Physical contact with hot vs. iced coffee has been shown to affect evaluation of the personal warmth or kindness of a hypothetical person (Williams & Bargh, 2008). In three studies, we investigated whether the manipulation of social context can modulate the activation of the metaphorical mapping, kindness as warmth. After priming participants with warm vs. cold temperature, we asked them to evaluate a hypothetical ad-hoc ally or adversary on the kindness dimension, as well as on other qualities used as a control. We expected more extreme evaluations of kindness in the adversary than in the ally condition, and no effects on other ratings. We thus replicated the classical effect of physical warmth on kindness ratings and generalized it to a German-speaking population. In addition, when the two German studies were combined, we found evidence suggesting a contextual modulation of the temperature effect: only out-group members, namely adversaries, were judged as more kind when participants had experienced physical warmth; the effect was not evident in the ally (i.e., in-group) condition. These studies suggest that context can modulate metaphorical activation; they therefore represent an initial attempt to add nuance to our understanding of when embodied metaphors affect our decisions.

Previous research has demonstrated that the scope of speakers’ planning in language production varies in response to external forces such as time pressure. This susceptibility to external pressures indicates a flexibly incremental production system: speakers plan utterances piece by piece, but external pressures affect the size of the pieces speakers buffer. In the current study, we explore internal constraints on speech planning. Specifically, we examine whether individual differences in working memory predict the scope and efficiency of advance planning. In our task, speakers described picture arrays to partners in a matching game. The arrays sometimes required speakers to note a contrast between a sentence-initial object (e.g., a four-legged cat) and a sentence-final object (e.g., a three-legged cat). Based on prior screening, we selected participants who differed on verbal working memory span. Eye-movement measures revealed that high-span speakers were more likely to gaze at the contrasting pictures prior to articulation than were low-span speakers. As a result, high-span speakers were also more likely to reference the contrast early in speech. We conclude that working memory plays a substantial role in the flexibility of incremental speech planning.

Using eye-tracking as a window on cognitive processing, this study investigates language effects on attention to motion events in a non-verbal task. We compare gaze allocation patterns by native speakers of German and Modern Standard Arabic (MSA), two languages that differ with regard to the grammaticalization of temporal concepts. Findings of the non-verbal task, in which speakers watch dynamic event scenes while performing an auditory distracter task, are compared to gaze allocation patterns which were obtained in an event description task, using the same stimuli. We investigate whether differences in the grammatical aspectual systems of German and MSA affect the extent to which endpoints of motion events are linguistically encoded and visually processed in the two tasks. In the linguistic task, we find clear language differences in endpoint encoding and in the eye-tracking data (attention to event endpoints) as well: German speakers attend to and linguistically encode endpoints more frequently than speakers of MSA. The fixation data in the non-verbal task show similar language effects, providing relevant insights with regard to the language-and-thought debate. The present study is one of the few studies that focus explicitly on language effects related to grammatical concepts, as opposed to lexical concepts.

We present in broad outline a theory of document acts, using the influential Supreme Court opinion in Marbury v. Madison (1803) as our principal test case. Marbury has a superabundance of rhetorical questions. They make up a significant and stylistically prominent portion of the total linguistic material of the text, yet they all but disappear from Marbury’s citation history and thus its content as an enduring jurisprudential entity. To account for these facts, we examine Marbury as a whole text addressing a particular situation, as a pastiche of constructions, and as a tool of jurisprudential decision-making. The intersection and independence of these ‘modes of being’ call for an overarching theoretical framework capable of accounting for facets of documents’ existence at three distinct but interpenetrating strata: system, artifact, and construction. We base our theory on primordial cognitive capacities for joint attention and joint commitments, with the strata as consequences of embodied human minds born into and embedded in intersubjective environments filled with and shaped by documents and their circulation. The closed system of United States Supreme Court opinions makes an excellent case for a theory of document acts that will eventually be used to understand and explain more open-ended systems.

Linguistic acceptability judgments are widely agreed to reflect constraints on real-time language processing. Nonetheless, very little is known about how processing costs affect acceptability judgments. In this paper, we explore how processing limitations are manifested in acceptability judgment data. In a series of experiments, we consider how two factors relate to judgments for sentences with varying degrees of complexity: (1) the way constraints combine (i.e., additively or super-additively), and (2) the way a comprehender’s memory resources influence acceptability judgments. Results indicate that multiple sources of processing difficulty can combine to produce super-additive effects, and that there is a positive linear relationship between reading span scores and judgments for sentences whose unacceptability is attributable to processing costs. These patterns do not hold for sentences whose unacceptability is attributable to factors other than processing costs, e.g., grammatical constraints. We conclude that tests of (super)-additivity and of relationships to reading span scores can help to identify the effects of processing difficulty on acceptability judgments, although these tests cannot be used in contexts of extreme processing difficulty.

Some more recent lines of research converge on claiming that human cognitive behavior in general and linguistic discourse in particular cannot reasonably be reduced to one monolithic system of cognitive activity. What this research suggests, rather, is that this behavior exhibits a dualistic organization. In the present paper, two frameworks representing this tradition are contrasted, namely Discourse Grammar and the dual process model. The former rests on observations on language structure and language use, while the latter was developed on the basis of neurolinguistic observations. The two frameworks converge on claiming that there is a significant correlation between linguistic categorization and hemisphere-based brain activity. The present paper argues that this correlation can be related to contrasting linguistic functions associated with each of the two hemispheres.