You may or may not be aware that I’ve been spending a fair bit of time on Twitter lately. This began as a work exercise – “what’s the point of this then?” – but has become something more as my enthusiasm has grown. And as my enthusiasm has grown my participation has grown.

This morning I realised I’d sent six posts to Twitter in an hour. Not many by some standards, but if you’re only following 20 people and one of them is me, it must seem like I’m absolutely caning it.

And that – together with this blog post on the fallacy that number of followers is a measurement of ‘influence’ – got me thinking about how we perceive audiences when writing online.

By “we” I mean “I” – I bet there’s some good research on this, but I’m just jotting thoughts down.

Twitter is like some other services I use – LiveJournal and Tumblr – in that the default mode is a flow of information, and the more people you are following the faster that flow is. With Twitter, because all content has a fixed length, it’s practical to follow a lot more people.

But how your activity on these services is perceived isn’t a function of how many people you follow, or how many people follow you. Perception of your activity is governed at the individual level by how many other people someone is following.

This means the same individual can appear laconic to one follower and loquacious to another. If my 6 tweets are 30% of someone’s twitterstream, that’s a fair chunk of their attention I’m trying to muscle in on. If they’re .3%, I’m barely noticed.

If I was posting this on my marketing blog, I’d say that this implies you’d be better off trying to get people who aren’t following many others to notice you than going for the ‘big birds’. But I’m not, so I’m more interested in how the “casual tweeter” processes this information.

And the answer, surely, is “very imperfectly”. It’s simply too hard to keep in mind the very different shares of attention you’re demanding and commanding.

So what I hypothesise happens is this: you assume on some level that the people you’re interacting with on Twitter (or in any decentred web space) are using the service in roughly the same way as you.

We project our own experience onto the ‘standard experience’ and use a service accordingly.

In other words, on Twitter people become more chatty in proportion to the amount of input they’re receiving. That’s mostly determined by how many people they’re following, but it’s also influenced by the number of @messages they get, and the application they’re using.

This is a completely testable hypothesis, and I’d be interested in seeing if it’s true or not.