Just a quick reflection on what I (often not thinking about it) do to increase chances of recognising patterns in a mess. Not scientific at all 🙂
I try to experience the field prior to the whole pattern-recognition exercise (or I get into pattern-recognising for the fields I’m familiar with). Knowledge (often tacit) of how things are/could be creates a bigger picture where new messy data have to fit – contrasting “prior” and “new” helps to see patterns.
I increase the mess by adding variety, more sources, more data. I guess once my brain can’t cope anymore with processing the volume it starts clustering things together – and those often turn into patterns.
I decompose to elements: I identify some basic elements/characteristics of the phenomenon and try to figure out more about them. Somehow having details worked out often brings them into a whole picture.
I search for metaphors in other fields. Especially in the times of being totally lost, I look beyond the field. Usually this means reading unrelated, but interesting books, talking to strange people, taking strange courses… I guess in this case my brain is still working on the original problem in the background, so it finds a way to translate it into whatever other strange field. Once I see the parallels I try to work them out, often finding missing links.
And I talk to people. Articulation of implicit bits and pieces mixed with unpredictability of someone else’s mind and fun of a conversation do wonders.
Archived version of this entry is available at http://blog.mathemagenic.com/2006/07/05.html#a1795; comments are here.