Foot or X down? Response compatibility but no effects of the object's typical location. Several failures to replicate Estes, Verges & Barsalou (2008)

The language embodiment hypothesis predicts that some object words (e.g., bird) elicit spatial activation of their typical location (which is up for bird) through a perceptual simulation of the experienced spatial characteristic of the object (Barsalou, 1999). Estes, Verges, Barsalou (2008) reported that centrally presented words denoting objects usually appearing in the upper/lower portion of the visual field interfere with the identification of unrelated targets appearing up/down on the display, respectively. In a series of 6 experiments, run in different laboratories, we tried to replicate this pattern of results, without success. Instead, we found a spatial compatibility effect so that targets appearing up/down on the display were identified faster with the response key placed in the upper (O)/lower (X) portion of the keyboard, respectively. Critically this effect was independent of the word’s typical location. The results challenge the conclusion drawn in Estes et al. according to which object words orient attention to the object's typical location.