Home / Synectics | Discovering Solutions Across Domains Through Analogy

Synectics | Discovering Solutions Across Domains Through Analogy

Synectics is a creativity method that encourages new ideas by connecting things that initially appear unrelated. It was developed by William J. J. Gordon and became widely known through his 1961 book, Synectics: The Development of Creative Capacity. Traditionally, it has been used as a facilitated approach to creative problem-solving, often involving groups.

Its distinctive aim is to make the familiar strange and the strange familiar: to step away from habitual ways of describing a problem and view it through experiences drawn from other contexts. An unexpected association, however, is not necessarily a useful solution. The historical importance of the method must be distinguished from evidence that a particular idea will work.

Four Kinds of Analogy

Synectics commonly introduces four mechanisms for shifting perspectives. These are prompts for creative exploration, not four scientifically guaranteed rules of inference.

Direct analogy. Look for systems in other domains that perform a similar function or address a comparable problem. For instance, designing a system that repeatedly supplies repair material to damaged locations might suggest investigating biological circulation and transport networks. The functional relationships matter more than superficial resemblance.

Personal analogy. Imagine becoming the object or process under investigation. Asking, “If I were the inside wall of this pipe, where would stress be concentrated?” might reveal neglected considerations. Identifying with an object, however, does not establish physical facts.

Symbolic analogy. Recast a problem using an evocative phrase, image, or apparent tension: “rigid yet flexible,” or “open yet protective.” Such formulations can prompt fresh perspectives but do not by themselves establish a feasible design.

Fantasy analogy. Temporarily remove real-world constraints and imagine the ideal outcome. “What if the material could find every crack and repair it repeatedly?” The fantasy can then be decomposed into potentially realizable functions such as detection, transport, and reaction.

What Does the Method Actually Do?

Synectics invites participants to move away from the original problem, explore analogies, and return with possible transfers. A session may involve defining the problem, searching for analogies, elaborating associations, transferring mechanisms, and evaluating feasibility. This description should not be mistaken for a uniquely valid or mandatory procedure.

The crucial question is not whether an analogy sounds interesting, but what exactly corresponds across the two domains. “Blood vessels resemble pipes” is a superficial comparison. “A network transports material repeatedly to locations where demand arises” expresses a relationship worth investigating.

This connects to Dedre Gentner’s structure-mapping theory, which emphasizes mapping relations rather than relying solely on surface attributes. Structure-mapping theory provides an independent account of analogical reasoning; it should not be confused with empirical validation of Synectics as a complete workshop method.

Why Synectics Deserves Critical Evaluation

Synectics offers techniques for escaping habitual patterns of thought. Yet following its traditional steps does not guarantee greater creativity or more successful innovations. Outcomes depend on the problem, the participants’ knowledge, facilitation, and the standards used to evaluate results.

There is also a selection problem: after a successful innovation, a compelling analogy can often be described retrospectively, while identifying a productive source analogy in advance remains difficult. Novelty alone is not evidence of usefulness, feasibility, or correctness. Synectics is worth understanding as an established creativity method, but its procedures should not be treated as authoritative simply because they have familiar names.

Can LLMs Turn Analogy into a More Systematic Process?

Large language models make it possible to explore analogies across more domains than a small workshop group could readily bring to mind. An LLM can help describe a problem in terms of functions and constraints, propose analogous systems from technology, biology, or social organization, and formulate candidate transfers.

A possible AI-assisted process would be:

  1. Describe the target function. Replace product-specific names with the underlying problem, such as “detect local damage and repeatedly deliver repair material,” while retaining essential constraints.
  2. Search for mechanisms in other domains. Look for documented examples in biological circulation, replenishment systems, or distributed supply networks.
  3. Map relationships explicitly. Identify what corresponds to the source, transport path, demand signal, feedback, and repair mechanism.
  4. Generate transfer hypotheses. Suggest how the underlying mechanism might be adapted to the original problem.
  5. Identify mismatches and constraints. Examine physical compatibility, cost, size, reliability, and technical feasibility.
  6. Verify sources and test proposals. Confirm that cited examples exist and operate as claimed, and distinguish evidence from speculation.

The self-repair example above illustrates a reasoning process; it does not establish that a proposed new system is technically feasible.

Research on LLM-based analogical reasoning has reported encouraging results on some tasks. However, other studies show persistent limitations in mapping detailed relations across domains. A sensible division of labor is therefore to use AI to widen the search and generate candidates while keeping structural correspondence and feasibility open to independent evaluation.

Synectics in ThinkNavi

ThinkNavi presents Synectics as a method for changing perspectives through analogy, not as a formula for producing original answers. A possible free tool could provide a structured worksheet for the four analogy prompts, allowing users to record source domains, mapped relations, transfer ideas, and objections. An LLM-assisted edition could support cross-domain candidate discovery, explicit mappings, hypothesis comparisons, and evidence tracking. These are proposed directions, not claims about functionality already released.

The objective is not merely to digitize a traditional workshop script. It is to develop analogy-based exploration into a more traceable, testable process for discovering alternative ways to frame and solve problems.

Related Topics

References