A role for action knowledge in visual object identification

Abstract

1712Which aspects of objects are important in visual identification? An object that is long, cylindrical, and tapered at one end is likely to be recognized as a "pen." Is it only the visual attributes of the pen that are involved in recognizing this object, or do certain nonvisual attributes stored within semantics (i.e., that a pen is held a certain way, and is used for writing) also play a role? We propose that in addition to the visual features of objects, other nonvisual features of these objects, specifically the actions habitually associated with these objects, play a key role in the visual identification and the naming of visually presented objects. We will briefly review the impact of visual and semantic similarity on visual identification before discussing the importance of action information and presenting the current studies. Visual Similarity Influences Visual IdentificationIn healthy participants, visual identification is influenced by visual similarity: Items that come from visually dissimilar categories are identified faster than items that come from visually similar categories (Dickerson & Humphreys, 1999;Humphreys, Riddoch, & Quinlan, 1988;Lloyd-Jones & Humphreys, 1997a, 1997b. Visual similarity also impacts the identification performance of patients with visual agnosia. In these patients, only the identification of certain categories of objects is impaired, a condition labeled category-specific visual agnosia (CSVA). In typical patients with CSVA, the visual identification of items like animals, fruits, vegetables, and musical instruments is impaired while the visual identification of items from other categories, such as tools, vehicles, and body parts is spared (Damasio, 1990;Dixon, Bub, & Arguin, 1997;Forde, Francis, Riddoch, Rumiati, & Humphreys, 1997;Gainotti & Silveri, 1996;Tranel, Logan, Frank, & Damasio, 1997). Arguably, the items that are typically impaired in CSVA patients come from categories where the exemplars are more visually similar to one another than the items that are spared. Gaffan and Heywood (1993) have highlighted the importance of visual similarity by showing that even nonhuman primates have more difficulty identifying drawings of animals than drawings of man-made objects, presumably because the shapes of animals are more visually similar than the shapes of manmade objects.In contrast to animals, objects such as tools share few visual attributes, and patients with CSVA show relatively few problems identifying them. However, for humans who know about animals and about tools, visual similarity may not be the only factor influencing visual object identification; semantic similarity may also play a role. Consider a patient who confuses depictions of lions and tigers but not depictions of hammers and saws. The felines are more visually similar than the tools, but also share more semantic features. Using depictions of real objects (e.g., comparing identification times of different felines to those of different articles of tools) it is difficult to disentangle the effects of vis...

Size-Sensitive Perceptual Representations Underlie Visual and Haptic Object Recognition

“…[44], [45]), and that these representations might contain size-specific information. However, the activation of action representations seems unlikely to have had an important influence on our results for a number of reasons.…”

Learning to use novel objects: A training study on the acquisition of novel action representations

“…Interestingly, one recent study suggests that pairing novel objects with specific actions (e.g. pull, slide or twist) may facilitate the ease with which these objects can be identified (Desmarais et al, 2007). This effect is likely mediated by the acquisition of motor representations that become associated with the object and that in turn facilitate object recognition.…”

“…Desmarais, Dixon, & Roy, 2007;Vuong & Tarr, 2006) or by expertise visual training with the objects (Behrmann, Marotta, Gauthier, Tarr, & McKeeff, 2005;Tarr & Cheng, 2003). However, relatively little is known about the effects of direct action experience on object recognition.…”

Perceptual and Conceptual Visual Rhetoric: The Case of Symmetric Object Alignment

“…Identifying an object can be described as establishing a correspondence between some external stimulus in the visual field (a percept) and our internal representation of it (a concept) that, among other things, specifies its perceptual attributes (Desmarais, Dixon, & Roy, 2007;Humphreys & Forde, 2001). Yet establishing this link always involves another type of visual information: the way we see a certain object, or how it is shown to us.…”