Thursday, November 22, 2007

Searching & Finding

It doesn't make any sense to look for something that isn't hiding. Why do the things people search for need to be hidden? Do they search for things because they're hidden, or are things hidden because they're searching?
-- Miyuki Miyabe

This intriguing assertion is posited by a character in Miyabe's novel Brave Story.

Two things particularly interested me about the statement when I first read it: that it sounds right, but at the same time it feels wrong.

Agreed. Things are not lost until we try looking for them. If loose change falls out of my pocket into the sofa, but I don't notice, is it lost? I don't think so. I didn't notice its going, so it is not missed.

A quick check of Websters turns up nine definitions of lost, some of which apply to loose change, some which don't. And one which is contradictory: "2 a: no longer possessed b: no longer known". Well, it is no longer in my possession, but I barely knew it to start with, so there is not much "loss".

In fact, things are only lost once we are looking for them and they are not where we expect to find them. For example, when I come home I tend to deposit what I am carrying somewhere in the house: my watch, phone, and wallet. They belong in a drawer in the kitchen, but I often drop them on a bookshelf at the top of the stairs when passing. When I go to find them again, if they aren't in the drawer, I'll go look on the shelf. They can hardly be said to be "lost", since I know where they will be within a limited set of possibilities.

However, I tend to being both lazy and forgetful. So I often drop them somewhere else in my haste, and do not remember where. Then when I go to look for them, they are truly "lost" because they are outside the scope of where they ought or could be expected to be.

But the words used in the quote are "hiding" and "hidden", not "lost".

You could say the things I misplace are "hidden", since in looking I have difficulty finding them, even if I am practically staring right at them. On my desk in particular, they are hidden like a tree in a forest -- one of an innumerable collection of things that make up the clutter I call my work area.

So, yes, they are "hidden" even if they are not hidden from view. Because, like being lost, they are outside the realm of their expected locations and I have trouble finding them against an unfamiliar backdrop, even if that backdrop is something I deal with day after day. I have frequently had the experience of looking for a book on the bookshelf, not finding it on the shelf I expect, then searching all of the bookcases in the house to no avail. Then, on a second go round, I find it on a shelf I had already searched. I simply didn't know how to "see" it in its new and unexpected location.

So in a sense, things are not hidden until you search for them and fail to find them where expected. And it is also true you search for things because they are hidden, in that you can hardly be said to be searching when you look within the scope of expected locations.

To make this discussion real, think of it in terms of something we use every day: search.

  • On the simplest level, things (web pages, information) are "hidden" because there is just too much information on the web for us to know where it all is -- or even might be -- and search engines help us "find" that information. Is it "lost"? Often not. Because in many cases we do not know whether the information exists or not. We are searching hoping that something will show up. So it is "hidden" but not "lost". This is the simplest view of search: search as discovery.
  • Sometimes we are searching for things we know exist -- sites we have visited before or information we have been told to search for. In this case, we are looking but the information is neither hidden nor lost, since we know it is out there somewhere on the web and search engines help us find it -- as expected. This is the second view of search: search as locating. Just like the wallet and watch I expect to find on the bookshelf
  • Sometimes, whether we know it exists or not, we look for information using the wrong words. We might misspell a name ("Dwayne Allmann") or look for synonyms to the words in the content we are looking for ("fix" instead or "patch"). The consequence is a failure to find the item, in which case it is truly "lost". Internet search engines do a lot to try and save us from this dilemma; they recommend correct spellings and support stemming, synonyms, and fuzzy logic to broaden the results. However, even these techniques may not solve the problem and we must try again and again to define a search that matches our requirements. This is search as hunting.
  • Finally, even if you construct the proper search -- you look in all the appropriate places -- you may fail to "see" the item you want. Most internet searches produce hundreds or thousands of results. The search engines do their best to prioritize the results (called relevancy) so the most likely are at the top of the list. In other words, where you expect to find them. However, if the items you need are not at the top, you need to do a second search: searching through the search results. This can be extremely frustrating -- just like searching through your entire house for keys, wallets, glasses, or whatever -- because you cannot find the items you need. You do not recognize the title, the abbreviated description, or the location/URL as meaningful. At this point, the information is truly "hidden" from you because you cannot distinguish it from the forest of other results, just as I cannot find a book on my bookshelf if it is outside the bounds of where I expect to find it. This is search as loss. In fact, the desired results may never be found. (Many people give up before even looking at the 2nd or 3rd page of search results. The field of possibility is so vast it discourages exploration.)

So, given this situation, why is the quotation both right and wrong? It is the first sentence of Miyabe's quote that holds the key: It doesn't make any sense to look for something that isn't hiding.

There could well be an issue of translation here. Because, although "hidden" and "hiding" are different tenses of the same verb, they have significantly different connotations in English. "Hidden" is passive; it implies something that cannot be seen or found. "Hiding" is active; it implies the object is deliberating taking action to hide itself. And to say we aren't looking for something unless it is "hiding", would eliminate most inanimate objects from the equation.

It is true: it doesn't make sense to look for something that isn't hidden -- even if it is in plain view. And it explains the frustration and despair many users feel when they need to look for information, since they often have no clue how large the field of possibility is before they start. They assume it is hidden.

One of the stories I like to tell is the experience I had when interviewing consultants to determine how they looked for knowledge about previous projects. The overwhelming response was, in order:

  1. Ask someone in the office.
  2. Call someone they think might know.
  3. Send email to people they think may know someone who may know.
  4. If all else fails, look online.

These were experienced, tech-savvy engineers; they knew how to construct search queries; they had a good sense of what information should be available within the corporation. But assuming the information was hidden, their #1 preference was to look for a guide within their community of peers.

This tendency has been repeated time and again across the diverse audiences I have supported. Why do things people search for need to be hidden? Because they are searching. Because they are outside the realm of known possibility. Because their need exceeds the bounds of personal knowledge.

Part of the work of making things findable is bounding the field of possibility for the searcher. if you can make it clear that their search is bounded within a scope of likely candidates (rather than everything on the intranet, for example), you can encourage them to search earlier and have more faith in the results.

One of the keys within knowledge management, or the design of any information space, is establishing confidence in your audience that your structures form a clear and reliable scope of possibility for the classification of information you support. In other words, your systems are the shelves where they are likely to find their keys and wallet. This means turning searching into finding and the hidden into the found.

No comments: