Language, Cognition and Neuroscience ISSN: 2327-3798 (Print) 2327-3801 (Online) Journal homepage: https://www.tandfonline.com/loi/plcp21 Abstraction and concepts: when, how, where, what and why? Eiling Yee To cite this article: Eiling Yee (2019) Abstraction and concepts: when, how, where, what and why?, Language, Cognition and Neuroscience, 34:10, 1257-1265, DOI: 10.1080/23273798.2019.1660797 To link to this article: https://doi.org/10.1080/23273798.2019.1660797 Published online: 02 Oct 2019. Submit your article to this journal View related articles View Crossmark data Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=plcp21 LANGUAGE, COGNITION AND NEUROSCIENCE 2019, VOL. 34, NO. 10, 1257–1265 https://doi.org/10.1080/23273798.2019.1660797 INTRODUCTION Abstraction and concepts: when, how, where, what and why? Eiling Yee a,b aDepartment of Psychological Sciences, University of Connecticut, Storrs, CT, USA; bConnecticut Institute for the Brain and Cognitive Sciences, University of Connecticut, Storrs, CT, USA ABSTRACT ARTICLE HISTORY It is increasingly apparent that sensorimotor information is a constitutive part of conceptual Received 23 July 2019 knowledge. Yet all concepts, even highly concrete ones (e.g. dog) include information that is Accepted 13 August 2019 abstracted across individual episodes of experience, departing somewhat from direct sensory or KEYWORDS motor input. This process of abstraction is the essence of conceptual structure. This Special Issue fi Concepts; semantic memory; brings together developmental, experimental, computational and cognitive neuroscienti c abstraction; abstract perspectives on abstraction. The contributions address questions like: When during development concepts; conceptual do our concepts become less directly tied to sensory or motor knowledge? How (and where) in development the brain does the process of abstraction happen? And what is the role of a concept’s label in abstraction? In answering these questions, the contributions highlight that context–the developmental contexts of our first episodic experiences, and the linguistic contexts that accompany the development of conceptual knowledge in both children and adults–is at the root of conceptual knowledge. Over the past two decades, there have been an increas- reasons, conceptual knowledge must be more than ing number of demonstrations that sensory and motor simply sensorimotor – it seems that it also involves (or embodied) information is a functional part of concep- some “abstraction” away from direct sensory or motor tual knowledge (for discussion, see Matheson & Barsalou, information. 2018). Most of this evidence comes from adults, but a In this article, we use the term abstraction to refer to role for sensory and motor information in conceptual the process that, supported by sensitivity to statistical knowledge is also consistent with evidence that young regularities in our environment (e.g. regularities across children’s conceptual systems are acquired and organ- sensory input and our motor responses to that input), ised via their sensory and motor experiences (for allows us to form and store (or, in exemplar/retrieval- review, see Wellsby & Pexman, 2014). based models, allows us to compute at retrieval) seman- At the same time, however, it is apparent that not all tic information gleaned across our experiences. Via this of conceptual knowledge can be directly attributed to process of abstraction we aggregate across individual simple sensory or motor input. At one end of the spec- episodes of direct sensory or motor experience, such trum, there are many concepts that do not have any that what is most common across our experiences with obvious sensory or motor correlates (e.g. the concept a given object or event gains prominence, and idiosyn- concept). But there are also intermediate cases, such as cratic properties, like a leaf sticking to your dog’s coat, the concept joy, which not only correlate more with carry little weight. It is through this kind of abstraction internal bodily states than with external sensory infor- that we are able to discern what various objects and mation but these internal sensory correlates can vary sig- events have in common, and group them together into nificantly (consider an exuberant joy vs. a calm joy; for concepts. We use the term abstracted to refer to the con- review, see Winkielman, Coulson, & Niedenthal, 2018). ceptual knowledge (which is dynamic in the sense that it In fact, even the concept dog, insofar as it maps on to is continually updated via experience) that is the output animals of a large variety of shapes and sizes, arguably of this process of abstraction. We favour the term should have a relatively indistinct sensorimotor rep- abstracted here because it accentuates that there was resentation (for discussion see Barsalou, 1999). Further- some process, operating across experience, from which more, there is evidence that competent language users the knowledge arose. can derive (at least some) meaning from language As suggested earlier, it may be that all conceptual without relying directly upon sensory and motor knowledge can, in a sense, be described as abstracted. systems (for discussion, see Louwerse, 2011). For these We therefore use the term in a graded fashion, referring CONTACT Eiling Yee [email protected] © 2019 Informa UK Limited, trading as Taylor & Francis Group 1258 E. YEE to knowledge that is “more abstracted” (e.g. the knowl- the creature. In fact, the different levels in hierarchical edge of what an animal is) and “less abstracted” (e.g. descriptions of semantic knowledge (e.g. Collins & Quil- knowledge of what a dog is, or even of who a particular lian, 1969) can be thought of as degrees of abstraction, dog, Zoey, is). We use the term abstract to refer to knowl- with superordinate (i.e. more inclusive) levels corre- edge that does not have any obvious sensory or motor sponding to more abstraction, and subordinate levels basis (e.g. knowledge about the concepts concept or corresponding to less (Rosch, 1978). And insofar as truth) – and, on our view, can therefore only be attained these more abstracted representations of concrete by relying heavily on language or on abstracting across things (e.g. animal) lack clear sensory or motor correlates, diverse episodes of experience. Thus, we consider they are similar to abstract concepts, as defined above.2 abstract knowledge to be conceptual knowledge from The major questions discussed in this Special Issue the more abstracted end of the continuum. Abstract con- focus on exactly this central property of concepts – cepts are composed, predominantly, of abstract knowl- that all concepts, even concrete ones (e.g. dog) require edge, but may have sensorimotor correlates (e.g. the abstraction away from their sensorimotor correlates. concept joy; for related discussion, see Barsalou, This process of abstraction is the essence of conceptual Dutriaux, & Scheepers, 2018). structure. It is from within this perspective that the But why have concepts in the first place? That is, what papers in this Special Issue ask: is the benefit of (via the process of abstraction) develop- ing this conceptual knowledge? Most people who study (1) Given that young children’s conceptual systems are concepts would agree that concepts allow us to make acquired and organised via their sensory and sense of and organise the world – they allow us, when motor experiences, when, and how, during develop- encountering something new, not to have to learn ment does abstraction away from direct experiences from scratch what it does, how we should interact with occur (Sheya & Smith, 2018; Pexman, 2017; Sloutsky it, and how it might change. In this sense, having con- & Deng, 2017; Gentner & Asmuth, 2017)? cepts gives us the ability to generalize – which we (2) What is the role of language, and in particular, a con- define here as the ability to apply knowledge that we cept’s label, in abstraction (Pexman, 2017; Sloutsky & have learned to new situations. Thus, when we use con- Deng, 2017; Gentner & Asmuth, 2017; Connell, 2018; cepts to make inferences or predictions about what will Lupyan & Lewis, 2017; Jones, 2018; Davis & Yee, happen when we encounter something new (e.g. any 2018)? new instance of a dog, or a leaf, or a pencil), we are gen- (3) How do we generalise from (apply) this abstracted eralising (these definitions of generalisation and abstrac- knowledge (Sheya & Smith, 2018; Pexman, 2017; tion follow the conventions in the statistical learning Gentner & Asmuth, 2017; Connell, 2018; Lupyan & literature; see Altmann, 2017).1 The notions of abstrac- Lewis, 2017; Jones, 2018; Davis & Yee, 2018)? tion and generalisation recur repeatedly throughout (4) What is the underlying neurobiology that supports this Special Issue. This is appropriate because (1) we abstraction and generalisation – how/where are have concepts in order to use them and (2) there can such processes grounded in brain function (Sheya be no generalisation without first developing/abstract- & Smith, 2018; Gentner & Asmuth, 2017; Jones, ing the knowledge that we use to make generalisations. 2018; Davis & Yee, 2018)? The point that we use concepts to organise the world (5) How are abstract concepts (e.g. truth and fairness), relates to the notion of different levels of abstraction, i.e. which seem to be devoid of physical and perceptual differences in degree of detail or precision, where the instantiation, learned and represented (Pexman, less detail there is, the greater the level of abstraction/ 2017; Gentner & Asmuth, 2017; Connell, 2018; the more abstracted the concept (e.g. goldendoodle vs. Lupyan & Lewis, 2017; Davis & Yee, 2018)? dog vs. animal). The idea is that the abstraction process (as described above) can produce these different levels Sheya and Smith (2018) begin by arguing for the of abstraction (for discussion of this relationship, see importance of the interactions between the body and Burgoon, Henderson, & Markman, 2013). That is, all its environment on cognitive, and hence, conceptual abstractions minimise the idiosyncratic details of the development.