Symposium Description

Abstract concepts are a controversial and widely debated topic, especially within the theoretical discussion about the embodied vs symbolic nature of language and meaning.

Supported by an extensive body of empirical research, the embodied account of cognition argues that cognition (and therefore language) is tightly related to perceptual and motoric experience. However, the Achille heel of the embodied account of cognition is precisely the (still debated) nature, structure, processing, and modeling of abstract concepts. In particular: how does perceptual experience affect our understanding and semantic representation of abstract concepts (idea, theory, argument), which by definition lack perceptual referents?

This topic is investigated from different angles in our symposium, thanks to the special lectures provided by world class computer scientists and cognitive scientists.


In fields such as cognitive psychology and neuroscience, it has been argued that abstract concepts can be grounded in bodily experiences (see review by Pecher et al., 2011), that neural substrates dedicated to processing sensorimotor information are activated also when we process the words used to describe such concepts (Pulvermüller et al., 2012), and that emotions play a critical role in the understanding and processing of abstract concepts (Vigliocco et al., 2014). In this framework, some scholars suggest a reconciliatory view, claiming that conceptual processing can be both symbolic and embodied (Barsalou et al., 2008; Zwaan, 2014), that conceptual representations are multimodal in nature (Semin et al., 2013), that they can involve the activation of two functionally independent but interconnected representational systems (Paivio 1971;2007; see McRae, Jones 2013 for a review), and that there are redundancies between the semantic information encoded by language and perceptual information (Louwerse, 2011). Such redundancies allow, for example, congenitally blind patients to perform similarly to sighted participants in property generations tasks (Lenci et al., 2013), and can be exploited to model computationally semantic representations starting from corpora of text in a distributional fashion, or exploiting the lexical information encoded in WordNet (Vossen, 2014-2019-ULM-2).

All these views will be presented and discussed I  a high-profile academic debate.

The host

At the Metaphor Lab Amsterdam, which is one of the research groups affiliated to ABC, we conduct interdisciplinary research projects that tackle the structure and functioning of metaphor in thought, language, and communication.

Among the ongoing projects, COGVIM, Cognitive Grounding of Visual Metaphor, aims at understanding how the visual and the verbal modes of expression construct and represent metaphor in a comparative fashion. Starting from the analysis of samples of representative visual and linguistic metaphors extracted from corpora, corpus-driven (semi-automated) distributional methods are applied to 3 different databases of semantic information in order to retrieve and model the similarity that characterizes the two terms compared in a (visual or linguistic) metaphor. Such unprecedented large scale analyses of conceptual content, based on distributional analyses  will help to shed light on the similarity that characterizes metaphorical comparisons in language and in images. The results of COGVIM will deliver a model for the cognitive grounding of visual metaphor.

In this respect, the nature of abstract concepts is particularly relevant for metaphor researchers, because it has been argued for several decades that metaphors are matters of thought, that we think through metaphors, and thanks to these cognitive devices we can understand abstract concepts by comparing them to concrete ones, which in principle can be experienced through our bodies and our sensory modalities (see the Conceptual Metaphor Theory, proposed by Lakoff, Johnson 1980). In this view, we understand concepts such as life through concrete concepts such as journeys, and through deeper cognitive structures such as the source-path-goal image schema (e.g. Gibbs 2006). While the classic metaphor view suggests that abstract concepts are embodied in sensorimotor experiences through metaphorical mappings and image schemas (see for example Gibbs 2006; Dodge, Lakoff 2005) this view has been recently challenged by several metaphor scholars (e.g. Bowdle, Gentner 2005, Steen 2011). The results of the symposium will contribute to shed light on this controversial topic, and will provide a solid ground for developing future theories on how images (and in particular metaphorical images) convey abstract concepts by visual means.