Incremental and predictive discourse processing based on causal and concessive discourse markers: ERP studies on German and English
Incremental and predictive discourse processing based on causal and concessive discourse markers: ERP studies on German and English
复制标题
基于因果和让步话语标记的增量和预测话语处理：德语和英语的 ERP 研究
DOI：
--
复制
发表时间：
2014
期刊：
影响因子：
--
通讯作者：
F. Delogu
中科院分区：
文献类型：
--
作者：
Heiner Drenhaus;Vera Demberg;J. Köhne;F. Delogu
基于因果和让步话语标记的增量和预测话语处理：对德语和英语的ERP研究Heiner Drenhaus（drenhaus@coli.uni-saarland.de）萨尔兰大学，66123 Saarbrucken Germany Vera Demberg（vera@coli.uni-saarland.de）萨尔兰大学，66123 Saarbrucken德国Judith Kohne（judith. uni-bamberg.de）Otto-Friedrich-Universitat班贝格，德国Francesca Delogu（delogu@coli.uni-saarland.de）萨尔兰大学，66123萨尔布吕肯德国摘要有一些证据支持话语连接词促进文本理解的假设和话语连接词处理是高度递增的假设。然而，很少有人知道的方式，不同的话语连接帮助理解，以及他们是否可以引发即将到来的内容的预测。在目前的ERP研究德语和英语，我们调查的效果一致性和连接类型（因果/让步）的预测目标所指（2x2设计）。我们发现了一个前端中心的积极让步连接相比，因果关系的条件和N400样的效果，为目标名词短语的不一致相比，一致的条件。我们的研究结果表明，在连接器（正面积极性）的上下文更新的效果，并提供证据的话语连接器的基础上预测即将到来的内容。保留字：句子处理，ERP，N400;正面积极性，可预测性，话语连接词，话语关系介绍话语连接词（例如因此，然而，相反）已被证明有助于理解，至少在使用得当时（例如，米利斯和贾斯特，1994年）。然而，不太清楚的是，不同类型的话语连接语以何种方式帮助加工，特别是，它们是否会引起对即将到来的语言材料的预测。现有的关于话语关系处理的研究表明，话语连接词与话语的早期部分（例如，Traxler等人，1997年）。以往关于话语关系加工的研究也表明，人们在加工文本时存在一些普遍的认知偏差，这导致他们比让步话语关系更容易推断出因果关系和连续关系（Murray，1995; Sanders，2005; Bengym，2005）。Kuperberg等人（2011）的ERP研究与此特别相关，他们发现了将相邻句子解释为因果关系的默认证据：因果无关的场景比中间相关的场景诱发更大的N400，反过来，比因果相关的场景诱发更大的N400。因果和连续的话语关系通常是预期的话语表达者的观点也与最近对生产数据中话语关系标记的分析一致：Asr & Demberg（2012）发现因果和连续的话语关系不太可能使用显式话语连接符标记，而不是其他（可能不太期望）话语关系。这一观察结果符合统一信息密度假设（Levy & Jaeger，2007），该假设认为，当可选语言元素（如话语连接词）不传达太多新信息时，即当它们标记预期时，它们可以被省略。话语关系。两者合计，这些研究结果导致的预测，话语连接信号的不连续性应该更难处理，因为他们需要重新解释的话语关系，假设默认。在目前对德语和英语的ERP研究中，我们比较了因果语篇连接词和让步语篇连接词的处理。从理论的角度来看，让步话语连接词代表了与因果连接词的有趣比较，因为它们被认为代表了相同类型的关系，但极性不同（Louwerse，2001; Sanders等人，1992年），有时被称为“消极因果关系”（Konig & Siemund，2000年）。先前的工作比较了视觉世界范式中因果关系和让步的处理（Kohne & Demberg，2013）表明，因果和让步话语标记语都可以快速整合到在线理解中，并且处理让步会导致寻找替代方案。虽然视觉世界范式在通过视觉注意揭示句子理解过程中的预测方面很有用，但它并没有对潜在的过程提供太多的见解。Kohne和Demberg在阅读中使用眼球运动的实验还表明，因果关系中的高度增量处理和预测。然而，没有证据表明，在特许权方面存在着递增处理和快速融合的情况。作者推测，他们的结果可能是由于处理优惠或将其与即将推出的内容整合的难度增加。
Incremental and predictive discourse processing based on causal and concessive discourse markers: ERP studies on German and English Heiner Drenhaus (drenhaus@coli.uni-saarland.de) Saarland University, 66123 Saarbrucken Germany Vera Demberg (vera@coli.uni-saarland.de) Saarland University, 66123 Saarbrucken Germany Judith Kohne (judith.koehne@uni-bamberg.de) Otto-Friedrich-Universitat Bamberg, Germany Francesca Delogu (delogu@coli.uni-saarland.de) Saarland University, 66123 Saarbrucken Germany Abstract There is some evidence for both the hypothesis that discourse connectors facilitate text comprehension and the hypothesis that processing discourse connectors is highly incremental. Less, however, is known about the way in which different discourse connectors help comprehension and whether they can elicit predictions of upcoming contents. In the present ERP studies on German and English we investigated the effect of congruency and connector type (causal / concessive) on the prediction of a target referent (2x2 design). We found a fronto-central positivity for the concessive connector compared to the causal condition and an N400-like effect for the target noun phrase for incongruent compared to congruent conditions. Our results suggest an effect of context-updating at the connector (frontal positivity) and provide evidence for prediction of upcoming content based on the discourse connector. Keywords: Sentence Processing, ERP, N400; Frontal Positivity, Predictability, Discourse Connectors, Discourse Relations Introduction Discourse connectors (such as therefore, however, instead) have been shown to facilitate comprehension, at least when used appropriately (e.g., Millis & Just, 1994). It is, however, less clear in which way different types of discourse connectors help processing and, in particular, whether they elicit predictions of upcoming linguistic material. Existing work on the processing of discourse relations revealed that discourse connectors are rapidly and incrementally integrated with earlier parts of the discourse (e.g., Traxler et al., 1997). Previous research on the processing of discourse relations also suggests that people have some general cognitive biases when processing text, which lead them to more easily infer causal and continuous than, for instance, concessive discourse relations (Murray, 1995; Sanders, 2005; Brehm, 2005). Of particular relevance here is an ERP study by Kuperberg et al. (2011), who found evidence for a default to interpret adjacent sentences as standing in a causal relationship: causally unrelated scenarios induced a larger N400 than intermediately related scenarios which, in turn, elicited a larger N400 than causally related scenarios. The view that causal and continuous discourse relations are generally expected by comprehenders is also consistent with recent analyses of discourse relation marking in production data: Asr & Demberg (2012) find that causal and continuous discourse relations are less likely to be marked using explicit discourse connectors than other (presumably less expected) discourse relations. This observation is in accordance with the uniform information density hypothesis (Levy & Jaeger, 2007), which holds that optional linguistic elements (like discourse connectors) may be omitted when they do not convey much new information, i.e. when they mark an expected discourse relation. Taken together, these findings lead to the prediction that discourse connectors which signal discontinuity should be more difficult to process because they require a re-interpretation of the discourse relation that was assumed by default. In the present ERP studies on German and English, we compare the processing of causal versus concessive discourse connectors. Concessive discourse connectors represent an interesting comparison to causal connectors from a theoretical perspective, as they have been argued to represent the same type of relation but to differ in polarity (Louwerse, 2001; Sanders et al., 1992), and have sometimes been referred to as “negative causals” (Konig & Siemund, 2000). Previous work comparing the processing of causals and concessives in the visual world paradigm (Kohne & Demberg, 2013) showed that both causal and concessive discourse markers can be rapidly integrated into on-line comprehension and that processing the concessive leads to a search for alternatives. While the visual world paradigm is useful in revealing predictions during sentence comprehension through visual attention, it does not give much insight into the underlying processes. Kohne and Demberg’s experiment using eye-movements in reading additionally indicates highly incremental processing and prediction in causals. However, no evidence for incremental processing and rapid integration was found for concessives. The authors hypothesize that their results may be due to an increased difficulty of processing concessives, or of integrating them with upcoming content.
登录
查看更多内容
影响因子：
24.8
作者：
Kutas M;Federmeier KD
通讯作者：
Federmeier KD
影响因子：
3.7
作者：
Voss JL;Federmeier KD
通讯作者：
Federmeier KD
影响因子：
5.7
作者：
Bridger, Emma K.;Bader, Regine;Mecklinger, Axel
通讯作者：
Mecklinger, Axel
DOI：
10.1016/s0926-6410(03)00086-7
发表时间：
2003-06-01
期刊：
COGNITIVE BRAIN RESEARCH
影响因子：
--
作者：
Kuperberg, GR;Sitnikova, T;Holcomb, PJ
通讯作者：
Holcomb, PJ