Thursday, June 24, 2010

Puzzling out a solution


Most people do jigsaw puzzles the 'normal' way.

Not our four-year-old.

Most of us pour the pieces out, look at the lid of the box, sort the pieces into little piles (edges and distinct colours / patterns), look at the lid of the box, and assemble the puzzle -- constantly referencing the picture on the lid of the box.

Not Caelan.

He doesn't seem to believe in the value of an absolute image to aid the assembly process. He will start the puzzle with some concept of what he is aiming for (a world map, a forest scene, a truck), spread the pieces out, and then just start assembling them based around colour / pattern (from what we can tell). All this time, the lid of the box is lying idly who-knows-where.

He is also guided by shape. Once he has mastered a puzzle (that is, can assemble it competently picture-side up), one of his little tricks is to reassemble it upside-down -- and he doesn't do this by looking at the pictures on the obverse, but by the shape of the pieces. (This gets a little harder when the puzzles get up past 100 pieces.)

It is an intriguing process to watch, and he is generally able to assemble puzzles quite quickly (he has just started the same process again with a new 200-piece puzzle this morning). Where as most of us are essentially using a tightly (slavishly?) self-referenced replication process, he is using an interesting combination of creative and interpretive skills.

For him, if the visual cue is right, the next key is shape and fit. If the fit is wrong, then he hunts for another piece.

As a friend has expressed it (from within the world of strategic conversations), he appears to be using less a process of building from a set of specifications (though in the end, the whole thing can only go together one way), and more a process of building from within available parameters and intent (aiming for the creation of a tree or a whale or a car).

The process is fractal (thanks again, David), as Caelan seeks to solve the puzzle by turning it into a series of mini-puzzles (he will work on a section, find some level of resolution in it, and then begin work on another section, until eventually the completed sections butt into each other).

It will be interesting to watch how this way of dealing with puzzling situations is applied in other life circumstances where there is more than one way for a final solution to 'lock together' ...

No comments: