This newsletter is published at techpolicy.substack.com
An excerpt from Edition 23 is reproduced below.
Of letting bycotts be boycotts, banned if you don’t, Q-A-going-going-gone?
Let bycotts be boycotts?
No, that’s not a typo, bycotts are a thing (see image)👇
And I guess you’ve figured out where I am going with this. We’ve seen this faux-trage before.
- Actor 0 says or does something.
- Group 1 supports it, Group 2 does not.
- Group 2 outrages, says what about this, what about that, what if the shoe was on the other foot. Group 1 mocks Group 2, and points out that the shoe, shoe-rack and the entire shoe-factory + supply chain is often on the other foot, if applicable. Opinion: In this case, it was.
- Now, Actor 0 is faced with a choice : stand ground or back-track. In this case, they chose to back-track. The reasons for doing so are important, but unfortunately don’t seem to matter as much in the midst of an outrage cycle
- Group 1 is now angry, calls-out Actor 0. Group 2 rejoices.
- Some parts of Group 2 find issues with Actor 0s reasons for back-tracking or assert that Actor 0 hasn’t corrected enough.
- Group 2 now either shifts goal-posts or expands campaign. Meanwhile, Group 1 outrages.
And on and on we go.
By now, I am assuming you’ve read enough about the Tanishq campaign already, so I won’t the recount the events. I do want to add that I am not downplaying the implications of a business conglomerate choosing to go back on an ad that it believed was portraying communal harmony. I’ve also stated in past editions, how problematic terms like ‘love-jihad’ are and fall into the realm of ‘dangerous speech.
@beastoftraal’s blog post covers similar instances with brands like Surf Excel and even Closeup which apparently tried to take the middle-ground (shoes on both feet?) – but got shoes flung at them (figuratively only, I hope) nevertheless. Lakshmi Lingam, a professor at TISS, writes for The Leaflet about this being a representation of the fact that we live in post-truth times.
I looked at the following on Twitter:
- Tweets tagged with Boycott Tanishq on 12th October. (~40K Tweets, ~18.5K accounts)
- Tweets tagged with Boycott Tanishq on 13th October. (~40K Tweets, ~22K accounts)
- Tweets tagged with Boycott Tanishq on 14th October. (~28K Tweets, ~16K accounts)
- Tweets tagged Sack on 19th October (~32K Tweets, ~6K accounts). Note: I have chosen not to disclose the employee’s name, so related hashtags including the name have been redacted from the subsequent analysis. Incidentally, the tweet that first named this employee used the ‘bycott’ variant of the hashtag.
First, let’s look at the associated hashtags to get a sense of what the conversations could have been about:
12th October – Boycott
13th October – Boycott
14th October – Boycott
19th October – Sack
Some quick observations:
-Another type of ‘jihad’ (advertising) seems to have come to the fore
-Hinduphobia, anti-Hindu, and secularism found mentions too.
-Not surprisingly, ‘love jihad’ features prominently. It wasn’t the only Islamophobic trope that was there though. Some overlap with another hashtag about terrorism in madrasa’s also occurred.
-Surf Excel, Bollywood, and its ‘druggies’, as well as Akshay Kumar (named directly, as well as through a soon-to-be-released movie – Laxmi bomb), seem to be in the mix as well.
Next, let’s look at the Tweet rates in these datasets
Some points to note:
-12th October tapers off starting 7 PM UTC (1230 AM IST on 13th). It could just be a function of time and not a function of interest.
-That tapering off likely happened on the 13th too, but it is not as visible since the API didn’t pull results beyond 8 PM UTC (130 AM IST on 14th) and the peaks were lower than those on the 12th (~120 per minute v/s ~260 per minute)
-On 14th, the results start coming in from 12 AM UTC (530 AM IST) and continue on an upward trend till around 11 AM UTC (430 PM IST).
-On the 19th, I pulled the data just as it was trending at no. 1. It seems to peak in the afternoon and then gradually tapers off.
These are just indicative of course.
User Creation dates
-It should be immediately evident that the proportion of activity from accounts created in the last 3 months-ish is significant than other trends we’ve analysed in previous editions. Even so, in the case of the 19th October Sack hashtag, accounts created in October do see a bump.
I also looked into the top 15 dates on which accounts were created.
12th October (0 dates from the preceding 7 or 15 days)
13th October (2 dates from the preceding 7 days. 3 from the last 14)
14th October (4 dates from the preceding 7 days)
19th October (6 dates from the preceding 7 days. 7 from the last 14)
None of this smoking gun level stuff, but as the days progress there seems to have been a growing trend of participation by recently created accounts.
I am going to doom-prophet again, if this keeps up, I suspect we’re not far away from states forcing platforms to enforce some form of real-world ID linking to create new accounts. See the multiple levels of coercion there? And see who’s at the bottom of that chain
Not just me, us.
Banned if you do, Banned if you don’t
Another big information ecosystem story over the past week has been the Hunter Biden – New York Post situation. Twitter and Facebook came under scrutiny for how they handled it.
Facebook – Opted for Reduce pending fact-check i.e. control its spread on the platform until it could be fact-checked.
Twitter – Blocked the URL.
On the face of it, Facebook’s response seems more reasonable and easier to digest. Except, we should know better than to take things at face value. As Evelyn Douek points out in this episode of the Lawfare episode – pretty much all we know about this policy of theirs is that they rely on ‘signals’.
From the post where they mention it:
In addition to clearer labels, we’re also working to take faster action to prevent misinformation from going viral, especially given that quality reporting and fact-checking takes time. In many countries, including in the US, if we have signals that a piece of content is false, we temporarily reduce its distribution pending review by a third-party fact-checker.
Twitter, on the other hand, invoked its policies on hacked materials and disclosing personally identifiable information to take action. It wasn’t clear until much later that the latter had been used to block the URLs since the NY Post articles included unredacted personal information.
They have since updated the hacked materials to state that only hacked materials being distributed by the hackers or those working with them will be blocked. Meanwhile, the NY Post URLs and Twitter account are still blocked and locked respectively, since the personal information has not been redacted. One of the tweets in the explanatory thread said this:
Go here for the complete edition.