Written by Josita MAOUENE, Thea IONESCU on . Posted in Special issue: Embodiment And Development, Guest Editors: Josita MAOUENE, Thea IONESCU, Volume XV, Nr. 4

In the last two decades, we have observed an expansion of the definition of cognition (Gentner, 2010a). While the dominant view until the 1980s considered cognition as a separate entity from the body (and for some it still is), that view is declining as growing evidence indicates that cognition cannot be separated easily - or for some not at all – from the body's morphology and its sensorimotor systems nor from the context it is embedded in. As Gentner (2010a) shows in her recent analysis of the history of psychology in cognitive science, embodiment research is growing.

What about it in cognitive developmental psychology? Is the embodiment perspective growing? It certainly is in domains concerned with unveiling certain sensorimotor processes such as those involved in learning how to crawl, walk, avoid obstacles, etc. (Adolph, 2008), learning eye-hand-object coordination (Street, Jones & Smith, 2011), searching for hidden objects (Perry, Samuelson, & Spencer, 2009), or with the processes involved in feedback about the consequences of infants’ actions and how they could entice infants to start reaching (Needham & Libertus, 2011). The growth is also perceptible in domains that go beyond the boundaries of what Piaget called the sensorimotor stage of intellectual development. As such it can be extended to the domains of categorization, language, mathematics, and prelinguistic communication. Some of us study now the relation between social visuo-motor processes and higher level-cognition such as word learning (Pereira, Smith & Yu, 2008), how visual motor movement is connected with statistical word learning (Smith and Yu, in press), and how visual intelligence connects to symbolic process (Smith & Jones, 2011). It is also being investigated in the work on tactile categorization of object properties in infants (Sheya and Smith, 2010), in how gestures influence thought (Beilock & Goldin-Meadow, 2010) and mathematics (Goldin-Meadow, Cook & Mitchell, 2009), how purposeful intentions of others are grasped by very young infants through the perception of body movements, facial expressions and gesture (Gallagher & Hutto, in press). However, there are still domains where an embodiment perspective is, to our knowledge, inexistent (but not unthinkable). For example, the role of the sensory-motor processes in the development of knowledge about the physical world and causal reasoning (Luo & Baillargeon, 2010; Shutts, Condry, Santos, & Spelke; 2009), the role of sensory- motor processes in representational insight (Troseth, Bloom, and Deloache, 2007) and the role of embodiment in analogical reasoning and structural mapping (Gentner, 2010b). We wish to argue that there is also room for embodied and situated cognition for “declining areas” such as representation of knowledge, conceptual semantics, and computational semantics (Gentner, 2010a), in particular in the modeling of developmental semantic networks (Steyvers & Tenenbaum, 2005; Hills, Maouene, Maouene, Sheya & Smith, 2009a,b). Ultimately embodied and situated cognition perspectives will have to propose whether there are such a thing as modal networks or not (Schubert & Semin, 2009) and we predict that the question of the code will reemerge (Pylyshyn, 1981).

All the contributions to this Special Issue reflect developments that are of central relevance and interest to the topic of learning in cognitive science from the point of view of embodiment. The special issue starts with an introductory paper on Embodiment and Development in Cognitive Science from Laakso. Laakso’s introduction aims to give the reader a sense of the main issues regarding the role of embodiment in cognition and development. The paper begins with the history of the “disembodied” perspective that dominated cognitive science from the origins of the discipline through the 1980s, and the history of the “embodied” perspective. Then, it surveys some arguments for the embodied cognition hypothesis and describes some of the empirical evidence that weighs in its favor, with a focus on the developmental literature in particular. In the end it points out important challenges for supporters of the embodied cognition hypothesis.

After this introductory paper, experimental and correlational papers will follow with chronological age of the tested population as the organizational principle. The second contribution is that of Dueker, Portko & Zelinksy on 6- to 11-month-old infants and their mothers entitled Meaningful touch in naturalistic contexts: Haptic input as a cue to the referent of Infant Directed Speech. They report that infants are more likely to be touched, particularly held or held-up when adults speak about topics for which there are no immediately useful physical referents (other than themselves) than when they speak about objects that are present in the immediate environment. Further, infants are likely to experience more simultaneous touch contact points when adults are speaking about a topic that does not have an immediate physical referent than when speaking about a nearby object. But, the location of touch did not vary by topic. They hypothesize that haptic input could affect infant attention to various aspects of the context and also, that haptic input could facilitate multimodal coordination. The third paper is a correlational study of Maouene, Sethuraman, Laakso & Maouene examining the speech of 12 to 23 month-olds entitled The body region correlates of concrete and abstract verbs in early child language. The authors report evidence that embodiment is an important factor to consider in children’s acquisition of verbs, with two major findings: first, those verbs that are more highly associated with a single region of the body, according to adult judgments, are acquired among the first in vocabulary development; second, associations with a specific region of the body may be a better predictor of whether a verb is learned early than the concreteness or abstractness of the verb. The fourth contribution is that of James and Bose with 5- to 7-year-olds entitled Self-generated actions during learning objects and sounds create sensori-motor systems in the developing brain. These authors report four important findings that contribute to enrich the embodied perspective on learning. First, in children, learning about objects actively results in very different brain system recruitment than observing others act. Second, sounds learned actively result in motor, auditory, and visual system recruitment more than passive observation upon subsequent auditory presentation. Third, videos of objects that are learned actively result in motor and visual system recruitment more than passively learned objects. And fourth, when given all learned information about objects through multi-sensory presentation, active learning results in greater visual system recruitment than passive observation. The fifth contribution is that of Wakefield and James with 4- to 7-year-olds and adults and the Effects of Sensori-Motor Learning on Melody Processing across Development. They present two studies on how abstract actions affect perception, and how this may change across development. They address this question by teaching children (4-7 year-olds) and adults to sing melodies, with or without an abstract motor component, and by using functional Magnetic Resonance Imaging (fMRI) to determine how these melodies were subsequently processed. Results show that developmental change occurs during melody processing. Children, contrary to adults, do not process motor signs as meaningful, although associations among sensori-motor systems are still created. These associations lead to motor system recruitment during purely perceptual tasks only when movement is incorporated into the learning episode. Later, in adulthood, motor movement is coded as meaningful to the learning episode. Therefore, as we mature, abstract motor movement become meaningful to the learning event.

Review papers follow, although they are not what one might call "classic style" reviews. The sixth contribution is that of Yoshida and Burling, a paper at the crossing between experimental and theoretical purposes entitled A new perspective on embodied social attention. Yoshida and Burling present a perspective paper, where they first discuss the role of early social input in language learning as it has been treated in the literature so far and then introduce their recent theoretical and empirical embodied direction. The authors hypothesize a link between sensori-motor experiences and embodied attention—specifically how different bodies produce different kinds of attention. Understanding the role of bodily events (the child’s and the child’s social partners’) in early visual experiences provide insight into the development of learning mechanisms and the processes involved in learning disabilities. They share with us their first data on normally developing hearing children, deaf children of deaf families, and children with autism who were observed in a social context using a new child-centered technology. The final contributions are a review of Wagner Cook on Abstract thinking in space and time: Using gesture to learn math and a commentary from Samuelson entitled Abstract thinking in space and time: using the environment to learn words. Cook first briefly discusses some of the evidence for embodied cognition in general, focusing on language and motor processes as these are most relevant for thinking about gesture as an embodied representation. She then discusses evidence that hand gestures are functionally involved in mathematical thinking in both children and adults. Finally, she addresses some possible criticisms and data that are difficult to explain from this perspective. Samuelson’s comments on Cook’s arguments, highlighting how this view of math as embodied offers new insights for our understanding of classic developmental themes, in particular, the continuity versus discontinuity dichotomy. In addition, she presents a brief summary of recent work on how children use their bodies in another realm typically thought of as abstract—understanding referential intent. She presents an embodied account of how children disambiguate speaker intent in novel naming situations and argues that, as in the case of embodied math, an embodied view of cognition can help elucidate developmental mechanisms.

In surveying some of the latest elements brought to the debate about embodied cognition by studies with adult subjects, the contributions of the authors in this special issue help delineate which objections we can discard and which ones we cannot (see Laakso, this issue). For example, arguments according to which abstract concepts like beauty cannot be accounted for by the embodied theory of cognition are not problematic (Mahon & Caramazza, 2008). Cook and Samuelson remind us that very abstract entities like numbers are experiences as are beauty and justice and thus pertaining to the spatiotemporal context and the perception-action systems involved in their processing. However, we do see issues with the mirror neurons account (Arbib & Bota, 2006) since some development data (James & Swain, 2011 and James & Bose, this issue) as well as lesion data in adults (Arevalo, Balso & Dronkers, in press) seem to cast doubt on their existence in humans. An associative theory of embodiment might be a better option. Similarly, for embodied semantics, the results of James and Maouene (2009) and Arevalo, Balso & Dronkers (in press) do not support a strict interpretation of the homuncular topology view where language would be restricted to the network responsible for execution and action, but supports the view of a greater motor-language network of associations. Developmental science will help in that regard - the question is whether adult cognitive development research will take into account this compelling evidence from research on child development.