strata elite men-either lower-level officials, or businessmen with cadres as parents had only known each other for several months. Xiao Hong and his friends nonetheless mimicked playful feasting, which I believe popularly indexes long-term intimacy among working-class men. I did my best to keep up with the fast-moving banter, but was hindered from contributing by my incomplete understanding of Sichuan dialect and my cautious personality. Mostly I quietly observed Peng was part of this group. Since I stayed close to Xiao Hong, who had welcomed me into the group, I did not know Peng. Over several meetings, Peng began giving me a hard time about not opening up (fang bu kai) and being real friends (gemenr) with them. He asked why I said little, and why I replied so politely to questions. My propri ety and restraint was not the way real friends should behave, he seemed to be saying. He was implicitly accusing me of lacking genzhi Ich feast value-concepts operate doubly: as the means to pedagogically shame participants into drinking or otherwise harming themselves, and as the end of ritual ction itself, a state of affective communitas. Genzhi refers to one's character or dispo- sition. Such a person does not calculate what impact their words will have on others Rather, this sort of person speaks and acts directly from feeling. Because a genzhi per- son does not keep track of winning and losing with friends, he or she is apt to take losses for them. The person lacking genzhi is envisioned as unable to loosen up among friends and as unable or unwilling to engage in social interaction. Like other cultural values utilised ritually, genzhi indexes behaviour in sharp contrast to individually max ng market calculation, referencing emotional authenticity with consequences for one's health, finances, and image. A genzhi persons words issue directly from feelings; there is no careful self operating beneath a mask. One is therefore vulnerable to agen- tive contagion from others. Peng was accusing me of just such detached, self-contained caution, and tried sev- eral times to resocialise me. One night drinking beer at a stand on the Binjiang Road promenade he challenged me to take off my shirt; another night in a karaoke hall he dared me to mock a group of college students for their weak singing. Ha! he crowed when I refused, 'He cant make fun of them! Even if the ritual challenge, often a toast, is itself a calculated move in competitive play, one so challenged who does not respond nimbly can be outmanoeuvred into appearing to resist the other person's ethical'gift'of self-sacrifice One therefore shows selfish intent to hold on to one's sober, socially correct demeanour at the cost of group feeling, such as when bank employee Luo was outmanoeuvred by driver Yangs ritual toasts, which were presented as shows of respect. Luo found himself in the untenable position of fighting against cultural ideals of manly sacrifice for ones brothers. Drinking therefore signifies entering meaningful, committed social interac tion. Challenging someone who holds back is therefore, I argue, a game that involves masochistically, dramatically surrendering selfish control over one's agency. Analysing genzhi through the lens of agency, Pengs and driver Yangs ritual actions make more sense. My caution in both cases indicated a calculation and self- preservation that made me an unattractive ritual partner. By not committing violence e 2014 Australian Anthropological Societystrata elite men—either lower-level officials, or businessmen with cadres as parents— had only known each other for several months. Xiao Hong and his friends nonetheless mimicked playful feasting, which I believe popularly indexes long-term intimacy among working-class men. I did my best to keep up with the fast-moving banter, but was hindered from contributing by my incomplete understanding of Sichuan dialect and my cautious personality. Mostly I quietly observed. Peng was part of this group. Since I stayed close to Xiao Hong, who had welcomed me into the group, I did not know Peng. Over several meetings, Peng began giving me a hard time about not opening up (fang bu kai) and being real friends (gemenr) with them. He asked why I said little, and why I replied so politely to questions. My propri￾ety and restraint was not the way real friends should behave, he seemed to be saying. He was implicitly accusing me of lacking genzhi. Such feast value-concepts operate doubly: as the means to pedagogically shame participants into drinking or otherwise harming themselves, and as the end of ritual action itself, a state of affective communitas. Genzhi refers to one’s character or dispo￾sition. Such a person does not calculate what impact their words will have on others. Rather, this sort of person speaks and acts directly from feeling. Because a genzhi per￾son does not keep track of winning and losing with friends, he or she is apt to take losses for them. The person lacking genzhi is envisioned as unable to loosen up among friends and as unable or unwilling to engage in social interaction. Like other cultural values utilised ritually, genzhi indexes behaviour in sharp contrast to individually max￾imising market calculation, referencing emotional authenticity with consequences for one’s health, finances, and image. A genzhi person’s words issue directly from feelings; there is no careful self operating beneath a mask. One is therefore vulnerable to agen￾tive contagion from others. Peng was accusing me of just such detached, self-contained caution, and tried sev￾eral times to resocialise me. One night drinking beer at a stand on the Binjiang Road promenade he challenged me to take off my shirt; another night in a karaoke hall he dared me to mock a group of college students for their weak singing. ‘Ha!’ he crowed when I refused, ‘He can’t make fun of them!’ Even if the ritual challenge, often a toast, is itself a calculated move in competitive play, one so challenged who does not respond nimbly can be outmanoeuvred into appearing to resist the other person’s ethical ‘gift’ of self-sacrifice. One therefore shows a selfish intent to hold on to one’s sober, socially correct demeanour at the cost of group feeling, such as when bank employee Luo was outmanoeuvred by driver Yang’s ritual toasts, which were presented as ‘shows of respect’. Luo found himself in the untenable position of fighting against cultural ideals of manly sacrifice for one’s ‘brothers’. Drinking therefore signifies entering meaningful, committed social interac￾tion. Challenging someone who holds back is therefore, I argue, a game that involves masochistically, dramatically surrendering ‘selfish’ control over one’s agency. Analysing genzhi through the lens of agency, Peng’s and driver Yang’s ritual actions make more sense. My caution in both cases indicated a calculation and self￾preservation that made me an unattractive ritual partner. By not committing violence 366 © 2014 Australian Anthropological Society B. D. Harmon and B. Harmon