Using Narratives to Sharpen Testing Skills

Published on October 5, 2025

As testers, we spend much of our time reviewing requirements, specifications, and user stories. We’re looking for ambiguities, inconsistencies, and contradictions. However, these analytical muscles can be exercised anywhere, including in narrative fiction. Let’s dig in!

Here I’ll share an example from a science fiction novel that demonstrates how easy it is for even published, edited writing to contain spatial and logical inconsistencies. The skills you can use to spot these problems are the same ones you use when reviewing a feature specification or tracing a user workflow.

I should note that isn’t the first time I’ve used this particular author’s work for testing exercises. In a previous post, I examined how a plot point in another story didn’t didn’t align with cause-and-effect logic. Today’s example is different: it’s about spatial and temporal consistency. Both are critical testing skills, just applied to different kinds of ambiguity.

The Setup

In the prologue of Jack McDevitt’s novel The Engines of God, two human explorers, Priscilla Hutchins (a pilot) and Richard Wald (an archaeologist), are visiting Iapetus, one of Saturn’s moons. They’ve come to examine an ancient alien artifact: a three-and-a-half-meter tall ice sculpture of a winged, clawed creature. This “Monument” has stood on the frozen plain for over 20,000 years.

Crucially, at least per the text, the explorers begin their investigation at the statue itself, located in the center of a valley on the plain. Around the statue are footprints: six-toed, barefoot tracks left by the alien who carved and placed it millennia ago. The humans follow a raised walkway (a “ramp” installed by the Park Service to preserve the original tracks) as they trace the alien’s movements across the landscape.

Your Challenge

As you read the following passage from the Prologue, try to map out the alien’s movements. Where did she land her ship? Where did she walk? In what order did events occur? Can you create a mental diagram of the geography and timeline? More importantly: do all the details actually line up consistently?

This portion of the prologue is reproduced here with the permission of Ace Books, part of Penguin Random House.

They followed the ramp out onto the plain. Off to one side they could see the booted tracks of the astronauts. Approximately a kilometer and a half west, her prints appeared.

There were two sets, going in opposite directions. She wore no shoes, and the length of both the foot and the stride, measured against the anatomy of the ice figure, suggested a creature about three meters tail. They could distinguish six toes on each foot, which was also consistent. “Almost as if,” Hutch said, “the thing climbed down and went for a walk.”

Chilling thought, that. They both glanced reflexively behind them.

One set of tracks proceeded west into the uplands.

The other wheeled out across the plain, on a course well north of the artifact. Astronaut prints, and ramps, followed in both directions. Richard and Hutch turned north.

“The bare feet shook them up,” said Richard. “Now, you and I could match the trick, if we wanted.”

After about a quarter-kilometer, the prints stopped dead in the middle of the snow. Both sets, coming and going. “There must have been a ship here,” Hutch said.

“Apparently.” The snow beyond the prints was untouched.

The ramp circled the area, marking off a space about the size of a baseball diamond. Richard walked completely around the circle, stopping occasionally to examine the surface. “You can see holes,” he said, pointing them out. “The ship must have been mounted on stilts. The prints show us where the creature first appeared. It — she — walked off the way we’ve come, and went up into the hills. She cut a slab of rock and ice out of a wall up there. We’ll go take a look at the spot. She fashioned the figure, put it back on board, and flew it to the site.” He looked in the direction of the ice figure. “There are holes back there, too.”

“Why haul it at all? Why not leave it up in the hills?”

“Who knows? Why put something here and not there? Maybe it would have been too easy.” He tapped the ramp with his toe. “We’re in a valley. It’s hard to see, because the sides are low, and the curve of the land is so sharp. But it’s there. The ice figure is located precisely in the center.”

After a while they went back the other way, and followed the tracks into the hills. The walkway plunged through deep snow and soared over ravines. The prints themselves twice went directly up to sheer walls and stopped. “They continue higher up,” said Richard.

“Anti-gravity?”

“Not supposed to be possible. But how else would you explain any of this?”

Hutch shrugged.

They entered the ravine from which the ice and stone for the figure had been taken. A block had been sliced cleanly out of one wall, leaving a cut three times the visitor’s height. The prints passed the place, continued upslope, and petered out on thick ice. They reappeared a little farther atop a ridge.

The ground dropped sharply away on both sides. It was a long way down.

Richard strode along the ramp, submerged in his thoughts, not speaking, gazing neither right nor left. Hutch tried to caution him that the energy field provided fair traction at best, that the light gravity was treacherous. “You could sail off without much effort. You’d fall kind of slow, but when you hit bottom, there would be a very big splash.” He grunted, and went a little easier, but not enough to satisfy her.

They continued along the crest of the ridge until the tracks stopped. It was a narrow place. But with a rousing view of Saturn, and the breathless falling-off of the worldlet’s short horizon.

Judging from the confusion of tracks, the creature might have been there for a time. And then of course she had doubled back.

Richard gazed down at the prints.

The night was full of stars.

“She came up here before she cut the ice,” said Hutch.

“Very good. But why did she come here at all?”

I put a little more material than you needed in order to respond to the challenge. I hope you’re intrigued and go buy the book! It really is quite good and it’s part of a series.

Analysis

If you tried to map out the alien’s movements, you likely encountered some frustrating contradictions. Let’s break down the problems using the same analytical approach we would apply to a technical specification.

Issue #1

Probably the best description for this issue would be contradictory directional references.

The problem: Richard says the alien “walked off the way we’ve come, and went up into the hills.”

Why this doesn’t work: Hutch and Richard came from the statue. They didn’t come from the hills. Richard is standing at what he identifies as the landing site, looking at footprints. If the alien walked “the way we’ve come,” she would have walked toward the statue. However, Richard says she went to the hills instead.

In terms of a testing parallel, this is like a user flow that says “After completing checkout, return to the shopping cart” when the user actually came from the product page. The reference point is inconsistent with the established context.

Issue #2

This issue might best be framed as one involving missing footprints.

The problem: Richard explains that the alien “fashioned the figure, put it back on board, and flew it to the site.” The text also mentions “There are holes back there, too” (at the statue location, implying the ship landed there).

Why this doesn’t work: If the alien carved the statue near/on her ship at the landing site, then flew it to the statue location, there should be no footprints connecting these two points. Yet the narrative seems to imply continuous tracks leading from the landing site toward the statue area.

The testing parallel here is that this is like a specification that describes a user uploading a file directly to cloud storage, but then references validation steps that could only occur if the file passed through a local staging area first. The described mechanism doesn’t support the claimed outcomes.

Issue #3

This issue is all about spatial geometry confusion.

The problem: The text says “Approximately a kilometer and a half west, her prints appeared.” Later, after turning north, “After about a quarter-kilometer, the prints stopped dead in the middle of the snow.”

Why this doesn’t work: These appear to describe the same landing site (where prints “appeared” and where they “stopped”), but the directions and distances don’t align. If they’re 1.5km west of the statue and then walk a quarter-kilometer north, how are they at the same place where the prints originated?

The testing parallel is that this is like an API specification that describes an endpoint as /api/v2/users/{id} in one section and /api/users/v2/{id} in another, both allegedly referring to the same resource. The inconsistency suggests the author hasn’t fully worked out the structure.

Issue #4

This issue is all about ambiguous temporal references.

The problem: “After a while they went back the other way, and followed the tracks into the hills.”

Why this doesn’t work: The phrase “went back the other way” suggests retracing their steps, but they’re actually going to a new location (the hills/ravine) they haven’t visited in this scene. “Back” implies return; this is actually proceeding forward to unexplored territory.

The testing parallel here is that this resembles workflow documentation that says “Return to the dashboard” when the user is actually navigating to a new reports screen for the first time. The language suggests a return journey when it’s actually initial exploration.

Extra Challenge!

I’ve identified four contradictions in this passage. But there’s at least one more spatial oddity I didn’t discuss. Can you spot it?

The Corollary with Testing

This passage demonstrates several important testing principles.

  1. Mental Models Matter. When reviewing requirements, try to build a complete mental model of the system, workflow, or data flow. If you can’t construct a coherent model, there’s likely ambiguity or contradiction in the specification.
  2. Track Your Reference Points. Notice how the confusion stems partly from unclear reference points (“the way we’ve come”). In specifications, phrases like “the previous screen,” “the original value,” or “the default state” can be similarly ambiguous. Always ask: previous to what? Original when?
  3. Follow the Implications. The author stated the statue was flown to its location, but didn’t think through what that meant for footprints. Similarly, spec writers often describe a feature without considering all its logical consequences. If X is true, what else must be true? What becomes impossible?
  4. Draw It Out. If you’d tried to literally draw a map of this scene, the contradictions would have become obvious immediately. The same applies to testing: sketching user flows, state diagrams, or data models often reveals inconsistencies that aren’t apparent in prose.
  5. Even Professionals Miss This Stuff. This novel was written by an award-winning author, published by a major publisher, and presumably reviewed by editors. Yet these spatial contradictions made it through. The lesson? No matter how experienced the team, ambiguity and inconsistency can slip through. That’s why testing is valuable.

Jack McDevitt writes compelling science fiction with intricate scenarios that should hold together logically, but sometimes the complexity trips him up. What’s particularly interesting is the contrast between my two posts.

  • My 2016 post examines a plot logic failure in terms of cause-and-effect reasoning breaking down.
  • This post examines a spatial/descriptive failure. The geography and movements don’t form a coherent picture. It’s about consistency and reference points.

I think both are excellent testing exercises, but they target different analytical skills.

  • The first tests logical reasoning.
  • The second tests systematic modeling.

Your Turn!

The next time you’re reviewing a specification, user story, or requirements document, approach it like you approached this passage.

  • Can you map out the described flow or structure?
  • Do all the directional/temporal references align?
  • If the document claims X happens, are all the preconditions and consequences accounted for?
  • If you drew a diagram, would everything connect logically?

These skills transfer directly from fiction to functional specs because ultimately, both are attempts to describe a coherent reality through language. And language, as we’ve seen, is tricky.

Share