Surface vs. Structural Similarity

Two problems can share visible details without sharing the relationships that make a solution work. For example, a software subscription and a gym membership look different but both exchange continuing access for recurring payment. Structural comparison looks for such matching relationships rather than relying only on familiar objects or settings. Check this: what relationship makes your proposed comparison useful, and where does it stop applying?

Distinguish attributes from relationships

Surface similarities include shared objects, colors, names or settings. Structural similarities concern relationships, such as one part supplying another or one action causing an outcome. A useful analogy matches relevant roles and relationships across situations; it need not match every visible feature.

Use a familiar example carefully

Water systems and electrical circuits can be compared by relating pressure differences to voltage differences and flow to current. These are correspondences in how quantities relate, not claims that water and electricity are identical. The usefulness of an analogy depends on which relationships it preserves and which differences matter for the problem.

Let cues suggest comparisons, then check them

A surface match may help an example come to mind. It can be useful, and it is not automatically a distraction. Once you have a candidate analogy, compare the relationships needed for the proposed solution. Something that looks similar may work differently, while something from a different domain may share a useful pattern.

Avoid copying the visible result alone

Imagine two teams adopt the same office layout, but only one has a process for sharing information and resolving disagreements. Copying the layout alone does not copy that process or establish that it caused successful collaboration. This hypothetical contrast illustrates why a matching appearance is insufficient evidence for transferring a solution.

Make the comparison explicit

List the parts in each situation, identify the roles they play, and say which relationships match. Then identify differences that could break the inference. Relational words and explicit comparisons can help bring the common pattern into focus, but a structural match still needs checking against what is known about the target situation.

Surface and structural similarity are useful distinctions, not a rule to discard all object information.

• Gentner and Markman (1997) describe analogy and similarity in terms of structural alignment and mapping, emphasizing relational correspondences and the interactions between object matches and relational matches. [1]
• Gentner (2016) reviews evidence that learning and using relational language can foster relational reasoning, and explains how comparison can make shared relational patterns more salient. [2]

A proposed structural correspondence may omit a causally important difference, so it is a candidate for inference rather than proof that a solution transfers. Surface details and relevant relationships can correlate in real settings; identifying a useful relationship does not justify ignoring observations. What can be mapped also depends on how the situations are represented and what the learner knows. Explicit comparison and relational language can help, but do not guarantee a correct analogy.