Skip to main content
Category: Methods
Type: Combinatorial problem-structuring method
Origin: Fritz Zwicky, California Institute of Technology and Aerojet, 1940s–1969
Also known as: General Morphological Analysis (GMA); morphological box; Zwicky box; morphological field
Quick Answer — Morphological Analysis is a method that breaks a problem into independent dimensions, lists every combination, then drops the combinations that cannot coexist. Fritz Zwicky developed it at Caltech and Aerojet from the 1940s, and stated it most fully in Discovery, Invention, Research through the Morphological Approach (1969). The practical lesson is that empty cells matter as much as filled ones: a matrix that only records the ideas you already like is a mood board, not an analysis.

​
What is Morphological Analysis?

Morphological Analysis is a structured method for mapping the total set of combinations inside a multi-dimensional problem, then reducing that set to the combinations that can actually coexist.
The morphological method essentially is nothing more than an orderly way of looking at things.
Imagine packing a trip as if the choices were independent knobs: climate, duration, transport, lodging. “Hawaii + a weekend + driving” is a combination you can write down. It is not a trip you can take. That is the everyday cue. Unlike brainstorming, which harvests whatever the room shouts, Morphological Analysis first builds a complete field, then uses a pair-wise check to throw out contradictions. The covering instinct is closer to the MECE principle than to a sticky-note wall. It is also not the linguistic tool of the same English name: this page is about Zwicky’s combinatorial method, not splitting a sentence into morphemes.

​
Morphological Analysis in 3 Depths

  • Beginner: Name three to five dimensions of the problem and three or four options under each. Circle one combination you had not said aloud.
  • Practitioner: Run a pair-wise cross-consistency assessment. Keep only internally consistent configurations, and treat them as candidates rather than a ranking.
  • Advanced: The field is an inference model, not a catalog. Designate drivers, read off compatible outputs, and remember that covering the space is not the same as knowing which configuration is true.

​
Origin

Morphological Analysis began as a wartime engineering discipline, not as a creativity workshop. Fritz Zwicky (1898–1974) was a Swiss astrophysicist at the California Institute of Technology. In the 1940s, as Director of Research at Aerojet Engineering Corporation, he applied the method systematically to jet and rocket propulsion. In June 1947 he published “Morphology and Nomenclature of Jet Engines” in Aviation. After defining the class of chemically activated positive-jet plants, he arranged their parameters in a morphological box. The elementary matrices gave 4 × 4 × 3 × 3 × 2 × 2 = 576 basic plants. Some drawers, he noted, would be empty if the parameters that defined them contradicted one another. He stated the method as a general way of thinking in the Halley Lecture at Oxford on 12 May 1948, later printed as “Morphological Astronomy” in The Observatory. Morphology had meant the study of form, as in Goethe’s plants. Zwicky generalized it from shapes to the structural relations among phenomena, concepts, and ideas. He used the same habit on galaxy classification, the legal aspects of space travel, and a Palomar Schmidt sky survey. The book-length statement is Discovery, Invention, Research through the Morphological Approach (1969). Earlier, Morphology of Propulsive Power (1962) and the Pasadena symposium volume New Methods of Thought and Procedure (edited with Albert Wilson, 1967) carried the method to a wider methods audience.
Within the final and true world image everything is related to everything, and nothing can be discarded a priori as being unimportant.
The computer-aided form is later. In the 1990s, Tom Ritchey at the Swedish Defence Research Agency (FOI) built software support (MA/Casper, from 1995) and named the pair-wise reduction cross-consistency assessment (CCA). A typical field, he reports, can shrink by 90 or even 99 percent. A seven-parameter field with six conditions under each already holds 279,936 simple configurations, which is why the reduction is the work, not the listing. Ritchey calls the whole practice General Morphological Analysis (GMA), to keep it distinct from geology, botany, and linguistic morphology. The academic problem-structuring paper is Ritchey, “Problem Structuring using Computer-Aided Morphological Analysis,” Journal of the Operational Research Society 57(7) (2006): 792–801.

​
Key Points

Morphological Analysis earns its keep when the problem has several independent dimensions and you cannot yet write a causal model. Hypothesis-driven thinking bets on one answer and tests it. This method first shows you the set of answers that could exist.
1

Name parameters, not a pile of ideas

A parameter is a dimension that can take more than one condition: propulsion source, not “build a turbojet.” Three to seven parameters is a working range. If two columns always move together, they are one parameter wearing two hats. Write the question as a field you can fill, not as a slogan you already prefer.
2

Fill each column with mutually exclusive conditions

Under each parameter, list conditions that cannot be true at the same time: rail or air, not “travel.” Four to six conditions per column is enough for a first pass. If two cells can both be true in the same configuration, the column is not a true variable. This is MECE applied downward, one axis at a time.
3

Cross-check pairs; drop contradictions

Cross-consistency assessment compares every condition with every condition in the other columns. Ask only whether the pair can coexist, not which one causes the other. Logical contradictions (“open ocean” and “rapid police mobilization on foot”) go first. Empirical implausibilities go next, and should be labeled as judgments. A pretty matrix with no pair-wise cuts is still brainstorming.
4

Use the remaining field as a what-if laboratory

What remains is a solution space, not a winner. Designate one or two cells as drivers and read off the compatible rest. That is how a morphological field becomes an inference model. If a configuration still looks promising, prototype it with design thinking or hunt its inner contradiction with TRIZ. The box tells you what can exist. It does not tell you what to ship.

​
Applications

Use Morphological Analysis when the risk is a combination nobody named, not a single broken part you can isolate. Do not use it as a substitute for a vote, a forecast, or a root-cause finding.

Products and services still in pieces

Before a concept freeze, list the functions as parameters and the means as conditions. Walk the consistent configurations before sketching a favorite. Pair later with SCAMPER if you are remixing one existing object rather than rebuilding the field.

Policy, scenarios, and public systems

When the future will not sit still for a spreadsheet, build a scenario field and a strategy field, then test which strategies survive which worlds. The Swedish producer-responsibility study below is the named public case. A school board or clinic network can run a smaller version on staffing × site × hours × funding.

Career moves and early-career choices

Write industry, role, city, contract type, and company size as columns. Cross out the pairs you will not actually live: remote-only plus a lab that requires a bench, or a visa-tied role plus a two-month gap. The surviving cells are options, not a life plan.

Family and community choices that stall

A household renovation, a volunteer board, or a shared holiday can be a four-column box: budget, duration, who must be present, and non-negotiable constraints. Inconsistent cells end arguments that were never about taste. They were about two conditions that cannot share a weekend.

​
Case Study

The cleanest numbered window onto Morphological Analysis as a policy method is the Swedish study of extended producer responsibility (EPR) reported in April 2004. The Swedish Ministry of the Environment commissioned the Swedish Defence Research Agency (FOI). Maria Stenström and Tom Ritchey facilitated. EPR here means that firms which make, import, or sell products remain financially or physically responsible after use, instead of leaving the waste to government. The question was not “what is the one correct law?” It was how to hold a set of strategies against futures that could not yet be quantified. Two working groups of seven people each—one on the strategic environment, one on strategy—modeled with two morphologists. Participants came from the Swedish EPA, other authorities, two NGOs, and two private waste and recycling firms. The scenario field alone contained 20,736 formal configurations. That field was hyper-coherent: few pairs were internally impossible. Eight scenarios were chosen that together covered almost the whole space, from a raw-materials crisis to a “green paradise”. At a final joint meeting the two models were merged. One EPR strategy was selected as the best fit for each of the eight scenarios. Ritchey restated the case for an operations-research audience in the 2006 Journal of the Operational Research Society paper. The lesson is the method’s real bet: a wicked policy space can be turned into a laboratory of consistent configurations without pretending the future has a single number. Boundary note: the workshop structured options. It did not prove which future arrived, and it is not a causal claim about later Swedish waste law. Some morphological fields barely shrink; EPR’s scenario field was one of them. Reduction is a finding about the problem, not a score for the facilitator.

​
Boundaries and Failure Modes

Morphological Analysis fails when the columns are not independent. If “budget” is just a restatement of “duration × headcount,” the box multiplies the same fact and calls the product insight. It also fails when the work is to isolate a fault that must be stopped, or to estimate a rate. A crashed system and a fraud case need evidence along a causal chain, not a catalog of possible worlds. It fails, third, when you need a ranked decision this afternoon. Cross-consistency says which combinations can exist. It does not say which one the board should fund. Pair a later ranking tool if a vote is required. The common misuse is to fill a handsome matrix and then circle the cell you walked in wanting. Skipping the pair-wise cuts leaves you with combinatorial decoration. Another misuse is to treat 279,936 configurations as a virtue. Listing without reduction is how a method for perspective becomes a method for drowning.

​
Common Misconceptions

The English name collides with other fields, and the grid looks like a creativity game, so it is easy to copy the furniture and miss the reduction.
In natural-language processing, morphological analysis splits words into morphemes. In image processing, mathematical morphology erodes and dilates shapes. Zwicky’s method is neither. It is a combinatorial map of a problem’s dimensions. If your task is tokenizing a sentence, you are in the wrong article.
A seven-by-six field already holds hundreds of thousands of simple configurations. Ritchey’s point is that the work is to throw most of them out. A larger box with no cross-consistency assessment is a larger pile of untested conjunctions, not a deeper model.
Filling cells is the analysis phase. The synthesis is the pair-wise judgment of what can coexist, then the use of the reduced field as a what-if model. A table that never says no is brainstorming with extra lines. SCAMPER is the better tool if you only need seven prompts on an object you already have.
These pages sit next to the same problem: how to generate options without pretending that shouting, or a single favorite cell, has covered the space.

Brainstorming

Unstructured idea harvest; useful after the field exists, not as a substitute for filling it.

SCAMPER Method

Seven prompts for remixing one existing object rather than rebuilding a parameter field.

TRIZ

Resolve a contradiction inside a configuration; use Morphological Analysis when the issue is the field itself.

MECE Principle

No overlap, no gaps: the covering rule each column of a morphological box should obey.

Design Thinking

Prototype a configuration once the box has shown that it can exist in use.

Divergent Thinking

Generate many options; Morphological Analysis makes the generation exhaustive and then cuts it.

​
One-Line Takeaway

Name the dimensions first, then drop the combinations that cannot live together. The option you never wrote down is usually the one that blindsides you.