Saturday, January 8, 2011

Answer: Finding a story... FAIL!

I knew this day would come, and it's finally arrived.  We've been stumped.  

Lots of people worked on this, and several regular readers (Mark, gnetiq, Fred, Ahniwa) all left good comments and leads.  But nothing's worked out.  

To tell you what I've been doing... I posted this problem on a few social answering sites (Yahoo Answers, Answerland) and on a listserv that's devoted to answering exactly this kind of problem--that is, identifying stories and books from the merest of clues.  As it turns out, there's an entire community of librarians that face this problem every day--someone walks into the library and asks for "Poor and Wease" -- you have to know that what they meant is "War and Peace."  

So I signed up for the Fiction_L listserv and posted my question there.  (Side note:  It's been really interesting to see the kinds of questions people ask librarians.  "Can you find me novels set in Montana?"  I recommend it as a source of ideas about how people think about their difficult search problems.)  

SUSPENDED!  But nothing has worked so far.  So I'm going to declare this question SUSPENDED pending additional information or sudden discovery.  If (when!) we finally figure this one out, I'll be sure to do an analysis of HOW we could have found it (assuming we could).  

At this point, though, if I were a librarian, I'd be suspicious of the recollection.  As we've discussed before, recall of movies / books / short stories is very error prone.  It's not that individuals just have weak memories, it turns out we all recode what we experience, and then those recollections evolve over time.  Memory for an incident (or book or ...) isn't fixed into memory like an etching into glass, but it's lightly written and subject to weathering--more like water colors on rice paper that's left in the sun and elements of subsequent experience.  

by Peter Pirolli
As my friend Peter Pirolli points out in his work on Information Foraging Theory, you work on a problem until the expected utility of the next action (more searching, in our case) exceeds the total expected value of the work.  That's when you give up--when it just doesn't seem as though more work is going to yield anything.  

Pete's work was informed by some great studies of hummingbird's feeding in Alpine environments, where they have to choose when to move to another patch of flowers.  They move when the next search isn't likely to be successful.  

That's the situation we're in at the moment:  this information hummingbird needs to move onto the next food source, the next item on the dinner menu.  

But we'll remember this, and come back if the flowers re-open.  

Keep searching!  


  1. I kept on searching. I'm wondering if the memory is a mashup of a few stories?

    Searching [fairy tale king castle cliff -neuschwanstein tax] gave me the The Brave Little Tax-Man

    I don't remember the search that took me to The Wicker House It features a king with a castle on a cliff the ends up falling into the ocean at the end.

    I also went down the Lady Godiva path from the taxes angle and a result from a previous search.
    The search for [lady godiva tax castle cliff] took me to

    Again parts of each story match parts of the memory but not just one story matches all.

  2. I also failed to find the right story. I have learned that the next move to follow in such a case is to go back to the client in question for more details. Which country or continent? Is it a king or a prince or rajah? Are there other details? Can your friend Josh tell us anything more?

  3. From the librarian stand point, none of us would offer take this question as asked and run with it - any vague recollection such as this would be run through a pretty thorough reference interview to clarify and bring out as many details as possible.

    The smallest details make a difference: what color was the book? how big? what year do you think you read it?

    I'm not saying this isn't a stumper, just that in a practical context, no librarian who spend a lot of time researching this question as asked; they'd want more details before starting their search.