Twitter’s New @Replies re-design isn’t just stupid; it’s really stupid.

This week I noticed something odd on my desktop Twitter:

I was surprised everyone wasn’t up in arms over it. Twitter changes inspire a lot of emotion.

Not until today did I realize I am part of a test for the feature when I read Matthew Panzarino review of the new feature at Tech Crunch

Twitter is testing a new design of the @reply that is intended to ‘clean up’ the conversation view by completely removing usernames from tweets. I’m in the test bucket (a group of users that Twitter is having take the features for a spin to see how they react) and it is a mess.

Panzarino also gives an example of what it looks like from the user side. Here’s a snapshot of my @replies right now:

When I try to respond to a tweet, I have no idea to whom I’m responding. What, pray tell, could be the problem with that?

Let me throw a few problems at you.

First, you should know how Twitter has learned that being known as a platform that facilitates harassment is bad for business:

Bloomberg is reporting that Disney chose not to pursue an acquisition of Twitter in part because it thought the bullying and behaviour of some of the ailing social network’s users might damage the entertainment company’s image.

Those of us who use Twitter and have identities that make us more vulnerable to harassment know this all too well.

Some of us have developed strategies to help us mediate that risk.

I’ll share two strategies I have used and how this change undermines them.

Selective Visibility

Sometimes I use Twitter to engage in conversations that, if viewable to my entire timeline, could open me up to abuse from non-followers. Those non-followers aren’t always passive trolls who respond to things that their followers tweet into their timeline. Increasingly, these are active trolls. They search terms to find victims. They target users. And, they leverage network ties to force context collapse.

Context collapse “ refers to the infinite audience possible online as opposed to the limited groups a person normally interacts with face to face.” It’s an academic term, coined by danah boyd. Jenny L Davis and Nathan Jurgenson introduce the idea of power to context collapse, calling it context collision. Context collision is when a party uses their greater power — more users, more time to stoke flames of trolling, their role in media — to force audiences to collapse.

To combat context collision, I rarely respond to every twitter account that tries to engage with me. Strange twitter accounts lacking institutional symbols of legitimacy — a user picture, a bio that sounds like it was written by a human, a bio that does not have racist key words in it, etc. — often try to engage potential targets. The troll only happens after a lure has been thrown to test a subject’s vulnerabilities.

A type of digital literacy one learns when they are not a western, middle class white male on social media is how to read for lures in the subtext. When an alarm sounds for me, I may selectively reply to a tweet by removing suspicious parties included in the reply.

Anything that makes it harder for me to identify the users in a tweet increases the risk that I might respond to a troll lure.

Another strategy for managing selective visibility involves well-meaning bystanders. Twitter can increase the likelihood of bystander effect when harassment takes place before so many people. Bystander effect is:

a social psychological phenomenon that refers to cases in which individuals do not offer any means of help to a victim when other people are present. The probability of help is inversely related to the number of bystanders.

In a version of the bystander intervention programs designed to counter sexual assault, some users on Twitter will selectively include or exclude me from potentially harassing twitter conversations as a form of ally-ship.

For example, some of my Twitter users have been known to respond to racist vitriol lobbed at me on Twitter. They see it on my timeline but choose to delete me in their replies. In effect, they redirect the harassment onto themselves as a means of diffusing its impact for me and other people of color.

When the @replies make it difficult to do this it undermines a strategy used to improve the quality of positive network ties on a platform that amplifies negative network ties.

It is a strange feature for Twitter to roll out given its recent issues with harassment.

So What: The User Experience

So what if I don’t like the new @reply format on Twitter?

So what indeed, unless you’re Twitter and you care about my user experience.

Today that user experience involved me not responding to a good-humored tweet about what kind of hood music I work out to.

I wanted to respond because the engagement would have been authentic for me.

I knew the person who tweeted the request but not the person in the quoted tweet.

When I opened a reply tweet, I could not figure out how to delete the original tweeter. And, the poor visibility of who was included in the tweet created doubt that I would get it right.

I manage my visibility as a black woman, an academic, and a public writer by negotiating context collapse. The risk for me is greater than it is for some other users. And, my experience of the platform relies, to a significant degree, on my ability to navigate these fault lines.

Without the security of knowing to whom I am replying, I cannot safely tweet.

And I tweet a lot.

Maybe that’s why it might matter to Twitter.