In Defense of Live Tweeting at Talks and Conferences

Anyone who follows me on twitter knows that I often but not always live tweet talks.  There's a small debate on this brewing.  I'm behind on many things so this is a short note mostly in response to The Ethical Nag, but I'm taking time to post it because I think this is critical.

1. It is sometimes possible to tweet & listen to a talk at the same time.  In the old days we used to call writing during a talk "taking notes".  It was in fact recommended as a way to make sure you remembered the content.  I was forced to take notes against my will by some teachers at school – it's hard, it's a skill. But I learned that although I seldom go back and read them (though occasionally, and some of those occasions were very important), writing notes serves two very important purposes:
1. Taking notes keeps me paying attention even in a bad 1-hour lecture, because the task itself is hard and interesting.
2. What I write down I am much, much, much more likely to remember than what I don't, even if I never go back to read it.
Twitter allows me to go back and check my "notes" for a day or so, and serves the same functions for attention & mnemonics as note taking.  On the other hand, if I want the notes around long term, I don't use twitter, I write in a file, often a bibtex entry for the talk.  I've also learned that if I make todo lists at conferences I never get around to doing them when I come back, so I am sure to download papers and update my bibtex file for papers related to the talk, or great-sounding papers mentioned in the talk.  This takes priority over tweeting for me.  So the irony is, I tend neither to tweet talks I don't want to remember, nor talks where I am really learning a lot.  Rather, I tweet talks in the middle, where it is going slowly enough I can keep up, or where I have heard it before so I'm just acting like a tweep and sharing my favourite parts.

2. Tweets are more for the people at the meeting / talk than those not. What I like most about twitter at conferences is finding out what is happening in other sessions, or finding out what other smart people are thinking and feeling about the talk. Sometimes we crowd-construct questions during the talk. It's a positive externality that sometimes other tweeps also benefit. I get very frustrated if the fact I tweet well discourages others. I want to read multiple threads, multiple perspectives. I want to know if I missed an important point while downloading a paper, or if something I thought might be trivial was nonintuitive or even wrong to someone else. I want to engage with the speaker later even if I have completely forgotten the talk because the organisers scheduled 6 talks before the 10am coffee break. And I do. I get email from speakers days later. It's great.

There certainly are a lot of hazards to tweeting at live talks, so here's just a few pointers.

1. Never tweet about results presented as preliminary or secret! (duh)
2. Every tweet must always  identify the speaker & the conference so the idea is properly attributed if it gets retweeted.  I have had clever things I said retweeted 30 times attributed to the tweeter, not me, even though he had mentioned my name five tweets earlier.  Who would know?  Fortunately, this is usually doable with a hashtag or conference account & a last name.
3. Don't bore the pants off uninterested followers.  If you also tweet about breakfast, this isn't an issue.  But if you have an account that is usually high content per tweet, then this can be an issue.  There are two solutions. 
1. I have a secondary account where I tweet about breakfasts, politics & conferences, while I vaguely try to restrict my tweeting on my main account to high content.  I don't expect people to follow the secondary account; I expect them to find the tweets through hashtags.  If scientists do follow that account & not my primary one, I tweet at them to check they've noticed they do this.  (Some people actually prefer my blathering account because it's more snarky.)  If I find a particularly interesting paper, I tweet that on my main account just like I would if I'd run into it from another type of research activity.
2. Another option is to start each tweet with the conference account.  If you start a tweet with any @reference, only people that follow both you & that reference (or anyone else @mentioned in the tweet) will see the tweet.  This should be people who follow both you & the conference, that is, other attendees or wannabe attendees.  I got this innovation from Jaime Teevan.  This also covers half of point 2 above, so you only need to add the speaker's name after that.

So are there costs & concerns associated with tweeting at conferences?  Yes, just like every other possible human activity.  Your job as a scientist is to balance those costs with the benefits of communication, just like you do every single day with every other decision you make, including whether to do some experiments or read this blog post.  But I'm glad a lot of you choose to tweet at conferences and engage me when I do.  Thanks!  Because I learn a lot.