Spatial Representation of Time in Mandarin-English Bilinguals Jennifer Wu Cognitive Science Honors Thesis University of California, San Diego Advisor: Benjamin Bergen June 2012 Introduction Previous work has shown that time is often mapped to a spatial timeline, most likely to enhance and make communication about time, an intangible concept, easier to understand and utilize (Boroditsky, 2001; Weger & Pratt, 2008). In particular, cultural aspects such as writing and reading direction, as well as linguistic cues including spatialized time metaphors, may influence how one maps time in space. For example, we can “move meetings ahead” or “push deadlines back”, both of which are expression using spatial words. Not only is time associated with certain locations in space, these mappings are so prevalent that they produce compatibility effects, such that stimuli presented in a compatible orientation to the spatiotemporal mapping are responded to faster than stimuli presented in the opposite incompatible manner. For example, when “earlier” is on the left and “later” is on the right, this is a compatible orientation with time running left to right and thus, this would result in faster response times. (Boroditsky, 2001; Fuhrman et al., 2011; Miles et al., 2011; Weger & Pratt, 2008) Notably, although English speakers do not use left to right spatial time metaphors in language, evidence suggests that the reading and writing direction, as well as other graphical representations used in the culture (e.g. number lines), produce a mental timeline flowing left to right, with the left being “earlier” and the right being “later” (Boroditsky, 2001; Fuhrman et al., 2011; Miles et al., 2011; Santiago et al., 2007; Weger & Pratt, 2008). In contrast, for Mandarin speakers, there is an additional top to bottom vertical time mapping, with “earlier” being upward and “later” being downward (Boroditsky, 2001; Fuhrman et al., 2011). Several aspects of Mandarin, such as the traditional vertical reading and writing direction, as well as an increased use of vertical time metaphors (using the words “up” and “down” specifically to mark time), may contribute to the way Mandarin speakers represent time in space (Fuhrman et al., 2011). These two mental timelines appear to be combined in Mandarin-English bilinguals, such that both timelines co-exist within the individual (Boroditsky et al., 2011; Fuhrman et al., 2011; Miles et al., 2011). But what determines which timeline is used and what cues may bias a bilingual individual into using one timeline over the other? An experiment by Miles et al. (2011) asked Mandarin-English bilinguals to arrange photographs of two famous actors, Brad Pitt and Jet Li, in temporal order, but did not specify what orientation or axes in which to do so. They found that Brad Pitt’s pictures were most often arranged left to right along a horizontal axis, and in contrast, Jet Li’s pictures were most often arranged top to bottom along a vertical axis. These results suggest that the context of the pictures, more specifically, the ethnicity of the face, may have served as a cue that caused activation of one of the two existing spatiotemporal mappings in bilinguals. In most previous studies, the main manipulation was in the spatial orientation of the response method, such that adjacent buttons on a keyboard represented “earlier” or “later” in either compatible or incompatible orientations with a typical timeline (Boroditsky et al., 2011; Miles et al., 2011). For example, the “earlier” button would be non-linguistically determined to be the button on the left, whereas the “later” button would be on the right. This could potentially be mapped to a compatible left to right timeline, and thus, would constitute the compatible response condition. In these studies, the visual stimuli were presented sequentially in the same location. As previously mentioned, these studies found robust compatibility effects, such that responses were Page 1 faster in compatible conditions than incompatible conditions. However, a potential problem with these studies is the use of a spatially oriented response, which consequently, required a manipulation of the hands. It is possible that the robust compatibility effect is not reflective of the way people associate time with space, but rather, is simply representative of how people gesture about time. The task of moving of the hands itself may have driven the compatibility effects, and may not generalize to how people think about time, and moreover, how people may map time in space. Thus, using a different method to assess these spatiotemporal maps would be helpful in understanding where the compatibility effects come from. A different approach would be to utilize a non-spatial response and spatially oriented the stimuli. This approach would better allow one to determine if spatially oriented responses are required as a component of the previously seen compatibility effects, or if these mappings of time generalize to non-spatial tasks as well. Therefore, if compatibility effects generalize to a non-spatial task and if the ethnicity of a face can bias bilinguals into using one mental timeline over the other, then the prediction would be to see compatibility effects for Caucasian faces along the horizontal axis only (with no difference on the vertical axis), and to see a compatibility effect along both the horizontal and vertical axis for Asian faces. Methods a. Design & Predictions Face Type Asian OR Caucasian Face Axis Horizontal Vertical Compatibility Compatible Incompatible Compatible Incompatible Locations & Earlier on left Earlier on right Earlier above Earlier below Judgments Later on right Later on left Later below Later above Predictions Caucasian faces Faster Slower No difference Asian faces Faster Slower Faster Slower b. Materials The experiment recruited a total of 27 Mandarin-English bilinguals. There were three portions to the experiment. The first part was an online computer-based task created and run using E-Prime. The stimuli for this part of the experiment consisted of pictures of celebrities and famous people. A total of 50 different faces were used, ten of which were used in a practice block. All pictures were cropped to a 3-inch by 3-inch photo that included just the person’s head and shoulders. For each face, there was a “middle” photo that served as the baseline and comparison photo for each trial. There was also two target photos relative to the “middle” photo—an earlier and a later picture—such that there was a total of three pictures for each face used in the experiment. Subjects would only ever see two of the three photos at any given time (always paired with the baseline photo). Of the 40 critical faces, 20 were Caucasian, and 20 were Asian, each with equal numbers of male and female faces. (See Appendix for full stimuli set) Earlier Middle (baseline) Later Figure 1. Example of picture set for Caucasian face(Brad Pitt) with earlier-middle-later photos. Page 2 The offline portion of the experiment attempted to replicate the Miles et al (2011) experiment. This part of the experiment also used three pictures (earlier, middle, and later) per face, but only consisted of two targets: Brad Pitt and Jet Li (different photos those used during the earlier experiment). Each of the six pictures was sized to be 10-inch by 10-inch and backed with cardboard. The test board consisted of a 30-inch by 30-inch piece of cardboard. The final component of the experiment was a computer-based background questionnaire used to determine handedness, bilingual-dominance (Dunn & Fox Tree, 2009), and general language background of each subject. c. Procedure 1. Online Task Before the start of the experiment, the subject was given verbal instructions as well as written instructions on the computer screen in English before proceeding to the practice block. At the beginning of each trial, a prompt question that read “Is the second picture EARLIER?” or “Is the second picture LATER?” would appear to indicate which pictures to respond to. Using a go no-go paradigm, subjects were instructed to only respond when the answer to the question was “yes”, and they would do so by pressing the center button of a button box marked only with yellow tape. The question was visible on the center of the screen for 2000msec, which was subsequently replaced by the middle (baseline) picture of one set of faces. After a 500msec delay, the second (target) picture appeared on the screen in one of four positions on the screen: (1) 1-inch above the center picture, (2) 1-inch below the center picture, (3) 2.5-inch to the left of the center picture, or (4) 2.5-inch to the right of the center picture. These four positions served to highlight the two axes: horizontal (left and right) and vertical (up and down). The second picture to appear could have been either the “earlier” or “later” photo of that set. Subjects were asked to respond as quickly as possible after the appearance of the second image by deciding if the second picture represented the person depicted in an “earlier” or “later” point in time. If their judgment of the second picture gave a “yes” answer to the prompt question, they were asked to press the marked button, which would record their response time (RT). Otherwise, following a 2000msec delay from the appearance of the target photo, the experiment would automatically proceed to the next trial and their answer would be deemed as a “no” response. For example, if the prompt question was “Is the second picture EARLIER?” and the second picture was indeed an earlier photo of the person, then the subject should respond by pressing the marked yellow button. The experiment, excluding the practice, was divided into four blocks. Each block consisted of 80 trials, one for each possible pair of photos in a set (“earlier” paired with middle and “later” paired with middle), presented in a pseudo-random order.