Harvard referencing a tweet (a guide for students)

Undergraduates often ask “how do I cite a tweet in my work”? There is a very simple answer to this: treat it like a web page, and follow your citation system’s guidance for citing web pages. Students can also take this approach when citing material on other social networks too. A quick google search should reassure students that this advice is solid, and is common enough practice in the academic community. So this post perhaps seems redundant. Well it is and it isn’t.

Why write yet another post about this? Well obviously I now have the benefit of a copy and paste FAQ answer to send to students, so that’s good for me. More importantly there’s something lacking in most of the online guidance on citing tweets (and other material too). I’m going to step back from the question “how do you cite a tweet?” for a moment and ask the students “why are you citing tweets?”.

Why are you citing a tweet?

We can do a disservice to a student by giving them the textbook answer, the very simple formula that they can find on Google. If you’re a student reading this please stop and think about your intention in citing the tweet. What is it you’re actually doing?

There are, I think, three occasions when you might be looking to use material from Twitter in your work (or from Facebook, Instagram… let’s just agree that Twitter is a stand in now for current and future social media). Each of these three contexts leads us to a different answer on how to “cite” the tweet. So let’s look at that.

1. You are studying a data corpus made up of tweets.

OK, stop right there. If your primary work involves analysing tweets you probably aren’t actually going to cite them as such – not in the Harvard referencing sense. I mean, you might but you probably won’t. The problem of how to cite these tweets now becomes a matter of project design. There will be ethical considerations here: does your methodology require you to anonymise the data? Well then you ain’t gonna Harvard reference that, are you? Take a look at this open access journal article by Ruth Deller to see how she reports anonymised tweets within her work.

2. You are referencing a comment made on Twitter about something important.

Someone who you think of as an interesting scholar or commentator has said something on the Internet! It could be tempting to take that tweet and build it into a literature review. Is that actually the right thing to do? It could be, but look again at what they’ve said.

Have they quoted the salient point from an article and tweeted a link? Consider this tweet from academic Will Brooker (edit 2024-12-13: this Twitter account is deactivated):

Well this is depressing – women have to be 2.5 times more ‘productive’ than men to get funding http://www.theguardian.com/higher-education-network/blog/2014/apr/15/sexism-disadvantage-women-academics

Although he gives you a headline stat, you wouldn’t quote and cite this tweet. The more appropriate step to take is to go to the article and read it yourself, then quote from it and cite that, or keep following the trail until you find the most authoritative source of the data.

Have they made a comment that they expand upon elsewhere (and in more than 140 characters?). You should follow that up.

Are you 100% sure about the context in which their comment was made? Twitter is conversational, and conversation can take lots of interesting turns. Sarcasm, satire, jokes and retorts are all common and may need to be taken into account. Never forget that while some tweets are curated and published, others are merely snippets of chat that you’re eavesdropping on.

3. You actually are quoting something written on Twitter.

The tweet stands. It is a perfect and self-contained moment. It doesn’t relate to another article where the thought is developed fully. It is truly indicative of the thing that you need to evidence. And it’s not part of your primary research. OK, that’s cool. Go cite it.

How to cite the tweet

As I said at the top of the page: treat it as a webpage, and follow your citation system. In Harvard you’re going to use author and date when you cite and then put the full reference into your bibliography. It won’t need to be separated out into another separate bibliography, it will just be there with all the books and articles, sorted alphabetically by author name – and by author name I mean their name, not their Twitter handle.

So, for example if I were to write:

Website statistics such as those offered by Harte (2011) suggest that it is the relatively banal information that is most useful to hyperlocal audiences.

I would then put something like this in my bibliography:

Harte, D. (2011) My Bournville blog nearing its busiest day ever (3000 visits; usual 250). Combined power of ‘when is Tesco open’ and ‘when are the carols’ [Online]. Twitter. 24th December. Available: https://twitter.com/daveharte/status/150631576189931520 [Accessed 3/1/12].

Note here: Dave never published a blog post about this, this was the only moment when he shared that data, and it does make a nice point. So it is in the third category of tweets in my list above. Also note that were I to quote an organisational Twitter account, I’d use an organisational author – so @bbcnews would be BBC News.

Remember: consistency

There are lots of web pages out there telling you how to cite social media content, and that’s great. However, there are differences in the way authors and institutions apply the rules of Harvard referencing. The reference I’ve put above was generated for me by EndNote, and every webpage in my EndNote database will inherit the same format. If you’re writing these by hand there’s the danger that you’ll take the template for the reference from a website that doesn’t match the rest of your formatting. Just remember: you’re citing this as a webpage, so be consistent with the way in which you are recording all the other webpages in your bibliography.