A systematic design method that breaks problems into independent dimensions, creates variant options, and surfaces three striking design combinations.
"What designs would we never propose unprompted?"
Fritz Zwicky's 1948 method for systematic design generation. Decompose the problem into 3-6 truly independent dimensions, generate 3-5 variants per dimension from conservative to provocative, then surface exactly three combinations from the combinatorial space — the most surprising, the most plausible, the uncomfortable-but-viable. Interesting designs are usually the ones nobody pitched.
- Makes the design space visible as a space, not as a list of options
- Conservative-to-provocative variants generate the surprise
- The hidden dimension that surfaces mid-session is often the gold
When to Use
When the team keeps designing by analogy to what they've already seen. When stakeholders disagree about what kind of design this is — that disagreement is usually a sign that the dimensions haven't been named yet. When the obvious answers feel too obvious and you need a deliberate way to step outside them.
How It Works
One role works in three passes:
- Dimensions: The Dimensionalist decomposes the design problem into 3-6 independent dimensions, naming each precisely and flagging any dependencies. Independence is the discipline — correlated dimensions produce a fake combinatorial space.
- Variants: Same role lists 3-5 variants per dimension, ranging from conservative (what most teams default to) to provocative (sounds wrong at first but might be the most interesting).
- Combinations: Same role curates exactly three combinations from the full space — most surprising, most plausible, uncomfortable-but-viable. Each one names its values per dimension and the trade-off it accepts.
What You Get
Three concrete design candidates (not thirty), what each rejected combination revealed about your assumptions, and often a "hidden dimension" — a choice axis that surfaced during the session but wasn't in the initial decomposition.
Related Methods
- Disney Creative Strategy — when you want to dream first and reality-check after, instead of mapping a combinatorial space
- Polarity Management — when dimensions turn out to be interdependent polarities rather than independent choices
- Six Thinking Hats — when perspective-based exploration fits better than design-space mapping
/libertee:morphological-box "your design problem"