real world columns: labels

I moved house recently. We did it ourselves and the operation brought many important questions up about the way we live. Are so many possessions really necessary? Do we need to store so many things that we never use? Should we just have possessions that are useful, or maybe just have things that are beautiful? And if we embark on a big throwing out campaign aren’t we just adding to the global rubbish heap? Another question that came up, as a user interface designer moving house, was: what is the best way to label the boxes? Initially I started labeling them with a categorization of the contents: ‘books’, ‘cutlery’ etc but the contents varied so much that this was difficult. Then I thought of labeling them with the area that the contents had come from: ‘mantelpiece’, ‘kitchen draw’. I added meta-information such as ‘fragile’ and ‘heavy’. Fragile is a good warning to add to a label, but heavy? That would be obvious to the person carrying the box, did I need to add it to the label? Wendelynne suggested labeling the boxes with the destination of the contents, this would avoid huge build ups of boxes at the other end, things could immediately be shunted into the correct places. Problem here was that we then spent ages with each box deciding where the contents should go before we could write the label.

So, a label is about what’s in the box, but the question is what’s in a label? Well, let’s start with the idea that we attach labels to items to help identify them. But wait, I carry my documents around in a cardboard folder in my bag, it’s an old folder and bears all sorts of crossed out labels from days gone by. The actual contents are current documents but the label says ‘furniture ideas’. This doesn’t matter because I never have to look at the label, I only have one cardboard file so I know that it’s current documents.

Right, where does that leave us in our definition? Labeling is done when things are in a context of many other things of the same type. Now consider my collection of postcards of famous buildings, they are all the same type and there are plenty of them, but I don’t label them, that would be daft because I can see exactly which is which! So, the label is used to convey information about the item that may not be apparent from the perception of the item itself. All right, but when I go to the wood shop they have racks of wood there in different sizes, and I can see it’s different sizes and I can measure the sizes but they still put labels on giving the sizes. So sometimes the label is used to convey information that is apparent but that costs time and energy to extract.

These problems with labeling things have parallels in the digital world of the computer. ‘Which box did we pack the can opener in?’ is the same as ‘which file / directory is the beginning of that article that I was writing in?’ Labels are vital, but so often they are insufficient. How often have I had to list the contents of a file to see what was really in it because the title no longer jogged my memory? Maybe I should spend longer choosing a good filename each time I create a file. Here we come up against the problem with all labeling systems, be they in the real world or in the digital world; they all require a certain amount of user effort to make the label. Not just physical (typing it in, finding pen and sticky labels) but also mental (what is the best way of describing it? How can I sum it all up in eight characters or less?). Users will always end up doing only the minimum amount of this work necessary to make a label. The Macintosh offers users many labeling techniques including assigning a comment field and a color code to files, but I know few people who actually take the time and trouble to make use of these facilities.

These label creation problems occur in both the real and the digital world, but the problem is exacerbated in the digital world because of an important difference; the real world is richer in detail than the digital, and people use and rely on this richness of detail.

When I’m looking for the can-opener I can’t remember what I wrote on the label, but I can remember that it was packed in a long white and blue banana box. I can use the extra detail as a sort of inherent labeling system. On my desk I can find the document about making tables in Web pages easily, not because it has a meaningful label on it, but because it is scruffy looking and has got a big coffee stain on it. I didn’t purposefully spend time scruffing it up and pouring coffee on it, it just happened naturally and I can make use of it.

This then is the challenge for the digital world; to create an environment of rich enough detail so that users can bring over the tricks and methods that they use in the rich environment of the real world.