findings gathered from the first and the second empirical research experiments. It was also investigated how the working model and the visualization tool supported interpersonal evaluation of knowledge. In practice, the model and the tool were tested in the experiments with distributed teams. The results of the experiments are reported in four research articles (Articles I–IV). Based on the analysis of the three research experiments, it can be concluded that when the distributed team members evaluate the other team members' thinking, they use several cognitive reasoning strategies. The findings indicate that the evaluation strategies, such as perspective-taking, comparing, attribution and categorization fulfill each other when the team members try to take the perspectives and shared knowledge of others into account. The results showed also that with the working model or the visualization tool it was possible to support only some strategies of interpersonal evaluation of knowledge at one time. The findings highlight the fact that interpersonal evaluation of knowledge is a multidimensional process. The dimensions which affect the evaluation of others' knowledge are externalized knowledge presented in communication, and an individual's knowledge about, for instance, others' expertise, which may not be externalized in communication. In future studies different levels of analysis are needed to understand how interpersonal evaluation of knowledge is related to the interactions between team members and with technological resources in practice. / Tiivistelmä Tässä tutkimuksessa tarkastellaan hajautetun tiimin jäsenten pyrkimyksiä arvioida toistensa tietoa. Yhä useammin sekä suomalaisten että globaalien organisaatioiden toiminta perustuu ongelmien ratkaisemiseen ja uuden tiedon kehittämiseen tiimeissä. Yleensä monimutkaisten ongelmien ratkominen edellyttää tiimiläisten tapaamisia ja keskusteluja, mutta se ei ole aina mahdollista esimerkiksi pitkien välimatkojen vuoksi. Tällöin tiimiläiset kommunikoivat pääosin teknologian välityksellä, mikä osaltaan tekee tiimiläisten keskinäisestä ymmärtämisestä ja siten myös ongelmien ratkaisusta haastavaa. Teoreettisesti tutkimus nojautuu kollaboratiivisen yhteisöllisen oppimisen teoriaan ja sosiokognitiiviseen oppimisnäkemykseen, joissa toisten tietojen arviointi ja pyrkimys vastavuoroiseen ymmärtämiseen nähdään oppimiselle tärkeinä prosesseina. Tutkimuksessa sovelletaan myös sosiaalispsykologian tutkimuksissa käytettyjä teorioita attribuutiosta ja kategorioinnista. Tutkimus koostuu kolmesta osatutkimuksesta, ja se seuraa Design Based Research -tutkimusotetta. Ensimmäisen ja toisen osatutkimuksien tulosten perusteella kehitettiin yhteisölliseen hajautettuun tiimityöhön työskentelymalli ja visualisointityökalu, jota hajautetun tiimin jäsenet käyttivät tietorepresentaatioiden ja jaetun tiedon arvioimisen tukena kolmannessa osatutkimuksessa. Yleisesti visualisointityökalun ja työskentelymallin avulla pyrittiin tukemaan hajautettujen tiimien jäsenten vastavuoroista ymmärrystä. Toisten tietojen arvioimiseen käytettyjä strategioita tarkasteltiin useiden laadullisten tutkimusmenetelmien avulla. Tutkimuksen tulokset osoittavat, että ymmärtääkseen toistensa näkökulmia hajautettujen tiimien jäsenet käyttävät kognitiivisina strategioina perspektiivinottoa, vertailua, attribuutiointia ja kategorisointia. Perspektiivinotossa pyrkimyksenä on ymmärtää toisen tiimin jäsenen näkökulma jaetun tehtävän sisällön kannalta. Vertailu perustuu omien ja toisten tiimiläisten tietorakenteiden erojen ja yhteneväisyyksien etsintään. Lisäksi tuloksista nousee esille erityisesti tarve ymmärtää, mikä tieto on jo jaettua tiimin jäsenten kesken. Tutkimuksen tulokset osoittavat, että yhteisöllisen ongelmanratkaisun tutkiminen hajautetussa tiimityössä vaatii yksilön kognitiivisten toimintojen analysointia osana sosiaalista tilannetta.

Fourteen young adults participated in a cross-modal picture priming study. Perspective shift processing, in four types of relative clause sentences and in control sentences, was assessed using reaction times. Predictions were: 1) the easier the perspective shifts, the faster the reaction times and 2) subject relative clauses would reveal a priming effect versus attenuated or no priming in object relative clauses due to difficulty following perspective. A priming effect was observed for 1- switch relative clause sentences and for control sentences, while no priming effect was observed for 0 switch, 1+ switch, or 2 switch sentences. Results suggest that variations in local syntactic constructions and word order facilitated relative clause processing. Violations of semantic expectations and noun-noun-verb distance in following perspective can both contribute to the complexity of relative clause processing.

This dissertation represents the first systematic study of the sign language of deaf children with autism. The signing of such children is of particular interest because of the unique ways that some of the known impairments of autism are likely to interact with sign language. In particular, the visual-spatial modality of sign requires signers to understand the visual perspectives of others, a skill which may require theory of mind, which is thought to be delayed in autism (Baron-Cohen et al., 1985). It is hypothesized that an impairment in visual perspective-taking could lead to phonological errors in American Sign Language (ASL), specifically in the parameters of palm orientation, movement, and location. Twenty-five deaf children and adolescents with autism (10 deaf-of-deaf and 15 deaf-of-hearing) between the ages of 4;7 and 20;3 as well as a control group of 13 typically-developing deaf-of-deaf children between the ages of 2;7 and 6;9 were observed in a series of studies, including naturalistic observation, lexical elicitation, fingerspelling, imitation of nonsense gestures, two visual perspective-taking tasks, and a novel sign learning task. The imitation task was also performed on a control group of 24 hearing, non-signing college students. Finally, four deaf mothers of deaf autistic children were interviewed about their children’s signing. Results showed that young deaf-of-deaf autistic children under the age of 10 are prone to making phonological errors involving the palm orientation parameter, substituting an inward palm for an outward palm and vice versa. There is very little evidence that such errors occur in the typical acquisition of ASL or any other sign language. These results indicate that deaf children with autism are impaired from an early age in a cognitive mechanism involved in the acquisition of sign language phonology, though it remains unclear which mechanism(s) might be responsible. This research demonstrates the importance of sign language research for a more complete understanding of autism, as well as the need for research into atypical populations for a better understanding of sign language linguistics. / text

本篇論文的目的在於探討兒童在親子對話中行使重新請求 (request reformulation) 的情況，研究問題如下: 1.兒童採取哪些重新請求的策略以達到請求的目的? 2.親子對話中，常見的重新請求的序列模式（patterns of request reformulation sequences）為何？ 2.如何從兒童重新請求的使用反映出角色取代能力 (perspective-taking ability )的發展? 研究對象為一位三歲和一位六歲的男孩。研究結果顯示隨著年紀的增長，兒童能使用更多元的重新請求策略。研究也發現造成請求失敗的原因隨著不同年齡的兒童有所差異。隨著年紀的增長，兒童面臨的失敗原因和挑戰日漸複雜、困難，兒童會依據不同的失敗原因採取重新請求的策略。最後，研究顯示不同年齡兒童採取的重新請求策略可以展現出他們不同階段的角色取代能力的發展。三歲的兒童無法跳脫自己的觀點而從別人的角度來看待自己的請求，所以重新請求著重於強調自己的需求。六歲的兒童較能夠從別人的立場看待自己的請求，所以他較有能力在考量對方的觀點和利益之後採用對雙方有益的策略。 / The purpose of this study is to explore how children at different age make reformulation to compensate for an unsuccessful request. Firstly, we aim to investigate what reformulation strategies children apply and the patterns of reformulation sequences. Second, we further aim to explore how children’s use of reformulation strategies reveals the development of perspective-taking ability. The data analyzed are natural conversations of two Mandarin-speaking mother-child dyads. Subjects in this present study are two male children. One subject is three years and six months old and the other is six years old. The strategies of request reformulation adopted in this study are mainly based on Levin and Rubin’s (1984) categorization. The results show that children would have a greater variety of reformulation strategies as they get older. Furthermore, aggravation and explanation are both children’s main strategies of reformulation. With the growth of age, children decrease the use of aggravation and increase the use of bargain and mitigation. Furthermore, the results of reformulation sequences reveal that the two children are confronted with different causes of the failure to obtain compliance, which influences their adoption of reformulation strategies. The younger child faced the communicative breakdown and his mother’s ignorance while the older child encountered his mother’s queries and disagreements. Finally, the two children’s application of reformulation strategies revealed their different ability to take the other’s perspectives. The younger child’s reliance on aggravation and speaker-oriented negotiation reveals that he is embedded in his own viewpoints and is less able to view his request from the hearer’s viewpoints while the older child is more able to view the request from the hearer’ perspective and take her benefits into account. Our findings throw some light on children’s use of request reformulation strategy and the development of the perspective-taking ability.

The purpose of this study was to replicate and extend recent findings in the functional contextual literature by 1) establishing complex deictic relational responding skills in 3 persons diagnosed with Schizophrenia and mild-moderate Mental Retardation and 2) assessing generalization through pre and post-instructional measures of Social Anhedonia and Theory of Mind functioning. Results suggest that increasingly complex levels of deictic relational responses were acquired and mastered by all 3 participants and that generalization extended to the Deceptive Container Task (ToM levels 4 & 5) and Hinting Task. Support is provided for the notion that perspective taking skills might be shaped through operant conditioning of deictic frames and that acquisition of these skills may generalize to novel stimuli and settings.

This paper researches the theoretical background needed for the implementation of literary texts with multicultural themes for use in EFL courses in Sweden and it offers several concrete didactical solutions for multicultural reading. The theory of multicultural education by J. A. Banks is presented with focus on the dimension of knowledge construction. The processes that are examined are the learning processes in the zone of proximal development by Vygotsky (1986), the concept of scaffolding by Woods, Bruner & Ross (1976) and the process of perspective-taking by Thein & Sloan (2013). These processes each employ a three-step sequence that moves students from their existing knowledge to new knowledge and revised personal opinions. The teacher’s role is to provide support during the learning process. The second part of the paper suggests different activities for the multicultural reading of a novel, used to make the process of multicultural knowledge construction and scaffolding visible. Teacher support includes text reduction, book discussion and language analysis. Discussion points found in Love Medicine start with revising the stereotypical images, discovering how personal experience influences knowledge or how ethnicity influences professional career choices. When teachers and students read literary texts with multicultural themes, students’ racial prejudice can be reduced (Banks 2004) and their ethical attitudes become more open (Thein & Sloan 2013). The literary work chosen for framing in the theory is Love Medicine by Louise Erdrich. The paper ends with a reflection over the limitations of multicultural reading projects.

Un système de dialogue permet de doter la Machine de la capacité d'interagir de façon naturelle et efficace avec l'Homme. Dans cette thèse nous nous intéressons au développement d'un système de dialogue reposant sur des approches statistiques, et en particulier du cadre formel des Processus Décisionnel de Markov Partiellement Observable, en anglais Partially Observable Markov Decision Process (POMDP), qui à ce jour fait office de référence dans la littérature en ce qui concerne la gestion statistique du dialogue. Ce modèle permet à la fois une prise en compte améliorée de l'incertitude inhérente au traitement des données en provenance de l'utilisateur (notamment la parole) et aussi l'optimisation automatique de la politique d'interaction à partir de données grâce à l'apprentissage par renforcement, en anglais Reinforcement Learning (RL). Cependant, une des problématiques liées aux approches statistiques est qu'elles nécessitent le recours à une grande quantité de données d'apprentissage pour atteindre des niveaux de performances acceptables. Or, la collecte de telles données est un processus long et coûteux qui nécessite généralement, pour le cas du dialogue, la réalisation de prototypes fonctionnels avec l'intervention d'experts et/ou le développement de solution alternative comme le recours à la simulation d'utilisateurs. En effet, très peu de travaux considèrent à ce jour la possibilité d'un apprentissage de la stratégie de la Machine de part sa mise en situation de zéro (sans apprentissage préalable) face à de vrais utilisateurs. Pourtant cette solution présente un grand intérêt, elle permet par exemple d'inscrire le processus d'apprentissage comme une partie intégrante du cycle de vie d'un système lui offrant la capacité de s'adapter à de nouvelles conditions de façon dynamique et continue. Dans cette thèse, nous nous attacherons donc à apporter des solutions visant à rendre possible ce démarrage à froid du système mais aussi, à améliorer sa capacité à s'adapter à de nouvelles conditions (extension de domaine, changement d'utilisateur,...). Pour ce faire, nous envisagerons dans un premier temps l'utilisation de l'expertise du domaine (règles expertes) pour guider l'apprentissage initial de la politique d'interaction du système. De même, nous étudierons l'impact de la prise en compte de jugements subjectifs émis par l'utilisateur au fil de l'interaction dans l'apprentissage, notamment dans un contexte de changement de profil d'utilisateur où la politique préalablement apprise doit alors pouvoir s'adapter à de nouvelles conditions. Les résultats obtenus sur une tâche de référence montrent la possibilité d'apprendre une politique (quasi-)optimale en quelques centaines d'interactions, mais aussi que les informations supplémentaires considérées dans nos propositions sont à même d'accélérer significativement l'apprentissage et d'améliorer la tolérance aux bruits dans la chaîne de traitement. Dans un second temps nous nous intéresserons à réduire les coûts de développement d'un module de compréhension de la parole utilisé dans l'étiquetage sémantique d'un tour de dialogue. Pour cela, nous exploiterons les récentes avancées dans les techniques de projection des mots dans des espaces vectoriels continus conservant les propriétés syntactiques et sémantiques, pour généraliser à partir des connaissances initiales limitées de la tâche pour comprendre l'utilisateur. Nous nous attacherons aussi à proposer des solutions afin d'enrichir dynamiquement cette connaissance et étudier le rapport de cette technique avec les méthodes statistiques état de l'art. Là encore nos résultats expérimentaux montrent qu'il est possible d'atteindre des performances état de l'art avec très peu de données et de raffiner ces modèles ensuite avec des retours utilisateurs dont le coût peut lui-même être optimisé. / A dialogue system should give the machine the ability to interactnaturally and efficiently with humans. In this thesis, we focus on theissue of the development of stochastic dialogue systems. Thus, we especiallyconsider the Partially Observable Markov Decision Process (POMDP)framework which yields state-of-the-art performance on goal-oriented dialoguemanagement tasks. This model enables the system to cope with thecommunication ambiguities due to noisy channel and also to optimize itsdialogue management strategy directly from data with Reinforcement Learning (RL)methods.Considering statistical approaches often requires the availability of alarge amount of training data to reach good performance. However, corpora of interest are seldom readily available and collectingsuch data is both time consuming and expensive. For instance, it mayrequire a working prototype to initiate preliminary experiments with thesupport of expert users or to consider other alternatives such as usersimulation techniques.Very few studies to date have considered learning a dialogue strategyfrom scratch by interacting with real users, yet this solution is ofgreat interest. Indeed, considering the learning process as part of thelife cycle of a system offers a principle framework to dynamically adaptthe system to new conditions in an online and seamless fashion.In this thesis, we endeavour to provide solutions to make possible thisdialogue system cold start (nearly from scratch) but also to improve its ability to adapt to new conditions in operation (domain extension, new user profile, etc.).First, we investigate the conditions under which initial expertknowledge (such as expert rules) can be used to accelerate the policyoptimization of a learning agent. Similarly, we study how polarized userappraisals gathered throughout the course of the interaction can beintegrated into a reinforcement learning-based dialogue manager. Morespecifically, we discuss how this information can be cast intosocially-inspired rewards to speed up the policy optimisation for bothefficient task completion and user adaptation in an online learning setting.The results obtained on a reference task demonstrate that a(quasi-)optimal policy can be learnt in just a few hundred dialogues,but also that the considered additional information is able tosignificantly accelerate the learning as well as improving the noise tolerance.Second, we focus on reducing the development cost of the spoken language understanding module. For this, we exploit recent word embedding models(projection of words in a continuous vector space representing syntacticand semantic properties) to generalize from a limited initial knowledgeabout the dialogue task to enable the machine to instantly understandthe user utterances. We also propose to dynamically enrich thisknowledge with both active learning techniques and state-of-the-artstatistical methods. Our experimental results show that state-of-the-artperformance can be obtained with a very limited amount of in-domain andin-context data. We also show that we are able to refine the proposedmodel by exploiting user returns about the system outputs as well as tooptimize our adaptive learning with an adversarial bandit algorithm tosuccessfully balance the trade-off between user effort and moduleperformance.Finally, we study how the physical embodiment of a dialogue system in a humanoid robot can help the interaction in a dedicated Human-Robotapplication where dialogue system learning and testing are carried outwith real users. Indeed, in this thesis we propose an extension of thepreviously considered decision-making techniques to be able to take intoaccount the robot's awareness of the users' belief (perspective taking)in a RL-based situated dialogue management optimisation procedure.