## Content This is a response to an oped in The Hindu on 20th January 2022 that attempts to point out problems associated with 'online anonymity' ([Link](https://www.thehindu.com/todays-paper/tp-opinion/the-many-problems-of-online-anonymity/article38294925.ece)) Note that my rebuttal is to arguments/suggestions made in the article. I do follow Karthik Srinivasan's work and often agree with his points. For the rest of the post, I will refer only to 'the author' so as to not personalise my arguments. I also recommend reading it first so that my opinion doesn't colour yours. **TL:DR:** <u>The role of anonymity in online spaces is a complex subject - there are advantages and disadvantages. But, given the present trajectory of spaces for dissent - probably not the area where we should be looking to turn policy knobs in.</u> I've organised this post into subheadings, based on my takeaways. Under each of them, I first paraphrase what I thought were the main arguments and follow that up with my responses, with quotes, if necessary. At the end, I've also included a collection of resources from Jillian C. York that add a lot of nuance to the anonymity conversation. I wouldn't blame you for skipping my response and going straight to [it](https://jilliancyork.com/2021/01/14/everything-old-is-new-part-2-why-online-anonymity-matters/). #### Reasons for anonymity 3 reasons are listed: 1. Being able to speak truth to power against vindictive governments. But governments may be able to trace whoever they want to given the resources at their disposal. 2. Participate without being judged for past experiences, non-heteronormative identities, or deeply personal experiences. 3. Preventing views of the online persona being linked/tagged to the person in the physical world. Let's start with point 3. > Yet another common reason for seeking online anonymity is to not let the views be tagged to the real person being spoken about, in the offline world. I don't think this is distinct from 1 and 2, in fact, I think it subsumes both 1 and 2. This is relevant because this is point from which the piece jumps off into listing the problems with anonymity. My take is that a separation of physical and digital persona's isn't the problem itself and is certainly not the place where the problems begin (see resources linked in [[#Related]]). Now, let's go to point 1. > The most famous reason for anonymity is to be able to speak the truth against vindictive governments. But no matter how someone tries, governments these days, with enormous resources, may be able to trace the person. First, in a world with shrinking space for dissent, preserving anonymity is vital. Even if we believe that with the vast amount of resources at their disposal, state authorities will *probably* find whomever they want to - there is still something to be said for raising the costs of doing so - for the sake of protecting the ability to dissent, which becomes even more important. There's another issue I have - if it is more or less a given that states may be able to trace whomever they want to - then anonymity already doesn't exist, no? *But I'll admit I may be nitpicking here.* #### Platform affordances 1. Platforms encourage anonymity in certain cases with examples of Glassdoor, Fishbowl and Reddit. Look, I don't have much sympathy for platforms. But I will say this, Facebook does have (for whatever it is worth) a 'real names' policy (enforcement is another matter, it is likely arbitrary). Facebook, as a platform is subject to the same set of problems listed below (and many many more) - and we won't find too many people making the argument that these problems exist at a smaller scale. #### Problems with anonymity I pieced together the following, across the article: 1. Not all anonymous handles are abusive but are seemingly responsible for '*most angry, abusive, abrasive, and obfuscatory conversations/replies*'. I am also including extreme opinions since that was referenced in paragraph 12. 2. Even if identity is known, there is limited recourse beyond 'judging' an opinion, or relying on other entities to take action. 3. Amplification of existing problems because people behave in 'undesirable ways' when physical world identities are dissociated from online personas - sharing misinformation and disinformation. 3. Disturbs balance in an 'online confrontation'. Anonymous party cannot be rationally or emotionally appealed to. Cannot be shamed. Allows evasion of judgement. Ok, let's start with 1. > And it’s true that not all anonymous handles tend to be abusive or hold extreme views. But it is equally true that the most angry, abusive, abrasive, and obfuscatory conversations/replies seem to come from anonymous handles. There are elements of seen and unseen factors here. Is there data to show that 'most angry, abusive, abrasive, and obfuscatory conversations/replies' indeed come from anonymous handles? And, if this is the case - we're obliged to ask additional questions: A. Are they driven by elite cues? Are there larger social problems that these are representative of? B. What is their broader impact? i.e. do they shape discourse in any meaningful way. Of note here - [a twitter study](https://blog.twitter.com/en_gb/topics/company/2020/combatting-online-racist-abuse-an-update-following-the-euros) in the aftermath of racist abuse of 3 England footballers (for missing penalties in the Euro 2020 finals v/s Italy) said: > We continued to remove violative content as it was posted on the platform in the days that followed. By 14th July, 1,961 Tweets had been removed proactively following the Final, with a total of 126 removed from reports. > > **99% of the accounts suspended were not anonymous** Said another way in same study: > our data suggests that ID verification would have been unlikely to prevent the abuse from happening - as the accounts we suspended themselves were not anonymous. **Of the permanently suspended accounts from the Tournament, 99% of account owners were identifiable.** Fine, this is just one example. And that's why I am saying we need to understand this a lot better than we do currently do, before we suggest bringing out legal instruments against anonymity (See [[#Related]] section below). Now, let's tackle 2. > And more importantly, even if someone gets to know the identity of the person who is being vile or abusive, they have absolutely no way of using that information in any meaningful manner beyond simply judging that person. They can perhaps tag the person’s employer or family members (if available/traceable). Even then, the tagged entities may decide to not do anything about it, and simply leave the opinion as it is, for it is that person’s ‘freedom of expression’. This pertains more to existing redressal mechanisms, as opposed to being specifically about anonymity since it starts with the assumption that an identity has been established. I'm generally in favour of looking at existing mechanisms first and figuring out how to improve them before bringing in new laws. But anonymity is not the root-cause here. Moving to 3 and 4. These are covered across multiple quotes: > The issue is not only about abuse or extreme opinions here but also of misinformation and disinformation. These are already massive problems. And anonymity, either by choice or enforced by platforms, gives the power to a person to evade judgment by public opinion. > Given the tendency of people to behave in undesirable ways when their real-world reputation is not affected by what they say online, the proliferation of both pseudonymous social media handles and platforms that encourage pseudonymous profiles may amplify already existing issues around online disinformation and fake news. > In an online confrontation, it’s almost as though one side has their eyes covered by a cloth and their hands tied to the back, while the other side has a bazooka in hand. You cannot rationally or emotionally appeal to a pseudonymous online entity. You cannot shame them into backtracking their disinformation. Again, I would go back to the questions A and B that I posed earlier. Some research/theorising does exist to show that elite cues still play a major role. [^1] [^2] [^3] *And to be fair, the author does not deny that anywhere*. But my sense is, that given 'performance' (See [[#Related]]) of majoritarian rhetoric is currently, pretty often, rewarded - and several people engage in hateful rhetoric even with the real identities linked to their handles (whether this is on open-feed/timeline-based platforms, or closed messaging spaces) - we may be past the evasion of judgement phase (it is, admittedly, hard to make this as a sweeping statement). Also, we have to factor in the role that identity (not identity as in IDs, but in the context of social, political, collective, cultural identities, etc.) plays in how we process information - arguments against a person so predisposed, wouldn't be too different (from the perspective of relying on emotional and rational appeals). Such appeals are unlikely to work. [^4] As someone who's been writing a [substack publication](https://techpolicy.substack.com) pretty much exclusively about the risks/dangers of misinformation and disinformation, and the health of India's information ecosystem - I will rarely make the argument that they aren't big problems. But it is my current belief that the ability to be anonymous is not the policy knob we should be looking to turn (see [[#Recommendations]] sub-heading). I would also recommend this [research paper](https://www.cambridge.org/core/journals/american-political-science-review/article/abs/psychology-of-online-political-hostility-a-comprehensive-crossnational-test-of-the-mismatch-hypothesis/C721597EEB77CC8F494710ED631916E4) that posits that individuals who engange in incivil/hostile interactions online, are likely to have those traits in the physical world as well. > Across eight studies, leveraging cross-national surveys and behavioral experiments (total *N* = 8,434), we test the mismatch hypothesis but only find evidence for limited selection effects. **Instead, hostile political discussions are the result of status-driven individuals who are drawn to politics and are equally hostile both online and offline. Finally, we offer initial evidence that online discussions feel more hostile, in part, because the behavior of such individuals is more visible online than offline.** Many of the problems attributed to anonymity, are in fact, social problems - often manifesting in non-anonymous spaces. #### Recommendations These are what I gathered as recommendations 1. Legal mandates, though it wasn't clear to me whether it applied to the people who shared false information, the platforms that allowed it, or both. > And anonymity, either by choice or enforced by platforms, gives the power to a person to evade judgment by public opinion. Only a legal mandate can hold them accountable for spreading lies, should the need arise. 2. And another reference > In an online confrontation, it’s almost as though one side has their eyes covered by a cloth and their hands tied to the back, while the other side has a bazooka in hand. You cannot rationally or emotionally appeal to a pseudonymous online entity. You cannot shame them into backtracking their disinformation. **You need to convince someone else (either in a social media platform company or in a law enforcement agency) to take action.** I would be very cautious about making any argument that sounds like it wants to mandate linking online personas to identity documents (as any talk of legal mandates are likely to lead to, whether we intended that or not). This is the bit about the article that worried me the most, to be honest. 'Undesirable behaviour' is very subjective. 'Mandating' action against it is fraught with risk of subjective/discretionary application. Look, we already have a number of ways to prosecute lies (I'm maintaining a [tracker](https://prateekwaghre.notion.site/Tracking-Arrests-Cases-FIRs-Threat-s-Detentions-related-to-posts-on-Social-Media-fbd46a35aad6490a8b2b57966b9020bb) of English-language news reports for arrests/FIRs/cases related to social media posts) - one theme that stands out that for social problems at such scale, state enforcement is going to be selective. We must factor in how the balance of power of between states, information infrastructure (platforms + news media, etc) or civil society will be affected. As a matter of principle, we should avoid taking paths that shift this balance of power away from civil society. ![[Balance of Power.png]] ###### Post-script *There is something to be said about sophisticated and even unsophisticated operations that attempt to astroturf public discourse that resort to mass creation of accounts. But the article doesn’t explicitly go there, so I won’t spend a lot of space on it, other than saying there are less blunt options than mandating id verification (gating feature access, etc.) * ## Related - On Anonymity: [Everything Old is New Again - Jillian C. York](https://jilliancyork.com/2021/01/14/everything-old-is-new-part-2-why-online-anonymity-matters/) : This is a fairly exhaustive resource for research and advocacy from the perspective of the right to anonymity / pseudoanonymity. - On Performance and the state of discourse: [[Prevalence, Performance and Politics]] ## Colophon %% title:: Response to "The many problems of online anonymity - The Hindu" type:: [[output]] tags:: url:: file:: creator:: Prateek Waghre %% created:: 2022-01-21 status:: [[brewing]] [^1]: [Mail-In Voter Fraud: Anatomy of a Disinformation Campaign](https://cyber.harvard.edu/publication/2020/Mail-in-Voter-Fraud-Disinformation-2020) [^2]: [The role of cable television news in amplifying Trump’s tweets about election integrity](https://firstdraftnews.org/long-form-article/cable-news-trumps-tweets/) [^3]: [Rasmus Kleis Nielsen on Twitter ](https://mobile.twitter.com/rasmus_kleis/status/1465355946851545088) [^4]: I wouldn't completely dismiss emotional appeals as this [article](https://www.bostonglobe.com/2022/01/04/magazine/what-reformed-extremists-taught-me-about-preventing-another-capitol-insurrection/) by Jessica Stern notes, but these appeals may need to come from people that already share social/parasocial relationships, instead of transactional social media interactions. > Many of the former terrorists in our sample attributed their capacity to disengage from these violent movements more to emotional factors than to intellectual ones: support from their families, finding a new sense of purpose through work (including counseling others trying to disengage), developing trusting relationships with prison personnel or probation officers, and strengthening other types of prosocial relationships