Cross-cultural similarity in relationship-specific social touching

Abstract

Many species use touching for reinforcing social structures, and particularly, non-human primates use social grooming for managing their social networks. However, it is still unclear how social touch contributes to the maintenance and reinforcement of human social networks. Human studies in Western cultures suggest that the body locations where touch is allowed are associated with the strength of the emotional bond between the person touched and the toucher. However, it is unknown to what extent this relationship is culturally universal and generalizes to non-Western cultures. Here, we compared relationship-specific, bodily touch allowance maps across one Western ( N = 386, UK) and one East Asian ( N = 255, Japan) country. In both cultures, the strength of the emotional bond was linearly associated with permissible touch area. However, Western participants experienced social touching as more pleasurable than Asian participants. These results indicate a similarity of emotional bonding via social touch between East Asian and Western cultures.

Keywords: bonding; cultural differences; emotion; social touch.

Figure 1.

Relationship-specific TAMs in (a) Japan and (b) the UK. The colouring displays the proportion of the sample reporting that being touched by this person in this area is acceptable to them. (c) Comparison of the proportion of participants per culture who allow touching in different areas. Red colour in the maps indicates that Japanese participants reported that area more acceptable, blue colour indicates that British participants reported that area more acceptable. The data in (c) are thresholded at p < 0.05, FDR corrected. After FDR correction, Z threshold with no correlation assumptions varied from 3.22 to 5.98, depending on the number of participants who had that particular individual in their social network. Red and blue labels indicate female and male members of the social network, respectively.

Figure 2.

Correlations between touchability index (TI), emotional bond and pleasantness. Each dot represents the average response for one member of the social network in each culture (e.g. mother), with linear regression line and confidence interval for the regression fitted separately for each culture. Bottom right panel presents the relative importance of regressors in determining the TI in a linear model.

Figure 3.

Interaction plot of the average TI for male and female participants (blue and red dots, respectively) with respect to male and female touchers for each culture. Error bars depict the 95% confidence interval. Red and blue lines indicate the interaction between toucher sex and the sex of the participant. Note: partners are excluded from the analyses, as the sex of the partner differs between the male and female participants.

Figure 4.

Visualization of ROI-specific cultural differences in TI versus emotional bond. Linear regression lines were fitted to each culture separately. Blue and red colours depict UK and Japan, respectively. Each dot represents the average response for one member of the social network (e.g. ‘British partner’ or ‘Japanese mother’). The grey intercept and slope terms mean no significant difference in the regression lines between the cultures. The coloured terms signal which culture had higher intercept (α) or slope (β) values, with blue and red colours indicating the British and the Japanese, respectively. The visualization is presented with averaged data in the interest of clarity, but the significance of intercept and slope are calculated from the full (un-averaged) data.