Skip to content

What is a Twitter network?

As I have learnt to use Twitter better, I have understood that one of the key things to keeping it useful is to follow the right person - for example, I like to follow various Malaysian politicians (e.g. @limkitsiang, @Khairykj, or @elizabethwong, amongst others) because when important political things are happening, they will be tweeting about it.

I installed TweetDeck recently, mainly because I found out it has a way of grouping the tweeters you follow - so now I have three groups: 'All Friends' (the default group), 'Academics', 'SoPo', and Facebook updates. Another thing that encouraged me to get TweetDeck was the realisation that much chatter amongst the academics probably happens while I'm asleep, due to timezones.

Anyway, this post is the second (the first was Social networks and commenting) that was sparked off by Would the real social network please stand up?

Thinking about Twitter, I agree with a commenter on the 'real social network...?' post (Adrian Chan) that a list of the people one follows would be more of a behavioural network: e.g. the groups I mentioned above are people I share interests with, but may never meet.

A Twitter network is also 'publicly articulated' in the sense that it's consciously expressed (i.e. through choosing people to follow) and people can see who you are following; in addition one can retweet in a name-dropping fashion, and people organising to meet up may display all the others who are in on the conversation. If you were to ask a tweeter who her 'personal network' was it would probably include some of those she follows, but her personal network would include many who are not also tweeters. If one were to trace all the 'followees' (i.e. those who are being followed) a group of people have, one could probably infer as to personal and behavioural networks - the personal would probably be denser. A 'Twitter network' is apparently multivalent, and seems to support the argument that networks depend what you're trying to measure, and how you go about doing it.
Facebook social network analysis visualisation
A Facebook publicly articulated network
Bernie Hogan


Some recent discussion amongst Malaysian bloggers about a soon-to-be-launched Malaysian Twitter monetisation service, Churp Churp (it is run by Nuffnang) makes me wonder about how their responses could relate to the different types of social networks.

The discussions has tended to centre around the inherent property of tweeting, that the tweet comes to you directly, whether you want to see it or not - as opposed to a blog, where you can choose to not read an advertorial (in fact, this is rather 'old media' in a way - like television; which must make it attractive to advertisers). So, Colin Charles (aka @bytebot) recommends that tweeters do not use the service, asking "do you want to alienate your followers?"; ShaolinTiger (@ShaolinTiger) argues that followers should be able to opt out of the sponsored tweets, but not have to unfollow the tweeter; David Lian (@davidlian) asks "Can you purchase conversation?" and argues that advertisers need to become part of the conversation, rather than pushing a message out through paid tweeters.

The symbolic aspect of tweeting, the exchange of pleasantries and informational titbits, is important to consider. Jeremy Woolf in Hong Kong makes an similar point to David Lian in talking about "gunners" who are paid to "seed" forums and the like - the process is like this:
"You identify a forum like Uwants or DiscussHk as an influential channel where discussions relevant to your brand, product or service are taking place. People care enough (or, at least are passionate enough) to share their feelings and ask probing questions. Instead of joining the conversation in a meaningful way by replying to posts or establishing a contributing and helpful role within the community, you instead hire a gunner to spam inappropriate comments at this influential audience." (Dear spammers – can we have our social media back?)

One of key difficulties of social network analysis is understanding the relative meaning of the different ties, and the classification of different types of networks helps in some measure to address this. To become part of a social network means that others need to derive positive meaning from associating with you; that meaning will derive both from their personal reaction, and the interrelated association with commonly valued practices. For the examples of social media, one needs to display commitment, relevance, and integrity. The latter does not preclude being a paid operator with vested interests, but only as long as disclosure is made; the motivation for participation is a central marker of authenticity and integrity.

With this in mind, we may speculate that tweeters may unfollow, or continue to follow, someone for reasons associated with the different types of networks. For behavioural social networks, if the content of the tweets starts being irrelevant to the initial interest, the tweeter will probably end up being unfollowed. So, if a politician starts telling everyone to buy cheap air tickets, he will probably be unfollowed. However, including a certain amount of personal, non-political tweets, is a good way to show commitment to the casual, conversational ethos of Twitter.

For the publicly articulated social networks (and assuming followees will realise when they've been unfollowed), there may be a delicate balance to be negotiated. For a personal friend with whom one has stronger ties, it would probably be easier to unfollow them (and explain via another channel why) than to unfollow someone with whom one has weaker ties, but not weak enough to not care about their reaction altogether.

Conclusions?
OK I've rambled on, and thanks for getting this far. What I'm trying to say is that by understanding what objections people have to sponsored tweets, we may understand more about why people tweet in the first place. It's related to:
1. Both building and maintaining personal networks that operate on meanings developed through relational practices. Social and cultural capital are generated here.
--> For example: I keep in contact with some offline friends, and we develop a mutual understanding about how much to tweet, and what kind of stuff to tweet about. This strengthens our social ties (social capital), and we learn more about each other's preferences and ideas (cultural capital).
2. Developing more functional informational networks directed at increasing cultural literacy and capital.
--> For example: I start to follow various academics in order to have an idea of what they're doing and talking about; I learn new buzz words, read recent articles they tweet about, and so on.

Coming up soon - how do blog networks and Twitter networks differ, and what are the consequences for monetisation strategies?

Social networks and commenting

A recent post by danah boyd (and Bernie Hogan) called Would the real social network please stand up? makes some interesting points about the dangers of assuming all social networks are comparable and concludes
"The truth of the matter is that there is no "real" social network. It all depends on what you're trying to measure, what you're trying to do with those measurements."

She outlines three types of networks:
• "Sociological 'personal' networks": measured in different ways, these would be 'ego networks' with the person in the centre choosing to associate themselves with all the others - e.g. by saying they are people who they would trust with a secret.
• "Behavioral social networks": these would be networks based on common practices. They may be observed but not experienced as important by the people involved, e.g. people taking the same train to work, or they may be more important to the persons - e.g. Grateful Dead fans.
• "Publicly articulated social networks": 'articulated' means that you consciously list them somehow (e.g. your list for Christmas cards), and the public part comes about when you tell others - the obvious example being Facebook 'friends', or blogrolls. These networks may be made of all kinds of people, some of whom may not reciprocate the social tie (e.g. think of an inveterate name-dropper); and these networks may serve different purposes. The symbolism of the ties are important here - i.e. I may friend you as a follow-up to an offline meeting, but there may be no real intention of deepening the relationship.

What are blog networks?
I was thinking - how would you classify blog networks? The common identification of the 'blogosphere' seems to be a behavioural network - from outside bloggers are often bunched together as one group, but from within most bloggers do not identify with the group as a whole. One publicly articulated social network is the blogroll - but there are different views on how useful they are in explaining meaningful ties for bloggers (e.g. Schmidt 2007). Leaving a comment, and responding to them, is a practice central to establishing and maintaining a 'publicly articulated social network' in blogging, but of course not every comment has the same meaning (e.g. see my 10 types of commenters).

I'd agree with Bernie Hogan (aka blurky) that it's important not to 'reify' networks, even though they can be visualised in compelling ways. Blogrolls have limited usefulness, but I would argue that mapping the comments reveals more meaningful relations. Here is an interesting example: at bit less than two years ago, bloggers who had clustered around the launch of All-Blogs, met up in 'Blog House' (Bloggers Allied). This is a mapping of the comments made on blog posts that discussed the blogmeet (the blue squares are blogs that received comments, and the red circles are people that made comments).
malaysia all-blogs blog house sna social network analysis

OK - it's all a bit confusing, but we can focus in on the two blogs with the most comments - who, not coincidentally, were the two major figures there.
malaysia all-blogs blog house sna social network analysis

With hindsight, it's interesting to notice how the two major figures there had little common online commenters - suggesting that their networks have different bases. These two leading bloggers were ostensibly working together towards a common purpose, but after the March 8, 2008 elections there was what was touted as a 'split in the blogosphere' - where they both had a public spat. When I asked some of those involved about the 'split', a common answer I got was: 'there never was a blogosphere - there are all kinds of bloggers, and they can do whatever they want'.

When analysing the social dynamics in the blogging field, it would be useful to think of different types of networks that are enrolled in different contexts: in practice, the networks only exist ephemerally, at the moment of their articulation - the danger of 'reifying' them comes from the ability to trace them on the web, which gives them a misleading permanency. A blogroll link may have been added two years ago, a comment may have been made pretty much at random in any blog.

I think that comments are a fundamental practice of bloggers, and investigating those is more important than - for example - looking at blogrolls or other links; though of course they are relevant too. Too many studies of blogs overlook comments, possibly because: 1) they are more difficult to crawl/mine with automated bots; 2) there is a decreasing rate of significance of comments as they increase in numbers (apart from being an index of the importance of the blog and/or the post) - studies that concentrate on the biggest blogs may therefore overlook them. The way I see it, a blog without the option of commenting is just a website, and analysing blogs without taking account of the comments is like trying to understand the social dynamics of a pub without paying attention to the pub goers.
tweetbackcheck