Reflection #6 – [2/8] – Pratik Anand

Paper 1 : A computational approach to politeness with application to social factors

Paper 2 : Language from police body camera footage shows racial disparities in officer respect

Both the papers provide interesting ideas related to respect and politeness in communication. While the second paper deals with a very hot topic of unfair treatment of minority races by the police in real world, the first paper takes a more academic approach towards behavior in online space.
The authors in paper 1 develop a framework to measure politeness in online conversation which is a very valuable resource. It has words with their politness scores. They painstakingly comb through interactions of editors in Wikipedia and annotate it as polite or impolite conversation. I really liked the fact that they acknowledged that these filtering might be biased as politeness is subjective. So, they took extra measure to choose only those words which are unanimously marked by all the annotators. Once completed, they used it to apply to StackExchange conversations and found that people who were polite earlier, became less polite once they gained a certain status or power after admin elections. It is a very fascinating result. Similarly, people who lost in those elections became more polite. What could be causing this ? Is this a general with power comes arrogance argument or something else ? Did those people start getting more involved and hence, got tired of being careful with words ? Also the paper doesn’t say about the people who got popular votes but didn’t apply for admin elections. Does their behavior change too ? The paper also mentions that people from certain part of US are more polite than the others. It brings an interesting point of culture. Some people are culturally polite in conversation by US and western standards. Other people are not fluent in English vocabulary enough and could be more direct and hence, appear to be impolite. Also, sarcasm plays a big role in conversations.Yeah, right” is a less polite phrase made up of more polite ones. The paper doesn’t talk about such limitations in its study.

The second paper is short but provides a crucial insight on how race based biases can influence a conversation. It analyses the transcripts from video feed of police officers who stopped someone and use it to determine how much polite they were in the talk. The test subjects are not revealed the race or sex of the people being investigated or the police officer but their conclusion shows that Non-Caucasians are talked less politely by the police irrespective of the race of the police officer. The paper measure in form of Respect and Formality. Though, formality is maintained for white as well black people. But it is drastically less for black people. An interesting observation is from Fig 5 that formality goes down with more time passing. Respect has a minima but increases on both ends of time of interaction. This trend can be explained by the fact that greetings are given at the start and the end of a conversation. A further direction of this paper should be facial and body gesture of the police officer as well as the apprehended user.

Overall, both papers bring attention to change in politeness either due to power and influence or racial difference which opens up new fields of study.

Leave a Reply

Your email address will not be published.