The Evolution of Likes as Social Gestures

I recently wrote about the emergence of “Likestreams” and how new tools and services are being created around our adoption of “likes” as a universal social gesture. Thanks to Svartling I came across this very interesting post titled “The like as interest and social gesture” written by Adrian Chan. It really made me give some more thought to “likes”, other similar gestures, and their implications.

Likes are not just the core social gesture on Facebook. They are a one-click sign of interest used on many kinds of social services. Likes are like social bookmarks — a simple expression of interest in a bit of social data. That is, a selection of one thing among many things, an expression simplified in order to communicate to an audience. Let’s break this down somewhat. For the Like isn’t a clear and direct expression of the user’s interest, or like. The reasons for this are several-fold. One gesture is not capable of capturing differences in degree. Clearly, when we like something, our like varies by degree A one-word gesture is not a linguistic statement. This limits the expression of interest. There will be ambiguity in the selection itself, owing to:

• The reason for liking is not provided;
• The kind of like is not provided;
• The purpose of sharing the like is not stated;
• The audience intended in sharing the like has some ambiguity (due to the medium);
• Any interest in soliciting conversation or commentary is ambivalent.

Adrian’s uncovers the many different contexts in which the like gesture is initiated. His post goes on to provide quite a bit more insight as you peel away the many layers of the like gesture and is worth reading. Since the Facebook like button can be placed on almost any online object it definitely can be a bit confusing to determine the context of the gesture depending on the location and provocation for the click.

It’s nice when the like gesture is an explicit action on a given service. If you’re a user of Cliqset you can now both create an aggregated likestream and view the ones from other users on their profile pages. Here’s mine where you can get a filtered view of only the items I’ve liked on Flickr, fast.fm, YouTube, and Twitter. I’m not sure of all the services supported here but it doesn’t seem to be importing my likes from Google Reader, Buzz or Facebook. Still it’s a great way to just get the data your friends have liked across multiple services. I just wish that I could get this in a single stream for all the users I follow. Better yet, sum up the same items that were liked by multiple people and prioritize those in the stream (you listening Darren? :))

Adrian talks about the ambiguity of the like and I also see many instances where that can be the case. In the case of Cliqset they are importing the explicit likes from a given service but there are also so many implicit gestures which could be construed as likes online as well. Generally speaking, every time I bookmark something on Delicious, or ReTweet an item, or Blip a song I am also most likely implicitly liking that item. In the case of Twitter, most users utilize the RT as a like gesture as opposed to clicking on the favorite button. I use Twitter Times to filter my news which is based on the tweets and retweets of links that were sent out by the people I follow. Implicit likes definitely throw a wrench into things and can complicate the process of determining likes.

Ok, this part of the post is going off the tracks and into another lane a bit but I’ve been focusing mostly on how the like could be used a gesture by my social graph to highlight content for me. I could see how certain aspects of “liking” could find their ways into other areas of our online lives. It’s a pretty strong datapoint that could become a universal and portable gesture then we could perhaps see them propagated based around the object they are in reference to.

For instance lets say I loved a song on last.fm that gesture of “liking” that object is limited to just that service and my social graph that I’ve built there. But that song may be available across many services and I may want my “like” to be shared to all of them. What if there was a way to associate an object to multiple services and propagate our likes to them. So what if after I loved the track on last.fm  associated objects could also be liked such as the same song on Pandora, a video for the song on YouTube, or the song on Ping.

So for the example with Ping it shows how our likes can be used as social recommendations for commerce sites. Another cool example for this would be if when I liked items over at GDGT if they then showed up on the Amazon product pages. It would be great to see recommendations from my social graph on product pages when I shopped online.

Ok, well that’s the latest brain dump on my current obsession with likes. I’d love to hear what others think about the evolution of likes.