Refine
Year of publication
Document Type
- Article (35125)
- Doctoral Thesis (6460)
- Monograph/Edited Volume (5513)
- Postprint (3236)
- Review (2280)
- Part of a Book (989)
- Other (870)
- Preprint (566)
- Conference Proceeding (528)
- Part of Periodical (450)
Language
Keywords
- Germany (197)
- climate change (171)
- Deutschland (137)
- European Union (78)
- Sprachtherapie (77)
- machine learning (75)
- diffusion (74)
- Patholinguistik (73)
- morphology (73)
- patholinguistics (73)
Institute
- Institut für Biochemie und Biologie (5355)
- Institut für Physik und Astronomie (5316)
- Institut für Geowissenschaften (3539)
- Institut für Chemie (3438)
- Wirtschaftswissenschaften (2635)
- Historisches Institut (2482)
- Department Psychologie (2311)
- Institut für Mathematik (2129)
- Institut für Romanistik (2105)
- Sozialwissenschaften (1882)
Referential Choice
(2016)
We report a study of referential choice in discourse production, understood as the choice between various types of referential devices, such as pronouns and full noun phrases. Our goal is to predict referential choice, and to explore to what extent such prediction is possible. Our approach to referential choice includes a cognitively informed theoretical component, corpus analysis, machine learning methods and experimentation with human participants. Machine learning algorithms make use of 25 factors, including referent’s properties (such as animacy and protagonism), the distance between a referential expression and its antecedent, the antecedent’s syntactic role, and so on. Having found the predictions of our algorithm to coincide with the original almost 90% of the time, we hypothesized that fully accurate prediction is not possible because, in many situations, more than one referential option is available. This hypothesis was supported by an experimental study, in which participants answered questions about either the original text in the corpus, or about a text modified in accordance with the algorithm’s prediction. Proportions of correct answers to these questions, as well as participants’ rating of the questions’ difficulty, suggested that divergences between the algorithm’s prediction and the original referential device in the corpus occur overwhelmingly in situations where the referential choice is not categorical.
Referential Choice
(2016)
We report a study of referential choice in discourse production, understood as the choice between various types of referential devices, such as pronouns and full noun phrases. Our goal is to predict referential choice, and to explore to what extent such prediction is possible. Our approach to referential choice includes a cognitively informed theoretical component, corpus analysis, machine learning methods and experimentation with human participants. Machine learning algorithms make use of 25 factors, including referent’s properties (such as animacy and protagonism), the distance between a referential expression and its antecedent, the antecedent’s syntactic role, and so on. Having found the predictions of our algorithm to coincide with the original almost 90% of the time, we hypothesized that fully accurate prediction is not possible because, in many situations, more than one referential option is available. This hypothesis was supported by an experimental study, in which participants answered questions about either the original text in the corpus, or about a text modified in accordance with the algorithm’s prediction. Proportions of correct answers to these questions, as well as participants’ rating of the questions’ difficulty, suggested that divergences between the algorithm’s prediction and the original referential device in the corpus occur overwhelmingly in situations where the referential choice is not categorical.
We report a study of referential choice in discourse production, understood as the choice between various types of referential devices, such as pronouns and full noun phrases. Our goal is to predict referential choice, and to explore to what extent such prediction is possible. Our approach to referential choice includes a cognitively informed theoretical component, corpus analysis, machine learning methods and experimentation with human participants. Machine learning algorithms make use of 25 factors, including referent’s properties (such as animacy and protagonism), the distance between a referential expression and its antecedent, the antecedent’s syntactic role, and so on. Having found the predictions of our algorithm to coincide with the original almost 90% of the time, we hypothesized that fully accurate prediction is not possible because, in many situations, more than one referential option is available. This hypothesis was supported by an experimental study, in which participants answered questions about either the original text in the corpus, or about a text modified in accordance with the algorithm’s prediction. Proportions of correct answers to these questions, as well as participants’ rating of the questions’ difficulty, suggested that divergences between the algorithm’s prediction and the original referential device in the corpus occur overwhelmingly in situations where the referential choice is not categorical.
Speakers’ referential choices differ in the degree of explicitness, ranging from very explicit expressions (such as lexical NPs, e.g., the boy) to less explicit expressions (such as pronouns, e.g., he, and null elements). We examine the referential choices of children with Specific Language Impairment (SLI), in order to differentiate between the linguistic and pragmatic abilities involved in the selection of appropriate referring expressions. Existing findings on referential choices by children with SLI are currently inconsistent and have mainly been reported based on narratives. We used an elicited production task to manipulate the referent’s accessibility by means of two factors: (a) contexts that instantiate different levels of contrast (one vs. two contrasts) and (b) the grammatical role of the expression (subject vs. object). We show that children with SLI and typically developing controls produce more explicit expressions for increased contrast levels and for objects than for subjects. Although children with SLI modify the explicitness of their referring expressions according to the accessibility of referents as typically developing children do, we also find varying production rates between the groups. We discuss how these differences in production rates surface as a consequence of language impairment, although the explicitness of referential choices remains otherwise largely unaffected.
Referential Coding Does Not Rely on Location Features: Evidence for a Nonspatial Joint Simon Effect
(2015)
The joint Simon effect (JSE) shows that the presence of another agent can change one's representation of one's task and/or action. According to the spatial response coding approach, this is because another person in one's peri-personal space automatically induces the spatial coding of one's own action, which in turn invites spatial stimulus-response priming. According to the referential coding approach, the presence of another person or event creates response conflict, which the actor is assumed to solve by emphasizing response features that discriminate between one's own response and that of the other. The 2 approaches often make the same predictions, but the spatial response coding approach considers spatial location as the only dimension that can drive response coding, whereas the referential coding approach allows for other dimensions as well. To compare these approaches, the authors ran 2 experiments to see whether a nonspatial JSE can be demonstrated. Participants responded to the geometrical shape of a central colored stimulus by pressing a left or right button, while wearing gloves of the same or different color as the stimuli. Participants performed the task individually, either by responding to either stimulus shapes (Experiment 1) or by responding to only 1 of the 2 shapes (Experiment 2), and in the presence of a coactor. Congruence between stimulus and glove color affected performance in the 2-choice and the joint tasks but not in the individual go/no-go task. This demonstration of a nonspatial JSE is inconsistent with the spatial response coding approach but supports the referential coding approach.
We report the results from two experiments investigating how referential context information affects native and non-native readers' interpretation of ambiguous relative clauses in sentences such as The journalist interviewed the assistant of the inspector who was looking very serious. The preceding discourse context was manipulated such that it provided two potential referents for either the first (the assistant) or the second (the inspector) of the two noun phrases that could potentially host the relative clause, thus biasing towards either an NP1 or an NP2 modification reading. The results from an offline comprehension task indicate that both native English speakers' and German and Chinese-speaking ESL learners' ultimate interpretation preferences were reliably influenced by the type of referential context. In contrast, in a corresponding self-paced-reading task we found that referential context information modulated only the non-native participants' disambiguation preferences but not the native speakers'. Our results corroborate and extend previous findings suggesting that non-native comprehenders' initial analysis of structurally ambiguous input is strongly influenced by biasing discourse information.
We report the results from two experiments investigating how referential context information affects native and non-native readers’ interpretation of ambiguous relative clauses in sentences such as The journalist interviewed the assistant of the inspector who was looking very serious. The preceding discourse context was manipulated such that it provided two potential referents for either the first (the assistant) or the second (the inspector) of the two noun phrases that could potentially host the relative clause, thus biasing towards either an NP1 or an NP2 modification reading. The results from an offline comprehension task indicate that both native English speakers’ and German and Chinese-speaking ESL learners’ ultimate interpretation preferences were reliably influenced by the type of referential context. In contrast, in a corresponding self-paced-reading task we found that referential context information modulated only the non-native participants’ disambiguation preferences but not the native speakers’. Our results corroborate and extend previous findings suggesting that non-native comprehenders’ initial analysis of structurally ambiguous input is strongly influenced by biasing discourse information.