Liminal spaces

If you’ve been on certain parts of the internet, you may have come across the concept of “liminal spaces.”¹ ⊕ ¹ “Liminal” is one of those words that I was surprised to see escape containment from academia and become common online. It was VERY big in papers in the early 2000s.

“Liminal” basically means a transition point between two things, where it’s neither one nor the other, with the connotation of being unsettled in some way because of that.  “Liminal spaces” on the internet are usually pictures of spaces that are unsettling or unreal for some non-specific reason. They feel like having dropped into a videogame level that hasn’t fully loaded, or accidentally ending up backstage. 

One way to think of them is as spaces that don’t work as places. This is unusual in the physical world and makes us feel weird. For all that geographers have written about sense of place and all the ink architects etc. have spilled, I think it’s relatively hard to make a real physical space not read as a place. That’s why we notice it, it’s unusual, even in places that haven’t been meticulously designed. 

It’s much easier in digital experiences. Left to their own devices, digital experiences will feel sort of weird and liminal, like you’re wandering down a long hallway with doors that keep repeating. I would argue that most of the common conventions of information architecture have arisen because they’re strategies that help digital experiences feel more like places.

Not all organization imparts this sense of place. Data can be extremely well-modeled without being architected, in this sense, I’d argue.

If we look at our old friend, the MARC record. This is well-modeled data. I know they feel old and clunky in a lot of ways, but it’s really astonishing how durable they’ve been. Lots of people thought about this and it’s been carefully constructed to be a certain way. But, I don’t think it has any of the hallmarks of placemaking. If we compare it with a website, we see some of the same content in there, it’s representing a lot of the same ideas, but they’re represented differently. 

The biggest difference, I think, is that the web page version of this book is more relational. Both the MARC record and the website give you a clear (if you know how to read it) indication of what you’re looking at, what the identity of this thing is and some stuff about it. But the website is much more concerned with showing you where “here” is and telling you where you can go next. 

If we take the exercise to a silly conclusion and look at a Tweet about the same book, we see that in what we will call a web-app experience, a lot has dropped off. It’s no longer trying to represent a thing that exists outside of itself, it’s a lot simpler. But, it’s still telling you what’s here and where you can go from here, via its relationships with other things like it.  Data can be modeled without being architected. And by architected, I mean making information work more like a place.

References