Recombination

Recombination models explore how new candidates can arise from arrangements of existing ideas or components. The search space can grow rapidly, but its growth rate depends on the allowed combinations, and useful innovation requires selection and development.

Recombination generates candidates by bringing existing components into a new arrangement. Its combinatorial potential depends on the allowed operation: choosing pairs from n distinct components gives n(n−1)/2 possibilities, whereas allowing every subset gives 2ⁿ subsets. Neither count says how many combinations are compatible, useful, or genuinely distinct.

The decision bottleneck often moves from generating possibilities to selecting and developing them. Name the function each component contributes, the interface between components, and the constraint that the combination might satisfy. Domain knowledge matters because an attractive analogy can fail at that interface. A larger library is an opportunity for search, not a mathematical guarantee of innovation.

When to use it

When innovation is needed and 'thinking harder' isn't producing breakthroughs; when cognitive diversity would expand the library of recombinant components; when cross-domain learning would create novel combination possibilities; when understanding that innovation is COMBINATORIAL rather than creative would change the approach to generating new ideas.

How it can help

Expand relevant components, specify their contributions, inspect compatibility, and test promising combinations against the problem. Balance breadth with expertise and evaluation capacity.