What does it mean that graph traversal is a thinking technique?

Following connections through your knowledge graph generates new insights.

Example: A researcher studying climate policy starts at the node "carbon tax" in her knowledge graph. She follows a depth-first path: carbon tax leads to economic incentives, which leads to behavioral economics, which leads to loss aversion, which leads to framing effects, which leads to political messaging. Six hops from where she started, she has a novel argument about why carbon taxes fail — not because the economics are wrong, but because the policy is framed as a loss rather than a gain. She never would have reached this insight by thinking harder about carbon taxes. She reached it by traversing her graph — by following connections through territory she had already mapped but never walked in this particular sequence.

Try this: Choose a concept you are currently thinking about — a problem, a project, an idea. Write it in the center of a blank page or document. Now perform three different traversals. First, go deep: pick one connection from that concept and follow it as far as you can, writing each hop as you go. Do not stop until you reach something that surprises you or until you run out of connections (aim for at least six hops). Second, go broad: return to your starting concept and list every direct connection you can think of — every related idea, every adjacent concept, every association. Write at least ten. Third, go random: close your eyes, point to one of the broad connections at random, and perform another depth-first traversal from that node. Compare the three paths. The depth traversal likely produced a single unexpected insight. The breadth traversal showed you the shape of your local knowledge. The random traversal combined the structure of depth with the surprise of chance. Note which traversal generated the most genuinely new thinking.