Limitations Two limitations need to be highlighted. First, due to COVID-19 disruptions, we were only able to conduct limited interviews and this material could only give us suggestions for approaching online data collection and analysis. Some observations from the interviews informed our analysis of the online data, but this is not meant to be statistically representative. Second, cases on Facebook were selected via convenience sampling. As such, each of the three groups we aimed to focus on could not be equally represented, nor could we consider every potential member of these groups. For example, in the case of targeting of religious minorities, we were able to find examples of anti-Muslim speech on Facebook, given it is the most prominent form of hate-speech in Myanmar. But we were not able to analyze posts relating to Christian, Hindu, or any other religious groups. Similarly, as for ethnicity, the study was only able to look at case studies relating to three groups: anti-Rakhine, anti-Kachin and anti-Shan speech. On the study of anti-LGBTQ speech, we were able to find only anti-gay and anti-same sex marriage conversations among others. 3. Some structures of counter-speech on Facebook 3.1 – Links between anchoring posts and comment threads Observations In analyzing posts and comment threads, we could loosely divide the anchoring posts into three groups: those containing speech that was attacking or took a negative stance regarding the groups we focused the research on, which we referred to informally as negative; those containing speech which took a generally positive or inclusive stance regarding those groups, which we referred to as positive; and those about the marginalized group but containing neither clear negative or positive stances about the groups, which we referred to as neutral. In the posts we analyzed, there was a clear correspondence between the nature of the original post and the comments that followed it. Negative posts were followed by comments that seemed to reverberate the negativity of the original posts. On the other hand, there were examples of positive posts talking about interfaith relationship and harmony, where the comments reverberated the positive aspect of such harmony. Positive comments ranged from short messages, GIFs, and stickers to long written messages. Nevertheless, this is not a rule. For example, on a post by a religious page which blames gay marriage for the Australian wildfires, the majority of the comments were counter to the post’s absurd claim rather than echoing the general homophobia expressed in the post. Posts carrying the same theme of interfaith harmony also sometimes met overwhelming negative comments, depending on the account that originally posted them. This leads to a second observation: posts 7