Some inconsistencies in the book: Is prejudice a normal cognitive process or an irrational ego-defensive maneuver? Can be both.

Are stereotypes the cause or the consequence of prejudice? Once again can be both, Stereotypes rationalize the emotional dispositions

“The Social Science Research Council then asked the Cornell University sociologist, Robin Williams Jr., to review the research on intergroup relations. Williams’s (1947) monograph, The Reduction of Intergroup Tensions, offers 102 testable “propositions” on intergroup relations that included the initial formulation of intergroup contact theory. Based on the scant research available, Williams (1947) stressed that intergroup contact would maximally. reduce prejudice when: (a) the two groups share similar status, interests, and tasks; (b) the situation fosters personal, intimate intergroup contact; © the participants do not fit the stereotyped conceptions of their groups; and (d) the activities cut across group lines. These general principles will be familiar to anyone versed in Gordon Allport’s framework.”

“From the student papers especially, Allport noted the contrasting effects of intergroup contact – often reducing but sometimes exacerbating prejudice. To account for these inconsistencies, Allport adopted a “positive factors” approach. Reduced prejudice will result, he held, when four positive features of the contact situation are present: (a) equal status between the groups, (b) common goals, © intergroup cooperation, and (d) the support of authorities, law, or custom.”

“What is critical is that both groups perceive equal status in the situation (Cohen, 1982; Riordan & Ruggiero, 1980;Allport’s Intergroup Contact Hypothesis 265 Robinson & Preston, 1976).”

In a robotic telepresence scenario this implies that both sides need to be hidden behind a robot. But what about the asymmetry in the starting conditions?

“research demonstrates that equal status in the situation is effective in promoting positive intergroup attitudes even when the groups initially differ in status (Patchen, 1982; Schofield & Eurich-Fulcer, 2001).”

“Effective contact usually involves an active effort toward a goal the groups share” (Military, athletics, work tasks etc).

In robotic telepresence that does mean a common game or task?

“Attainment of common goals should be an interdependent effort based on cooperation rather than competition.”

Does this take into account context, mediated goals, different value sets?

“Intergroup contact will also have more positive effects when it is backed by explicit support from authorities and social institutions.”

civil-rights legislation

Can it be that does conditions are in fact limiting and circumscribing intergroup contact? What happens when the contact is forced to take place within structures that are part of the cause of the conflict? Is this forced normalization?

“Allport held that his four conditions should be integrated and implemented together, rather than listing them as variables to be considered individually” So they considered only two: Intergroup friendship (equality and cooperation) and Structured programs for optimal contact (an organized intention to meet Allport's condition).

Have these studies considered long-term effects for thees contacts? Do we have an improvement or worsening of prejudice on a global level?

“with an ever-expanding list of necessary conditions, it becomes increasingly unlikely that any contact situations could meet these highly restrictive conditions (Pettigrew, 1986, 1998; Stephan, 1987)”

“intergroup contact typically leads to positive outcomes even when no intergroup friendships were reported and in the absence of Allport’s proposed conditions. Indeed, 95 percent of the 714 samples included in our meta-analysis reported that greater intergroup contact corresponds with lower intergroup prejudice; but only 10 percent of the contact measures involved intergroup friendship and only 19 per- cent of the samples reported contact under Allport’s conditions. In hisformulation, Allport held his optimal factors to be essential conditions for intergroup contact to diminish prejudice. But our results indicate that, while these factors are important, they are not necessary for achieving positive effects from intergroup contact. Instead, Allport’s conditions are better thought of as facilitating, rather than essential, conditions for positive contact outcomes to occur”

So it is known that these conditions aid in increasing prejudice, but they are not necessary. On the contrary it is not really known what are the negative factors that could even cause an increase in prejudice.

“In many ways, this stance reverses Allport’s approach. It starts with the prediction that intergroup contact will generally diminish prejudice, but the magnitude of this effect will depend on the presence or absence of a large array of facilitating factors – not just the four emphasized by Allport. In particular, this approach focuses special attention on those negative factors that can subvert contact’s typical reduction of prejudice”

• “participants who do not think intergroup contact is important show far less prejudice reduction than those who regard it as important” (Van Dick et al., 2004).
• “Emotions such as anxiety and threat are especially important negative factors in the link between contact and prejudice” (Blair, Park, & Bachelor, 2003; Stephan et al., 2002; Stephan & Stephan, 1992; Voci & Hewstone,2003; see also Stephan & Stephan’s ch. 26 in this volume)
• The type of intergroup matters too (difference race, sexual orientation, political opinion etc)

Relevant paper: https://www.academia.edu/1999282/An_Israeli-Palestinian_Dilemma_Contact_Encounters_the_Wall_and_The_Strange_Career_of_the_Contact_Hypothesis_working_

How about the uncanny valley of mediation?

“Allport warned that superficial contact between members of dif- ferent groups would, in fact, reinforce stereotypes, by failing to provide new information about each group,” Thus he wrote, “the casual contact has left matters worse than before” (Allport, 1954/1979, p. 264).

Allport warned: “whether or not the law of peaceful progression will hold seems to depend on the nature of the contact” (p. 262).

“ The decategorization model (Brewer & Miller, 1984) proposed minimizing the use of category labels altogether, and instead interacting on an individual basis. The recategorization model (e.g., Gaertner, Mann, Murrell, & Dovidio, 1989) suggested that intergroup contact could be maximally effective if perceivers rejected the use of “us” and “them” in favor of a more inclusive, superordinate “we” category. These two models can be seen as extensions to Allport’s notions of perceived similarity between groups and, to a lesser degree, equality of status. Another model (Hewstone & Brown, 1986), called categorization (or sometimes mutual intergroup differentiation) pointed out practical problems with personalized, as opposed to group-based, interactions, and instead promoted keeping group boundaries intact and salient during intergroup encounters. Nevertheless, despite their conceptual differences, all of the various models that followed Allport paid a tribute to his ideas in some fundamental way.”

Allport believed “contact must reach below the surface in order to be effective in altering prejudice” (p. 276). Incarnation/Physicality?

• Individual connection vs 'remembering' the group affiliation.

“Hewstone and Brown (1986) argued that, under decategorized contact, attitudes towards the outgroup as a whole would remain unchanged”

This brings a point about telepresence and avatars in general. How much should the group affiliation be kept and mentioned?

In a personal approach “they are likely to be subtyped, or cognitively processed as separate from the group as a whole, or treated as an individual with no connection to the overall group….Nevertheless, practically, for categories that are visually salient (e.g., race, gender), complete decategorization is unlikely to occur, thus providing some basis for the benefits of positive personalized interaction to generalize to attitudes to the group as a whole (Miller, 2002).”

“Although it is not necessary that categorysalience be maintained at all times (see van Oudenhoven, Grounewoud, & Hewstone, 1996), ideally it should occur before the outgroup individual is perceived as atypical of their group.” How would the robot show its group affiliation right away?

“Emphasizing categorization during contact, however, is not without its own dangers. Making categories salient risks exacerbating and reinforcing perceptions of group differences, which may result in anxiety, discomfort, and fear (Hewstone & Brown, 1986)”

A robot however is much less threatening, supposedly. Need to be careful of the uncanny affect causing anxiety

“A theoretical paradox had thus arisen: whereas interpersonal encounters were likely to be pleasant, they might fail to generalize without some salience of group membership; conversely, salient intergroup encounters could generalize to the whole outgroup, but might be undermined by the concomitant generation of intergroup anxiety (viz., “anxiety stemming from contact with outgroup members”; see Stephan & Stephan, 1985, p. 158; see also Smith & Mackie, ch. 22 this volume).

• cognitive subtyping - regarding an outgroup individual as 'exceptional'.
• Eliminating categories / prejudice - assimilation (my incarnation).

“The combined perceptions of disclosure (interpersonal) and typicality (intergroup) reduced subsequent bias toward new outgroup members, who themselves had no connection to the confederate other than shared group membership (Ensari and Miller)..Recent research shows cross-group friendship (i.e., intimate contact with a member of the outgroup) to be a particularly effective form of intergroup contact (Pettigrew, 1997). Moreover, even indirect contact (i.e., the mere knowledge that other ingroup members have friends in the outgroup; Wright, Aron, McLaughlin-Volpe, & Ropp, 1997), can lead to more positive intergroup attitudes. These findings indicate the importance of the intimacy of contact and its potential to personalize the outgroup member, but they also suggest that despite the importance of a personalized interaction, some level of category salience must also be maintained if positive attitudes are to generalize to other members of the outgroup category.”

In short, the interlocutor has to keep in mind they are conversing with an outgroup member, but they need to be able to develop a personal relationship. The robot has to be designed with some features that associate it with the group

“contact per se typically has a reliable and independent effect (e.g., bottom path of figure 17.1) on the reduction of prejudice.”

• Two perspectives: moderation (What happens when the contact works) and mediation (How it happens).

“Although affective factors are now considered to be particularly important (see Pettigrew, 1998), individuation and self-disclosure – factors that have both cognitive and affective elements have also been found to mediate the relationship between intergroup contact and improved outgroup attitudes (e.g., Turner, Hewstone, & Voci,2005).”

Optimal contact process: Three stages:

1. Decategorization / individuation: The contacted individual is separated from the group to which there are prejudice.
2. Categorization: After growing sympathy to the individual, its relation to the group is re-thought and creates a more positive outlook on the entire group.
3. Recatogrization: The outgroup is no longer considered 'out', but we are all seen in one common group.

“Pettigrew’s model suggests that mediators and moderators might work together to produce the optimal contact situation (see Brown & Hewstone, in press).

In a robot situation the mediating and moderating elements are the robot itself, its interface and nature of interaction - this also bears the question of what is the level autonomy of the robot? Can it act as a mediator? This would reduce the authenticity of the avatar.

Dovidio, Gaertner, and Kawakami (2003) name cognitive factors and affective factors that mediate the contact.

Learning general information about the outgroup does not necessarily help reduce prejudice, but learning specific information about the individual that separates them from the outgroup stereotypes helps.

Self-disclosure from the individual helps because it amplifies the individual complexity of every individual that is independent of the group stereotypes. Self-disclosure also has an affective value because it generates intimacy and trust.

”affective factors appeared to be much more influential. Specifically, Pettigrew and Tropp (2000) concluded that anxiety appears to be a more important link between contact and reduced prejudice than is increased knowledge of the outgroup..Intergroup anxiety is thought to stem from the anticipation of negative consequences during contact, such as embarrassment, rejection, discrimination, or misunderstanding, and may therefore be exacerbated by minimal prior contact with the outgroup and large status or numerical differences between the ingroup and the outgroup (Stephan & Stephan, 1985). Anxiety may also result from intergroup threat, either symbolic or realistic, adistinction that is prominent in Stephan and Stephan’s (2000) Integrated Threat Model. Symbolic threat is conceptualized as a threat to the value system, belief system, or worldview of the ingroup, whereas realistic threat is a threat to the political and economic power, or physical well-being, ofthe ingroup”

This relates to stranger fetishism and deep political emotions.

Of course cognitive and affective processes are linked.

“argument put forward by Wright et al. (1997), namely, that when one observes a fellow ingroup member interacting with an outgroup member, not only is there less immediate anxiety because the observer is not directly involved in the contact situation, but anxiety about future possible intergroup interactions will also be lessened by reducing the fear (affect) and negative expectations (cognition) that tend to cause anxiety prior to actual contact.”Or

Good justification for public space and joint conversations. And also robots that don't make you anxious.

Perspective taking of course leads to empathy, but how does one do that in a robot encounter? Has to be mediated/moderated some way? Or by showing first person data by the controller?

“a combination of positive contact and group salience during that contact resulted in the most positive evaluations of the outgroup. This methodological innovation can make a unique contribution to the future understanding of the contact hypothesis.”

Defines significant barriers to initiating contact:

1. Practicality: meetings are complicated to arrange, language barriers, status difference, physical barriers/distance.
2. Anxiety: anticipation of negative reactions leads to increased usage of stereotyping. During state of anxiety positive 'subtyping' is ignored.
3. Generalization (categorization): Tricky to achieve. It is still unclear how much group saliency is needed, and how to measure this. Group saliency levels can be explicitly determined in a robot telepresence scenario, using the design of the robot and of the deployment site .