Synonymous Commands in Interactive Fiction: When “Look” and “Study” Mean the Same Action

To give “look at the lens” and “study the lens” consistent feedback, first define exactly what an observation does. Then place expressions that meet that definition in the same row. The criteria are the target, the method, and the end state. Similarity of meaning makes an expression a candidate; it does not establish equivalence.

The old photography studio below is an entirely fictional teaching example. Its setting, inputs, responses, and judgments were constructed for explanation. They are not real user experiences, measured results, hands-on tests, or DramaFork product features. This article addresses only single commands whose meaning is already clear, not the full process of handling free-form input.

An old lens sits on a table, still assembled. A closed lens case stands beside it. The author needs to solve a specific problem: changing how a player asks to look at the lens should not sometimes reveal scratches and sometimes produce “You can’t do that.”

Start with the canonical action: “Observe the lens’s exterior.” Its target is this lens on the table. Its method is visual inspection under the current lighting and from the current position, without moving, wiping, or disassembling it. At the end, the lens remains in place with its structure unchanged, and the player learns about a scratch on the outer ring and specks of dust on the visible glass.

The end state includes both the object and the information obtained. Checking that the object has not changed is insufficient: neither “look at the lens” nor “inspect the inside of the lens” necessarily moves anything, but the latter requests information beyond the exterior.

This example also assumes that ordinary observation has no separate duration tiers and that no clues require closer scrutiny. Under these rules, “study” can describe the manner of looking without adding an operation. That assumption does not apply if your work makes observation time a cost.

Fill in the applicable target and canonical action first, then list equivalent expressions. Finally, test the boundaries against similar but non-equivalent commands. Do not start with a long list of verbs and force every object to fit it.

The third row reserves a place for a separate action. It does not mean the player is currently allowed to disassemble the lens. It records that “take apart” belongs to another operation, preventing the author from quietly reducing it to observation just to accommodate an input. This article does not supply disassembly conditions or outcomes, so the row cannot establish that the disassembly flow is complete.

When reusing the table, keep all five columns for every observation target. Expressions within one row may share feedback; moving between rows requires checking the target and state again. Both the lens and its case accept “study,” but they clearly provide different information.

This example accepts “look closely at the lens” because “closely” introduces no new form of contact, time cost, or obtainable information. The trade-off is giving up the distinction between a cursory look and close scrutiny: stronger wording cannot buy the player a hidden clue.

If the author wants an inscription to be discovered only through close scrutiny, that scrutiny should become a separate action, with its additional information and conditions specified and the table revised accordingly. You cannot declare two expressions equivalent while rewarding only “study.” That would still leave players guessing which word the author prefers.

“Rotate the lens to get a look” changes the method of observation and includes moving the object. The words “get a look” do not make it equivalent. “Wipe the lens” changes its surface state; “take the lens apart” changes its structure; “inspect the inside of the lens” requests a different scope of information. Each crosses a different boundary of the original action.

An expression with an unclear scope, such as “inspect the lens,” is left out of this table for now. The author must establish whether it means examining the exterior or checking functionality in this scene before deciding where it belongs. There is no need to force an unresolved meaning into an existing action just to make the vocabulary list look complete.

For a paper walkthrough, start every input from a copy of the same scene: the lens is assembled, its position is unchanged, and it has not yet been observed. This lets you compare wording without mistaking effects caused by sequence for a synonym problem.

Input: “Look at the lens.” The target is clear, the method is visual, and the end state matches the first row. Use the lens-exterior feedback and record that the player has learned about the scratch and dust specks.

Input: “Study the lens.” Under this example’s rules, it adds neither observation time nor a closer level of scrutiny, so it still uses the first row. If the draft says “Nothing catches your attention,” that contradicts the information outcome of the previous input. Revise the feedback instead of deleting this equivalent expression.

Input: “Take the lens apart.” The structural change exceeds the observation boundary, so hand it to the disassembly action. In this teaching example, the draft may carry the note “Disassembly rules still need to be defined,” but it must not report a successful observation and pretend to have handled the command.

Input: “Study the lens case.” The verb is the same, but the target is different. Use the second row and describe only the case’s exterior. If the feedback mentions padding inside, it has silently opened the case or looked through its walls. Remove that out-of-bounds information.

Finally, check consecutive inputs separately: first look, then study. If the scene has not changed, the second response may briefly recap existing discoveries. It must not add a clue because the wording changed or deny the scratch seen the first time. If the lighting or object state changes, rewrite the outcome under the new conditions; the feedback cannot be required to remain identical forever.

After a player enters “study the lens,” the feedback need not mechanically announce “You executed ‘look’.” Canonical action names primarily help authors manage mappings. Describing the visible findings directly in the prose is usually enough to tell players what happened.

To emphasize a boundary, you can retain “The lens remains where it was.” This establishes that the object has not moved, but does not by itself prove that it was not disassembled. An editor must still check the full response for any added rotation, wiping, or disassembly.

Before handing off the draft, substitute every equivalent expression in the table into the same scene and check that the information, object state, and costs agree. Then try the similar but non-equivalent commands to see whether any have been wrongly included. Two common failures are changing only the opening verb while keeping an old branch’s outcome, and replacing verbs globally without accounting for differences between targets.

This table applies to command text with an established target and a single, clearly defined action. It does not prove that a program can recognize arbitrary rewordings, nor does it cover omitted targets, chains of actions, or ambiguous intent. What the author delivers is a semantic agreement that can guide implementation and checking: which expressions share outcomes, and which differences must remain.