Argument Strength is in the Eye of the Beholder: Audience Effects in Persuasion

Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 1, Long Papers, pages 742–753, Valencia, Spain, April 3-7, 2017. c⃝2017 Association for Computational Linguistics Argument Strength is in the Eye of the Beholder: Audience Effects in Persuasion Stephanie Lukin, Pranav Anand, Marilyn Walker and Steve Whittaker Computer Science, Linguistics and Psychology Depts. University of California, Santa Cruz Santa Cruz, Ca. 95064 slukin,panand,mawalker,swhittak@ucsc.edu Abstract Americans spend about a third of their time online, with many participating in on- line conversations on social and political issues. We hypothesize that social media arguments on such issues may be more en- gaging and persuasive than traditional me- dia summaries, and that particular types of people may be more or less convinced by particular styles of argument, e.g. emo- tional arguments may resonate with some personalities while factual arguments res- onate with others. We report a set of exper- iments testing at large scale how audience variables interact with argument style to affect the persuasiveness of an argument, an under-researched topic within natural language processing. We show that be- lief change is affected by personality fac- tors, with conscientious, open and agree- able people being more convinced by emo- tional arguments. 1 Introduction Americans spend a third of their online time on social media, with many participating in online conversations about education, public policy, or other social and political issues. Our hypothe- sis is that online dialogs have important proper- ties that may make them a useful resource for ed- ucating the public about such issues. For exam- ple, user-generated content might be more engag- ing and persuasive than traditional media, due to the prevalence of emotional language, social affil- iation, conversational argument structure and au- dience involvement. Moreover, particular types of people may be more or less convinced by partic- ular styles of argument, e.g. emotional arguments may resonate with some personalities while fac- tual arguments resonate with others. Factual: Death Penalty Q1: I’m sure there have been more repeat murder- ers than innocent people put to death. As far as the cost goes, is that really an issue? Execution Room = $10,000. Stainless Steel Table = $2,000. Leather Straps = $200. Lethal Injection Chemicals = $5,000. Knowing this person will never possibly be able to kill again = PRICELESS R1: Actually the room, straps, and table are all multi- use. And the drugs only cost Texas $86.08 per ex- ecution as of 2002. Emotional: Death Penalty Q2: You mean, the perpetrator is convicted and the defender acquitted? Yes, that’s the rule and not the exception. Notice here how no-one ended up dead, or even particularly seriously injured. Addi- tionally the circumstances described are incredibly rare, that’s why it makes the news. R2: The defender shouldn’t even have been brought to trial in the first place. That doesn’t make it any bet- ter. Somebody breaks into your home and threat- ens your family with rape and murder, they de- serve serious injury at the very least. Table …

Content truncated to 500 words