[T]here are known knowns; there are things we know that we know.
There are known unknowns; that is to say there are things that, we now know we don’t know.
But there are also unknown unknowns – there are things we do not know, we don’t know.
Donald Rumsfeld, when United States Secretary of Defence
I realise that this quote has been mocked before but I have always found it be both clear and interesting, mainly because accepting that there are things that you don’t know that you don’t know is important. Because of the way our world works now, where most information is heavily filtered in one form or another, it is becoming more a world of unknowns unknowns (things that are so filtered that you didn’t even know that you could have known about them) then a world of known unknowns (things that you have yet to look into but know exist).
I have a student who is undertaking a project exploring ways of exposing the revision history of Wikipedia in a way that makes it immediately obvious if you’re reading something that is generally agreed upon or in massive dispute. The History and Discussion tabs in Wikipedia are, for most people, equivalent to unknown unknowns – not only do they not even realise what they are there for, they don’t think to look. This illustrates one of the most insidious forms of filter, one where the information is presented in a way that appears static and reliable, relying upon the mechanism that you use to give that impression.
How, for example, can a person inside the Chinese web search zone find pictures of Tank Man at Tiananmen Square, if all legitimate searches that might turn up anything to do with it have been altered? If no picture of Tiananmen shows protests or tanks, how do you even know to search for Tank Man? Even if you find a picture of a man standing there, in front of tanks, how do you then discover the meaning of the picture?
I was reminded of the impact of filtering while I was reading Metafilter the other day. One of the Front Page Posts (FPPs) dealt with the call to boycott a Fantasy writer/game article contributor who had advocated the use of rape in fantasy literature as an awesome way to make the story better (in a variety of ways). I started reading the article, because I assumed that I would take issue with this Fantasy author but wanted to read the whole story, and left the page up to see what sort of comments unfolded. Because this could take time, I ignored the page for about 30 minutes.
Then, when I reloaded later, the post had been deleted. Now, because of the way that Metafilter works, a deleted FPP still exists and can be located in the database, but it is no longer linked to the front page and can no longer be modified. So, suddenly, I had an island of effectively hidden and frozen information. Having read the contents, the comments so far, and the write-up, I was still quite interested to follow the story but the unfolding and contribution of other people in the comments thread, which is the greatest strength of Metafilter, was no longer going to happen.
Now there are many of these deleted FPPs in Metafilter, easily accessible if you search for them by number, but they are closed to comment. They are fragments of conversations, hanging in space, incomplete, cast in amber. You can see them but you can’t see the final comments that would have closed the debate, the petering out as the arguments faded, the additional links that would have been added to this shard of the data corpus by the 12,000 active account holders of Metafilter.
Now, of course, whenever you look at Metafilter, you’ll know that for every few stories that you see on the front page, there’s probably at least one deleted one. Whenever you look at Wikipedia’s illusion of a clean white page where everything looks like it’s just been printed, you may realise that this could hide hundreds or millions of updates and corrections behind the scenes.
How does this change your perception of the information that is contained in there?
While it is easy to point to traditional publishing, especially for text and reference books, and point out the elitist cabals and intellectual thuggery that permeated some of these avenues, we must accept that the printed book never changed once it had been printed. To change a printed book, you must excise, burn, overprint, paint, physically retrieve and then re-insert. There is no remote update. There is no way that an invisible war can be waged against the contents of your copy of uncorrected Biggles or that someone thousands of kilometres away can stop you from opening the pages of your history text that describe the Tiananmen Square protests.
We have always had filter bubbles but, at the same time, we had history and the ability to compare fixed and concrete entities with each other. Torn out pages left holes, holes gave us questions, unknowns were discovered. I try very hard to read across and out of my filter bubble, and I strongly encourage my students to do the same, but at the same time I have to remind both myself and them that we are doing what we can within an implicit filter bubble of known knowns and known unknowns.
By definition, even though I’m aware of the possible existence of things that have already been so well hidden from me that I will never find them in my life time, I have no idea where to look to find these unknown unknowns. Maybe that’s why I’m buying more books and magazines at the moment, reading so very widely across the written and the electronic, and trying to commit as much as possible to here?
Do we need to know what we don’t know? How will we achieve this? Is this just another twinge as we move towards a different way of managing information?
What will this post say tomorrow?