-
Hostile Counterspeech Drives Users From Hate Subreddits
Authors:
Daniel Hickey,
Matheus Schmitz,
Daniel M. T. Fessler,
Paul E. Smaldino,
Kristina Lerman,
Goran Murić,
Keith Burghardt
Abstract:
Counterspeech -- speech that opposes hate speech -- has gained significant attention recently as a strategy to reduce hate on social media. While previous studies suggest that counterspeech can somewhat reduce hate speech, little is known about its effects on participation in online hate communities, nor which counterspeech tactics reduce harmful behavior. We begin to address these gaps by identif…
▽ More
Counterspeech -- speech that opposes hate speech -- has gained significant attention recently as a strategy to reduce hate on social media. While previous studies suggest that counterspeech can somewhat reduce hate speech, little is known about its effects on participation in online hate communities, nor which counterspeech tactics reduce harmful behavior. We begin to address these gaps by identifying 25 large hate communities ("subreddits") within Reddit and analyzing the effect of counterspeech on newcomers within these communities. We first construct a new public dataset of carefully annotated counterspeech and non-counterspeech comments within these subreddits. We use this dataset to train a state-of-the-art counterspeech detection model. Next, we use matching to evaluate the causal effects of hostile and non-hostile counterspeech on the engagement of newcomers in hate subreddits. We find that, while non-hostile counterspeech is ineffective at kee** users from fully disengaging from these hate subreddits, a single hostile counterspeech comment substantially reduces both future likelihood of engagement. While offering nuance to the understanding of counterspeech efficacy, these results a) leave unanswered the question of whether hostile counterspeech dissuades newcomers from participation in online hate writ large, or merely drives them into less-moderated and more extreme hate communities, and b) raises ethical considerations about hostile counterspeech, which is both comparatively common and might exacerbate rather than mitigate the net level of antagonism in society. These findings underscore the importance of future work to improve counterspeech tactics and minimize unintended harm.
△ Less
Submitted 28 May, 2024;
originally announced May 2024.
-
Auditing Elon Musk's Impact on Hate Speech and Bots
Authors:
Daniel Hickey,
Matheus Schmitz,
Daniel Fessler,
Paul Smaldino,
Goran Muric,
Keith Burghardt
Abstract:
On October 27th, 2022, Elon Musk purchased Twitter, becoming its new CEO and firing many top executives in the process. Musk listed fewer restrictions on content moderation and removal of spam bots among his goals for the platform. Given findings of prior research on moderation and hate speech in online communities, the promise of less strict content moderation poses the concern that hate will ris…
▽ More
On October 27th, 2022, Elon Musk purchased Twitter, becoming its new CEO and firing many top executives in the process. Musk listed fewer restrictions on content moderation and removal of spam bots among his goals for the platform. Given findings of prior research on moderation and hate speech in online communities, the promise of less strict content moderation poses the concern that hate will rise on Twitter. We examine the levels of hate speech and prevalence of bots before and after Musk's acquisition of the platform. We find that hate speech rose dramatically upon Musk purchasing Twitter and the prevalence of most types of bots increased, while the prevalence of astroturf bots decreased.
△ Less
Submitted 28 January, 2024; v1 submitted 8 April, 2023;
originally announced April 2023.
-
No Love Among Haters: Negative Interactions Reduce Hate Community Engagement
Authors:
Daniel Hickey,
Matheus Schmitz,
Daniel Fessler,
Paul Smaldino,
Goran Muric,
Keith Burghardt
Abstract:
While online hate groups pose significant risks to the health of online platforms and safety of marginalized groups, little is known about what causes users to become active in hate groups and the effect of social interactions on furthering their engagement. We address this gap by first develo** tools to find hate communities within Reddit, and then augment 11 subreddits extracted with 14 known…
▽ More
While online hate groups pose significant risks to the health of online platforms and safety of marginalized groups, little is known about what causes users to become active in hate groups and the effect of social interactions on furthering their engagement. We address this gap by first develo** tools to find hate communities within Reddit, and then augment 11 subreddits extracted with 14 known hateful subreddits (25 in total). Using causal inference methods, we evaluate the effect of replies on engagement in hateful subreddits by comparing users who receive replies to their first comment (the treatment) to equivalent control users who do not. We find users who receive replies are less likely to become engaged in hateful subreddits than users who do not, while the opposite effect is observed for a matched sample of similar-sized non-hateful subreddits. Using the Google Perspective API and VADER, we discover that hateful community first-repliers are more toxic, negative, and attack the posters more often than non-hateful first-repliers. In addition, we uncover a negative correlation between engagement and attacks or toxicity of first-repliers. We simulate the cumulative engagement of hateful and non-hateful subreddits under the contra-positive scenario of friendly first-replies, finding that attacks dramatically reduce engagement in hateful subreddits. These results counter-intuitively imply that, although under-moderated communities allow hate to fester, the resulting environment is such that direct social interaction does not encourage further participation, thus endogenously constraining the harmful role that these communities could play as recruitment venues for antisocial beliefs.
△ Less
Submitted 23 March, 2023;
originally announced March 2023.
-
Resilience by Structural Entrenchment: Dynamics of Single-Layer and Multiplex Networks Following Sudden Changes to Tie Costs
Authors:
Paul E. Smaldino,
Raissa D'Souza,
Zeev Maoz
Abstract:
We examine a model of network formation in single-layer and multiplex networks in which individuals have positive incentives for social ties, closed triangles, and spillover edges. In particular, we investigate the influence of shocks to the network in which the cost of social ties changes after an initial equilibrium. We highlight the emergence of structural entrenchment: the retention of structu…
▽ More
We examine a model of network formation in single-layer and multiplex networks in which individuals have positive incentives for social ties, closed triangles, and spillover edges. In particular, we investigate the influence of shocks to the network in which the cost of social ties changes after an initial equilibrium. We highlight the emergence of structural entrenchment: the retention of structural features, such as closed triangles and spillover edges, which are formed under historically different conditions from those currently driving network evolution. This work has broad implications for understanding path dependence in the structure and dynamics of single-layer and multiplex networks.
△ Less
Submitted 24 August, 2017; v1 submitted 23 October, 2016;
originally announced October 2016.
-
Adoption as a Social Marker: Innovation Diffusion with Outgroup Aversion
Authors:
Paul E. Smaldino,
Marco A. Janssen,
Vicken Hillis,
Jenna Bednar
Abstract:
Social identities are among the key factors driving behavior in complex societies. Signals of social identity are known to influence individual behaviors in the adoption of innovations. Yet the population-level consequences of identity signaling on the diffusion of innovations are largely unknown. Here we use both analytical and agent-based modeling to consider the spread of a beneficial innovatio…
▽ More
Social identities are among the key factors driving behavior in complex societies. Signals of social identity are known to influence individual behaviors in the adoption of innovations. Yet the population-level consequences of identity signaling on the diffusion of innovations are largely unknown. Here we use both analytical and agent-based modeling to consider the spread of a beneficial innovation in a structured population in which there exist two groups who are averse to being mistaken for each other. We investigate the dynamics of adoption and consider the role of structural factors such as demographic skew and communication scale on population-level outcomes. We find that outgroup aversion can lead to adoption being delayed or suppressed in one group, and that population-wide underadoption is common. Comparing the two models, we find that differential adoption can arise due to structural constraints on information flow even in the absence of intrinsic between-group differences in adoption rates. Further, we find that patterns of polarization in adoption at both local and global scales depend on the details of demographic organization and the scale of communication. This research has particular relevance to widely beneficial but identity-relevant products and behaviors, such as green technologies, where overall levels of adoption determine the positive benefits that accrue to society at large.
△ Less
Submitted 21 July, 2016; v1 submitted 16 July, 2015;
originally announced July 2015.
-
Does Learning Imply a Decrease in the Entropy of Behavior?
Authors:
Paul E. Smaldino
Abstract:
Shannon's information entropy measures of the uncertainty of an event's outcome. If learning about a system reflects a decrease in uncertainty, then a plausible intuition is that learning should be accompanied by a decrease in the entropy of the organism's actions and/or perceptual states. To address whether this intuition is valid, I examined an artificial organism -- a simple robot -- that learn…
▽ More
Shannon's information entropy measures of the uncertainty of an event's outcome. If learning about a system reflects a decrease in uncertainty, then a plausible intuition is that learning should be accompanied by a decrease in the entropy of the organism's actions and/or perceptual states. To address whether this intuition is valid, I examined an artificial organism -- a simple robot -- that learned to navigate in an arena and analyzed the entropy of the outcome variables action, state, and reward. Entropy did indeed decrease in the initial stages of learning, but two factors complicated the scenario: (1) the introduction of new options discovered during the learning process and (2) the shifting patterns of perceptual and environmental states resulting from changes to the robot's learned movement strategies. These factors lead to a subsequent increase in entropy as the agent learned. I end with a discussion of the utility of information-based characterizations of learning.
△ Less
Submitted 18 February, 2015; v1 submitted 18 January, 2015;
originally announced January 2015.
-
Social Conformity Despite Individual Preferences for Distinctiveness
Authors:
Paul E. Smaldino,
Joshua M. Epstein
Abstract:
We demonstrate that individual behaviors directed at the attainment of distinctiveness can in fact produce complete social conformity. We thus offer an unexpected generative mechanism for this central social phenomenon. Specifically, we establish that agents who have fixed needs to be distinct and adapt their positions to achieve distinctiveness goals, can nevertheless self-organize to a limiting…
▽ More
We demonstrate that individual behaviors directed at the attainment of distinctiveness can in fact produce complete social conformity. We thus offer an unexpected generative mechanism for this central social phenomenon. Specifically, we establish that agents who have fixed needs to be distinct and adapt their positions to achieve distinctiveness goals, can nevertheless self-organize to a limiting state of absolute conformity. This seemingly paradoxical result is deduced formally from a small number of natural assumptions, and is then explored at length computationally. Interesting departures from this conformity equilibrium are also possible, including divergence in positions. The effect of extremist minorities on these dynamics is discussed. A simple extension is then introduced, which allows the model to generate and maintain social diversity, including multimodal distinctiveness distributions. The paper contributes formal definitions, analytical deductions, and counterintuitive findings to the literature on individual distinctiveness and social conformity.
△ Less
Submitted 11 February, 2015; v1 submitted 29 July, 2014;
originally announced July 2014.
