An Analogy Has to Survive the Crossing

The demonstration gave different kinds of information their own movement. Headlines occupied the foreground. Images and supporting text travelled in other layers. Chapters gave the reader places to arrive.

I brought it into a development conversation about Bookerish, our touring and venue-coordination product. We already had the sections of a product story: a problem to introduce, an explanation to develop and somewhere to take the reader. Here was a way those sections might become an experience of moving through the argument.

The attraction was more specific than wanting a page that moved. A headline and its supporting material have different jobs. Giving them different positions and movement offered a way to compose that relationship. The reader might encounter the proposition, then the material that developed it, with a sense of progression through the whole.

Our discussion began to map the existing story onto the demonstration's layers. We would need our own content and choices about its behavior. We did not need the controls that let someone tune the demonstration. I asked for a separate mock-up so we could see the proposal in our own setting before considering it for the homepage.

There is a particular pleasure in this kind of recognition. Something made elsewhere gives a shape to a possibility in work already under way. You can see enough of the connection to want to try it without yet knowing all that the attempt will require.

In our case, a phone made some of those requirements apparent.

Headlines and supporting copy could drift apart on the narrow screen. Text could be pushed beyond the visible area. The navigation stopped before the closing section. We had composed a story with a destination the reader could not reliably reach.

The repairs changed the relationship among the layers. Foreground and supporting text were given matching mobile movement speeds. Chapter positioning and navigation were revised. Later, I pointed out that the visible navigation needed another stop for the final section.

Some of what we changed was what had made the demonstration distinctive. Independent movement looked promising when we considered the composition. On the smaller screen, it could separate material that needed to be understood together. Keeping those elements together mattered more than keeping them faithful to the original motion.

This is where borrowing begins to require judgment. It is possible to reproduce the conspicuous parts of an idea while losing the relationship that made them useful. The movement was easy to notice. The harder question was what it was doing for the story, and whether it still did that when the content and available space changed.

Research on analogy gives that first recognition an appropriately limited role. In experiments by Dedre Gentner, Mary Jo Rattermann and Kenneth Forbus, the similarities that helped people recall an earlier story differed from the relationships they judged useful for drawing an inference. Being reminded of something and being able to reason from it are distinct accomplishments. In our design work, the resemblance supplied a promising question; the prototype let us begin asking it in the receiving setting.

Not every transfer demands an elaborate experiment. In another Bookerish exchange, I pointed to Homebridge's presentation of structured information and asked for a comparable organization of our text resources. Full links and a readable hierarchy crossed between business contexts as familiar documentation practices. The discussion then questioned whether repeating a large set of links in every document was useful and proposed a lighter treatment.

There was no need to make that small act of reuse heroic. The correspondence was close, the change straightforward. We can identify the reference and the subsequent implementation without claiming that the formatting improved retrieval by an AI system.

The motion prototype travelled further. It took a composition demonstrated with one set of material and asked it to carry another, for readers using different screens. Its transformation was part of finding out what we had borrowed.

Of course, adaptation can become such a generous explanation that almost anything qualifies. If every important relationship changes, the original reference may be little more than a prompt to begin. Here, something identifiable remained after the mobile repairs: different kinds of information arranged into a navigable story. We could change the movement while continuing to examine that possibility.

I thought the prototype was promising enough to consider for later use. We had built something we could inspect, but had not established a benefit to readers. Whether it explained the product better than a conventional page remained open. A simpler page might have done the job with less effort.

That uncertainty does not make the initial encounter worthless. It does keep my interest in the result from becoming evidence on behalf of an audience that had not supplied it.

In Listen After the First Take, I describe recognizing a related process in development years after working as an actor. That recognition belongs to my present understanding of the two crafts. It cannot, by itself, tell us that a theater experience caused a particular engineering decision. Some connections become available only when we look back; others, like this design reference, are there in the record before we make the attempt.

We do not need to resolve the complete history of an influence before trying something. We do need to let the new setting answer. In the Bookerish prototype, answering meant altering an attractive feature of the source rather than reproducing it more faithfully.

The closing section did not need to move like the demonstration. It needed to be reachable. Finding that out changed the borrowed idea without making the original encounter irrelevant.

See also

• What the evidence allowed us to say — Clarifies why an implemented change and a demonstrated benefit to users require different support.