Context Of Dialogue Research Articles

Knowledge-Grounded Response Generation with Deep Attentional Latent-Variable Model

End-to-end dialogue generation has achieved promising results without using handcrafted features and attributes specific to each task and corpus. However, one of the fatal drawbacks in such approaches is that they are unable to generate informative utterances, so it limits their usage from some real-world conversational applications. In order to tackle this issue, this paper attempts to generate diverse and informative responses with a variational generation model, which contains a joint attention mechanism conditioning on the information from both dialogue contexts and extra knowledge. The experiments on benchmark DSTC7 data show that the proposed method generates responses with more grounded knowledge and improve the diversity of generated language.

Learning Multi-Level Information for Dialogue Response Selection by Highway Recurrent Transformer

With increasing research interests in dialogue modeling, there is an emerging branch that formulates this task as next sentence selection, where given the partial dialogue context, the goal is to determine the most probable next sentence. To model natural language information, recurrent models have been applied to sequence modeling and shown promising results in various NLP tasks (Sutskever et al., 2014). Recently, the Transformer (Vaswani et al., 2017) has advanced modeling semantics for natural language sentences via attention, achieving improvement for sequence modeling. However, the Transformer focuses on modeling the intra-sentence attention but ignores inter-sentence information. In terms of dialogue modeling, the cross-sentence information is salient to understand dialogue content, so that the response selection can be better determined. Therefore, this paper proposes a novel attention mechanism based on multi-head attention, called highway attention, in order to allow the model to pass information through multiple sentences, and then builds a recurrent model based on the Transformer and the proposed highway attention. We call this model Highway Recurrent Transformer. This model focuses on not only intra-sentence dependency, but also inter-sentence dependency in the structure of dialogues. Experiments on the response selection task of the seventh Dialog System Technology Challenge (DSTC7) demonstrate that the proposed Highway Recurrent Transformer is capable of modeling both utterance-level and dialogue-level information for achieving better performance than the original Transformer in the single positive response scenario.

Context and knowledge aware conversational model and system combination for grounded response generation

End-to-end neural-based dialogue systems can potentially generate tailored and coherent responses for user inputs. However, most of existing systems produce universal and non-informative responses, and they have not gone beyond chitchat yet. To tackle these problems, 7th Dialog System Technology Challenges (DSTC7-Track2) was developed to focus on building a dialogue system that produces informational responses that are grounded on external knowledge. In this study, we propose a Memory-augmented Hierarchical Recurrent Encoder-Decoder, called MHRED, that grounded on both multi-turn dialogue context and external knowledge. Furthermore, we apply a combination of multiple dialogue systems. Our final system is an ensemble that combines three modules: a generation-based module, a retrieval-based module, and a reranking module. First, responses are generated by MHRED, and retrieved from a pre-defined database focusing on informativeness. Next, the reranking module sorts these candidates using several hand-crafted features, and finally it selects a response with the highest score. Therefore, this system can return diverse and meaningful responses from various perspectives. Experimental results show that our proposed MHRED outperforms strong baseline models and combining multiple dialogue systems significantly improves the automatic evaluation and human evaluations.

Directional attention weaving for text-grounded conversational question answering

Compared to the traditional single-turn question answering (QA) based on text, also known as machine reading comprehension (MRC), conversational MRC tasks aim to enable models to understand the question according to not only the information in the passage but also the previous question-answer pairs. In this paper, we introduce a directional attention weaving (DAW) mechanism to represent dialog context more comprehensively in two aspects. First, we use self-attention on each level of the question-aware passage hidden representation to collect global information for each question as conversation history input, which is referred to multi-round self-alignment. Second, different from existing methods using forward recurrent neural networks (RNNs) to pass the limited dialog context information, which is inefficient to deal with long conversation history and topic return, a unidirectional attention mechanism is proposed to build the history-aware passage representations by computing the attentions along the dialog progression. The DAW mechanism is used to build our model named CANet. We evaluate the proposed model on the CoQA dataset and achieve competitive results in published models, outperforming the published baselines on the hidden test dataset by a substantial margin. The further analysis shows that our DAW mechanism can be flexibly applied to powerful single-turn MRC models, and effective to augment the representations of dialog history to cater to long conversation and topic return settings.

Music, Politics, and the Liminality of the Havana Jazz Plaza Festival in the Obama Era

AbstractAfter the Obama administration (2009–17) began authorizing musical exchanges with Cuba in 2009, Havana's music festivals became a primary site for transnational interactions and a public face for US-Cuban engagement while politicians worked towards normalization in secret. This article uses field research from the Havana International Jazz Festival, interviews with festival participants, and media coverage to explore Cuban music festivals as politically liminal spaces where musical and political life commingled to reflect the changing US-Cuban relationship. While diverse lineups attracted international tourists, artists faced bureaucratic challenges to legally traverse the Florida Straits and create music in the context of intercultural dialogue. Despite these difficulties, Havana as a festival space encouraged musicians to defy genre conventions, explore cultural commonalities, and negotiate social differences on stage during the Havana Jazz Plaza Festival. These transnational interactions culminated in Arturo O'Farrill's albumCuba: The Conversation Continues,which was recorded by US and Cuban musicians in Havana during the 2014 jazz festival and is characteristic of festival exchanges in its representation of a more harmonious international relationship.

Subjective development process as a path to school learning: the classroom as a dialogic relational context (El proceso de desarrollo subjetivo como un camino para el aprendizaje escolar: el aula como un contexto relacional-dialógico)

This paper aims to discuss the subjective development process of a child in the classroom, considering aspects that have facilitated school learning. A case study is presented of a second-year student in a public elementary school in Brazil, who, at the beginning of the research, had difficulties with socialization and learning. The theoretical framework used is the Theory of Subjectivity, within a cultural-historical approach, together with its epistemological and methodological proposals, Qualitative Epistemology and the constructive-interpretative method. The main instruments were conversational dynamics and various interactive sessions, which have been expanded to the classroom context, since they generated valuable dialogical, social and relational context. The participant has become active, has gained her place in the classroom, has engaged in activities, and has thus significantly advanced in her learning. The outcomes achieved by the child point to the importance of creating everyday pedagogical strategy contexts which are less expositive and more dialogical, so new networks of relationships favour the coordination of operational processes of learning with processes of subjective development.

Exploiting Text Matching Techniques for Knowledge-Grounded Conversation

Knowledge-grounded conversation models aim at generating informative responses for the given dialogue context, based on external knowledge. To generate an informative and context-coherent response, it is important to conjugate dialogue context and external knowledge in a balanced manner. However, existing studies have paid less attention to finding appropriate knowledge sentences from external knowledge sources than to generating proper sentences with correct dialogue acts. In this paper, we propose two knowledge selection strategies: 1) Reduce-Match and 2) Match-Reduce and explore several neural knowledge-grounded conversation models based on each strategy. Models based on Reduce-Match strategy first distill the whole dialogue context into a single vector with salient features preserved and then compare this context vector with the representation of knowledge sentences to predict a relevant knowledge sentence. Models based on Match-Reduce strategy first match every turn of the context with knowledge sentences to capture fine-grained interactions and aggregate them while minimizing information loss to predict the knowledge sentence. Experimental results show that conversation models using each of our knowledge selection strategies outperform the competitive baselines not only in terms of knowledge selection accuracy but also in response generation performance. Our best model based on Match-Reduce outperforms the baselines in the comparative studies with the Wizard of Wikipedia dataset. Also, our best model based on Reduce-Match outperforms them with the CMU Document Grounded Conversations dataset.

An Encoding Mechanism for Seq2Seq based Multi-Turn Sentimental Dialogue Generation Model

Abstract For the multi-turn dialogue generation task, Seq2Seq model lacks representation of sentimental information and dialogue context in the encoder, resulting in generating sentimentally poor and contextually independent responses. To address these problems, in this paper, we propose an encoding mechanism for Seq2Seq based multi-turn sentimental dialogue generation model. To make the Seq2Seq model contain sentiment, we build sentimental word vectors and sentimental sentence vectors to strengthen the sentimental information representation. To make the generated responses more context sensitive, we build semantic vector and contextual vector. Experiments show that Seq2Seq based multi-turn sentimental dialogue generation model with our proposed methods can generate more sentimental, contextually relevant and high-quality responses.

Social Activity of Contemporary Ukrainian Society: Threat to Internal Stability or Possibility of Social Dialogue

The article substantiates the social nature of the dichotomy of contemporary postmodern society through the analysis of the social content of possible forms of social activity. Using the terminology of S. Deetz’s theory of communication, the authors substantiate that today in Ukrainian society there are three main forms of social activity - consent; involvement; participation. The dominance of a certain form of social dialogue in society determines the nature, dynamics, direction, spatial and temporal social activity limitations. The results of an empirical study on the distribution of forms of social activity in contemporary Ukrainian society are presented. Emphasis is placed on the fact that one of the conditions for reducing the social destructiveness of postmodern society is the value context of social dialogue as a manifestation of social activity that regulates relations between social groups. In postmodern society, there is a multiplicity of variations of social interaction. The social nature of the constructive social activity of contemporary Ukrainian society “lies” in the plane of “social interaction - social dialogue - social trust”. Effective (constructive) social interaction between certain social groups (especially those which have the ability and desire to influence social processes actively) leads to the formation of new social relations, that, in turn, form opportunities for social dialogue between them. This dialogue influences the evolution (progressive reforming) of the structural organization of society, which is a prerequisite for the stability of society, decreasing its potential social destructiveness (especially in a situation of social challenges).

Introduction

Introduction Margaret Shrimpton Masson, Oscar Ortega Arango, and Nadia Celis Salgado The Italian theater director, Eugenio Barba, founder of Odin Teatret, wondered: “But, beyond the islands […] what is it that exists? What and who is to be found there?” (Barba 1986:470). To this, he answers: “There are people who live in a nation, in a culture. And there are people who live in their own bodies” (1986:471). This concern, which was first applied by Barba to the dramatic experience, and is at the base of his Anthropology of Theater and Spectacle, works better when we reflect not only on the identity of the other, over there (beyond), but rather on the precise identity of the one who asks, situated here. In the Forward to the second edition of Truth and Method, Gadamer attempted to solve this concern based on Heidegger’s idea of horizon, which: “denotes the basic being-in-motion of Dasein that constitutes its finitude and historicity, and hence embraces the whole of its experience of the world. Not caprice, or even an elaboration of a single aspect, but the nature of the thing itself makes the movement of understanding comprehensive and universal” (Gadamer 2004:xxv). This assertion leads to the idea of the co-existence of a variety of partial horizons, each immersed in a totally encompassing horizon where human life and intellect play out. Consequently, the comprehension of the other is conceived as a system of relations that, as well as dissolving the theory-praxis dichotomy, can also form a network out of the utility of knowledge. This dossier explores the cultural production of the non-insular, mainland and diasporic Caribbean; a Caribbean space beyond the islands, one that we see as oscillating between a set of imaginaries drawn along pan-regional historical-cultural lines and national projects, that often leave these Caribbean spaces at the margin. Our research discusses the mainland Caribbean, seen as a series of networking trans-insular regional territories, that constantly challenge and explore hegemonic identities, cultural expressions and expectations. Mainland Caribbean spaces, often represented within both continental and national discourses as in some way “out of place,” “on the margins,” or even “empty,” display a rooted identity as well as rhizomatic interregional connections (Glissant 1997). Rooted identities, constructed in cultural texts through ideas and tropes of belonging, display their rhizomatic connections through complex regional performances where rootedness repeats and interacts in multiple layers and with overlapping patterns of identity. By using a transdisciplinary and multilingual approach, our research reveals both these specific connections as well as the shared identities that go beyond national identifications, to construct a regional, mainland Caribbean dialogue. In this sense, ideas such as the ones put forward by Portuguese [End Page 3] philosopher Boaventura de Sousa Santos are very appealing. He argues that a “diatopic hermeneutics” intends to “offer a translation process between orders of knowledge belonging to diverse cultural systems, as well as comprising a model for intercultural dialogue” (in, Vergalito 2009:20). In order to achieve this, we must strive to end cultural waste (brought about by western modernity) through the recovery of forgotten, marginal or disqualified cultural fragments; thus leading to the achievement of a comprehensive, liberating, multidimensional and intercultural “cosmopolitan reasoning.” This path to understanding the other and to a fusion with self from the fragmentary, is of primary importance, as it is at the very center of the encounter of the horizons of others: Diatopical hermeneutics is based on the idea that the topoi of an individual culture, no matter how strong they may be, are as incomplete as the culture itself […] The objective of diatopical hermeneutics is therefore not to achieve completeness […] but on the contrary, to raise the consciousness of reciprocal incompleteness to its maximum possible by engaging in the dialogue, as it were, with one foot in one culture and the other in another. Herein is its diatopical character. (De Sousa Santos 2002:273) It is from this context of diatopical dialogue that we approach the Caribbean. The writers and editors of this volume work and reside in different parts of the Caribbean and the diaspora, and came together sponsored by the research project “Representaciones literarias...

Leo Tolstoy And Writers of World Literature Literary Reflections – Elchin Efendiyev

It is a study of literary interaction questions and the identification of mutual enrichment patterns that currently acquire particular significance. The concept of foreign literary experience becomes ever more profound and diverse and is realized by a creative rethinking (not imitation or adoption) of another literature’s achievements. This paper aims to identify the profound influence of world literature on Tolstoy and vice-versa: the influence his creative works had on European literature. The paper shows the need to study the originality of Tolstoy’s artistic legacy’s foreign reception and, therefore, complement the overall picture of perception and functioning of the writer’s creation in the foreign literary context and cultural environment. The study of this theme is very significant from the standpoint of modern globalization, dialogue between cultures. The novelty here lies in the fact that the question of how Tolstoy’s works have been received within the context of creative cross-cultural dialogue has not been given sufficient attention within international comparative studies. There is no systematizing and summarizing research in the national science about a writer’s perception and peculiarities of appraisal of writer’s works involving the Azerbaijani studies material, drawing parallels with the national literature. For this consideration of Tolstoy’s work, the conception of Azerbaijani prose writing is taken to represent a World literary context. The outstanding playwright Elchin Efendiyev had due regard for Tolstoy’s creative work and his particular creative perception of the world. This work’s theoretical purpose is to develop a scientific paper that will expand understanding of the reception of an outstanding writer’s creative work by a western creative consciousness and will present a picture of international cultural ties.

Expression Orale: Amérique Du Nord by Michèle Barféty et al.

Reviewed by: Expression Orale: Amérique Du Nord by Michèle Barféty et al. Shannon Becker Barféty, Michèle, Patricia Beaujoin, Fabien Olivry, et Denis Liakin. Expression orale: Amérique du Nord. CLE International, 2019. ISBN 978-2-09-035176-7. Pp. 128. Part of the Compétences series, Expression orale is intended for students at the B1 level according to the Common European Framework of Reference for Languages, which corresponds to approximately the Intermediate Mid to Advanced Low level in ACTFL standards. It is important to note that this book is intended for learners who will be communicating with speakers of Québécois French. The book is divided into five units, each containing three lessons and followed by a quiz called a "bilan." The lessons are divided into sections called "imiter," "interpréter," and "s'exprimer," a thoughtful strategy that allows the learner to progress from exercises centered on identification and repetition, to answering open-ended questions based on written and oral prompts, and finally to text-based discussions leading to a debate. Importantly, Expression orale maintains a focus on communication throughout, only addressing grammar in boxes of "outils" that provide students with important sentences and phrases that they can incorporate into their production. Among other tasks, learners are asked to verify oral comprehension, to repeat dialogues, to use clues to guess the context of dialogues, to exchange information, and to role play based on visual and audio prompts. This culminates in reading a text, answering discussion questions, and preparing a debate with classmates. After each unit, learners complete the bilan, a brief quiz worth 25 points containing only two exercises: one that requires them to fill in the blanks of a written dialogue using the various outils, or grammatical structures, from the preceding lesson, and one that asks them to fill in the blanks of a short text using words from a provided list of vocabulary. A particularly interesting activity, which appears only in the first bilan and is thereafter replaced by the text completion task, asks learners to read a sentence and decide which of three sentences below it is the best reformulation, i.e., containing the same information and synonyms of key words. Native speaker dialogues are used throughout the book, with audio files available online and transcripts in the back of the book. The quality of the dialogue recordings is high, and the speakers articulate clearly and at an appropriate speed for this level of learner. It is easy to imagine using elements of this book even beyond a conversation course, for example in a survey course to introduce learners to Québécois French. Visually, the format of the chapters is easy to follow, with clearly delineated sections and "outils" boxes that stand out from the text, and numerous pictures throughout. [End Page 201] Shannon Becker Northern Illinois University Copyright © 2020 American Association of Teachers of French

Нараджэнне індывідуальнасці з духу беларушчыны: жыццёвыя траекторыі і канцэптуальна-эмацыйныя матрыцы беларускага інтэлігента ў рамане Віктара Вальтара Роджаныя пад Сатурнам

<p>Artykuł poświęcono analizie powieści Wiktora Waltera <em>Urodzeni pod Saturnem </em>(<em>Роджаныя пад Сатурнам</em>), opisującej życie i dylematy ideologiczne białoruskich studentów w międzywojennej Pradze. Wybór przedmiotu badań jest nieprzypadkowy: analizowany utwór posiada wysoką wartość artystyczną, nie był dotąd badany, ponadto w literaturze nie ustalono roli, jaką odegrała Praga w historii rozwoju życia intelektualnego na Zachodniej Białorusi w latach 20. XX w. Utwór opisuje nie tylko świat wartości, jakimi żyła ówczesna młodzież inteligencka, ale także jej kondycję psychiczną. <em>Urodzeni pod Saturnem </em>pokazuje także, w jaki sposób studenci białoruscy rozumieli takie pojęcia, jak „ojczyzna” i „inteligencja”, jak wyobrażali sobie przyszłość oraz swoje powołanie. I to właśnie rekonstrukcja płaszczyzn semantycznej, egzystencjalnej, psychologicznej i społecznej była głównym celem badania. W utworze <em>Urodzeni pod Saturnem </em>autor nawiązuje do tradycji powieści polifonicznej (według terminologii Michała Bachtina), prezentującej koncepcje i teorie filozoficzne w wypowiedziach bohaterów. Dlatego też główną strategią badawczą zastosowaną w niniejszym artykule jest podejście łączące ze sobą historię idei oraz tzw. historię emocji. Przeprowadzone analizy doprowadziły do konstatacji, że tragedia Tuchowskiego, głównego bohatera powieści, jest także tragedią pierwszego pokolenia białoruskiej inteligencji. Pokolenie to musiało udźwignąć ciężar odkrywania własnej indywidualności w sytuacji zerwania związku z modelem tradycyjnym preferowanym przez zbiorowość, jak również wzięło na siebie obowiązek konfrontacji z: nową ideologią narodowo-społeczną, pogwałceniem historii oraz koniecznością określenia własnej roli w świecie.</p>

But let your ‘Yes’ be ‘Yes,’ and your ‘No,’ ‘No.’ Meaning construction in medical encounters

This paper investigates interpretation in medical context. Our question is how institutional context influences the utterance meaning: if it is really triple layered (literal, utterance-type or pragmatic, Levinson 2000), or rather a continuum (Wilson 2016). Even idiomatic language use (Kecskes 2017) can induce uncertainty and obscurity, which can be and has to be solved in the given dialogue or discourse context (Wilson and Kolaiti 2017). The paper analyzes various medical encounters in a formal pragmasemantic model called ÂeALIS (Alberti and Kleiber 2014). The benefit of applying this system is that it represents the interlocutors’ mental states (beliefs, desires, and intentions) supplemented by the parameter of authority, by which the occurring mismatches can be captured formally. We have found that two main types of mismatch can be differentiated, and both of them can be originated from the fact that the context is not the same for the participants. Our findings support the view that meaning construction is rather flexible and context-sensitive: it can be considered as wandering along the meaning continuum without any clues.

Romance «Blacked Night» by O. Tadeush based on Sappho’s poem: dialogue in “large time”

Romance «Blacked Night» by O. Tadeush based on Sappho’s poem: dialogue in “large time”

Materiais do Ministério de Educação do Brasil: das concepções de linguagem às políticas linguísticas para o ensino de surdos

Neste artigo, analisa-se a política linguística do governo federal por meio dos materiais didático-pedagógicos direcionados à formação de professores do ensino de língua para surdos. A questão problema é: qual concepção de ensino de língua predomina como política oficial nesses materiais? O objetivo é descrever, contextualizar e analisar a política linguística proposta nesses materiais, indicando as mudanças que ocorreram, conforme os discursos constitutivos. Trata-se de uma pesquisa centrada na perspectiva dialógica da linguagem de Bakhtin e do Círculo, em especial, na concepção de linguagem de Bakhtin (2016) e Volóchinov (2017), no que tange a discussão de política linguística de formação dos professores da educação de surdos entre 1979 e 2010. Os resultados indicam que a primeira política linguística a concepção de língua na educação especial, adotada, entre 1979 e 1995, é baseada na reabilitação da fala e da escrita; a segunda, entre 1997 e 2002, é baseada na gramática de Língua Portuguesa e da Língua Brasileira de Sinais-Libras por meio de sinais, frases e diálogos em contextos; a terceira, entre 2003 e 2010, é baseada na leitura e produção de textos em Língua Portuguesa como L2, tendo a Libras como L1; a Língua Portuguesa escrita e a Língua de Sinais estão correlacionadas no espaço de Atendimento Educacional Especializado (AEE), como política de inclusão nas redes públicas de ensino.

Transfer Hierarchical Attention Network for Generative Dialog System

In generative dialog systems, learning representations for the dialog context is a crucial step in generating high quality responses. The dialog systems are required to capture useful and compact information from mutually dependent sentences such that the generation process can effectively attend to the central semantics. Unfortunately, existing methods may not effectively identify importance distributions for each lower position when computing an upper level feature, which may lead to the loss of information critical to the constitution of the final context representations. To address this issue, we propose a transfer learning based method named transfer hierarchical attention network (THAN). The THAN model can leverage useful prior knowledge from two related auxiliary tasks, i.e., keyword extraction and sentence entailment, to facilitate the dialog representation learning for the main dialog generation task. During the transfer process, the syntactic structure and semantic relationship from the auxiliary tasks are distilled to enhance both the word-level and sentence-level attention mechanisms for the dialog system. Empirically, extensive experiments on the Twitter Dialog Corpus and the PERSONA-CHAT dataset demonstrate the effectiveness of the proposed THAN model compared with the state-of-the-art methods.

Dynamic Working Memory for Context-Aware Response Generation

In human-to-human conversations, the context generally provides several backgrounds and strategic points for the following response. Therefore, many response generation approaches have explored the methodologies to incorporate the context into the encoder-decoder architecture, to generate context-aware responses that are remarkably relevant and cohesive to the given context. However, most approaches pay less attention to semantic interactions implicitly existing within contextual utterances, which are of great importance to capture semantic clues of the given dialog context, indeed. This paper proposes a dynamic working memory mechanism to model long-term semantic hints in the conversation context, by performing semantic interactions between utterances and updating context representation dynamically. Then, the outputs of the dynamic working memory are employed to provide helpful clues for the encoder-decoder architecture to generate responses to the given dialog. We have evaluated the proposed approach on Twitter Customer Service Corpus and OpenSubtitles Corpus, with several automatic evaluation metrics and the human evaluation, and the empirical results show the effectiveness of the proposed method.

Magisterium, Theologians, and the Need for Dialogue

Throughout history there have been theological tensions between official church teachers and church theologians, creating at times a divide between both the magisterium and theologians and also between theologians of different methodological approaches. We offer as examples of tension the declarations by the USCCB's Committee on Doctrine (CD) on the “inadequacies in the theological methodology and conclusions” of our book and of the books of three other contemporary theologians. These examples afford us the opportunity both to consider the theological tensions in general and to propose a solution to them. We establish some ecclesial context for dialogue with the CD, calling attention to four factors in this context: first, recent patterns of discourse between theologians and the magisterium in statements issued against particular theologians; second, an important change in the Catholic concept of church; third, an equally important change in how Catholic theologians set about doing theological ethics; and fourth, the reaffirmation of the importance of conscience by the Second Vatican Council in the 1960s and, more recently, by Pope Francis.

How teacher talk guidance during Invention activities shapes students\u2019 cognitive engagement and transfer

BackgroundA key question in K-12 STEM education is how best to guide students as they engage in exploratory learning activities so that students develop transferable knowledge. We investigated this question in a study of teacher talk guidance of an exploratory activity called Invention. In this study, teachers worked one-on-one with students, guiding them as they attempted to invent ratio-based equations of physical science phenomena. We applied the interactive, constructive, active, and passive (ICAP) framework as a theoretical lens through which to explore different forms of teacher talk guidance and resulting student talk. The ICAP hypothesis predicts that constructive engagement leads to greater learning than active engagement, which in turn leads to greater learning than passive engagement. However, students do not always enact the type of cognitive engagement that teachers prompt. In this paper, we work towards three goals: (1) to explore the forms of cognitive engagement prompted by teachers and enacted by students in their talk, (2) to test the ICAP hypothesis in the novel context of teacher-student dialog during Invention, and (3) to identify effective forms of teacher talk guidance for Invention activities and other exploratory STEM learning tasks.ResultsWhile the majority of student talk was active, teachers produced an even distribution of constructive, active, and passive prompts. Teacher and student talk types tended to align, such that students often responded with the type of cognitive engagement teachers invited, with the exception of passive talk. In general, teacher talk showed the most robust relationship with students' abilities to transfer, while teacher-student dialog demonstrated a weaker relationship with transfer, and student talk was not significantly related to transfer. Some evidence for the ICAP hypothesis was found, most prominently in teacher talk, where constructive prompts positively predicted transfer, active prompts were not related to transfer, and passive prompts negatively predicted transfer.ConclusionsThis research implies that teachers should use a large proportion of constructive prompts and relatively few passive ones when guiding students through Invention tasks, when the goal is to provoke transfer of learning to novel contexts. This work also extends the CAP portion of the ICAP hypothesis to teacher-student dialog and underscores the teacher’s critical role in encouraging students to cognitively engage with exploratory STEM tasks in effective ways.