Defining the Origin Relative to a Functional Part

Defining the Origin Relative to a Functional Part Origin has been defined as the intersection point of the axes of a reference frame (Miller & Johnson-Laird, 1976) and is taken to indicate where the reference frame is imposed on the reference object. Given theoretical suggestions that the reference object is represented in a relatively abstract, axial-based form (Landau & JackendoV, 1993; Regier, 1996; Talmy, 1983), it has typically been assumed that the origin is imposed on the basis of the geometric properties of the reference object, most usually at its center of mass (Gapp, 1995; Regier, 1996; Schirra, 1993). However, Carlson- Radvansky et al. (1999) demonstrated that the identity of the reference and located objects and their functional interaction play a significant role in defining the origin. Specifically, they presented participants with pairs of pictures of real-world objects and asked participants to place one object above or below the other object. Placements of the located object were assumed to reflect the best use of these spatial terms. Given that other paradigms have shown that the best use falls on the axis of the reference frame (Carlson-Radvansky & Logan, 1997; Hayward & Tarr, 1995; Logan & Sadler, 1996), these placements are taken to indicate the origin of the reference frame (for discussion, see Carlson, 2000). The pairs of objects were created with the constraint that the located object would typically be placed above or below a given part of the reference object in order to fulfill a particular function. This part is referred to as the functional part. For example, a tube of toothpaste (located object) is typically placed above the bristles (functional part) of a toothbrush (reference object) in order to fulfill the function of putting toothpaste on the toothbrush. A coin (located object) is typically placed above the slot (functional part) in a piggy bank (reference object) in order to insert the coin in the bank. The reference objects were photographed from a sideways perspective that oVset the functional part from the center of mass of the object. This enabled an examination of whether participants would define the best placement of ‘‘above’’ relative to the center of mass of the object, indicating reliance on geometric factors for defining the spatial term or relative to its functional part, indicating a sensitivity to the identity of the reference object and its functional relationship with the located object. Using Spatial Language 139Placements were measured relative to a line running through the center of mass of the objects, with deviations from this line toward the functional part coded as a functional bias. The critical result was that all objects showed a functional bias. Given that the objects varied in how far apart the functional part was from the center of mass of the object, the best way to characterize the functional bias was to express it in terms of the percentage of this distance, with 100% indicating a placement directly over the functional part and 0% indicating a placement directly over the center of mass. On average, located objects were placed at positions that corresponded to 72% of the distance between the center of mass and the functional part. This deviation was significantly smaller, although still positive (45%), when a functionally unrelated object matched in size and shape was used instead of the functionally related object (e.g., ring versus coin for the piggy bank; tube of oil paint versus tube of toothpaste for the toothbrush). The fact that there was a functional bias for these unrelated objects indicates an influence of the functional parts of the reference object on defining where the reference frame would be imposed. The fact that the functional bias was stronger for functionally related located objects indicates an influence of the functional interaction between the objects on setting the origin