Metaphorical conflict shapes social perception when spatial and ideological collide.

Kleiman et al. do some intriguing experiments. I give you their abstract first, which doesn't actually say how they did the experiments, and then give you some further description from their text. The abstract:

In the present article, we introduce the concept of metaphorical conflict—a conflict between the concrete and abstract aspects of a metaphor. We used the association between the concrete (spatial) and abstract (ideological) components of the political left-right metaphor to demonstrate that metaphorical conflict has marked implications for cognitive processing and social perception. Specifically, we showed that creating conflict between a spatial location and a metaphorically linked concept reduces perceived differences between the attitudes of partisans who are generally viewed as possessing fundamentally different worldviews (Democrats and Republicans). We further demonstrated that metaphorical conflict reduces perceived attitude differences by creating a mind-set in which categories are represented as possessing broader boundaries than when concepts are metaphorically compatible. These results suggest that metaphorical conflict shapes social perception by making members of distinct groups appear more similar than they are generally thought to be. These findings have important implications for research on conflict, embodied cognition, and social perception.

In the first experiment they asked subjects to categorize a series of pictures of Barack Obama and Mitt Romney. One group categorized the Romney pictures using their right hand (the P key)and Obama pictures with their left hand using the Q key - compatible with the right wing, left wing political metaphor. A second group was asked to identify Obama with their right hand and Romney with their left - in this case the physical action and the candidate's ideology were metaphorically incompatible. The interesting result was that:

...participants in the incompatible condition perceived the difference between the candidates’ ideologies as smaller than did participants in the compatible condition...Additionally, participants in the incompatible condition perceived the difference between the candidates’ stances on specific political issues as smaller than did participants in the compatible condition

A second experiment asked participants to estimate the ideology of the typical Democrat and Republican using a scale of 1 to 9 that was either compatible or incompatible with the metaphorical association linking spatial locations to political ideologies.

Participants assigned to the incompatible condition (n = 194) provided their response on a horizontally displayed scale with the values in the opposite sequence, that is, from 1 (extremely conservative) to 9 (extremely liberal). Note that this scale reversed the traditional spatial assignment and placed liberal views on the right and conservative views on the left, which metaphorically puts the physical location and ideology in conflict... consistent with predictions, participants who rated their perceptions on the incompatible scale perceived the typical Republican’s and typical Democrat’s attitudes as more similar than did participants who rated their perceptions on the compatible scale.

Two further control experiments were done.