How do you stop readers losing track of speakers?

Watch the briefing

Watch the full video on YouTube

Full briefing transcript

Can we leave? Not until I see the chart. I gave it to you. With three people in the scene, a reader cannot rely on a simple back-and-forth pattern. Clear attribution is not a failure of character voice. Mark the changes where more than one speaker is plausible. Use a name, a plain tag, or an action that belongs to the speaker, then let the conversation move.

Our before passage places Mara, Ivo, and the inspector beside a gate. Can we leave? Not until I see the chart. I gave it to you. You gave me the old one. Then I'll fetch the new one. The order contains a request, a condition, and a dispute about a chart. But the untagged lines do not securely assign the request or the promise to a particular person.

Now the revision. Mara, Ivo, and the inspector stood beside the gate. Can we leave? Mara asked. Not until I see the chart, the inspector said. Ivo raised his empty hands. I gave it to you. You gave me the old one, the inspector said. Then I'll fetch the new one, Ivo said. Each change begins its own paragraph on the page, and the participants are identifiable.

Follow the roles rather than guessing the order. Mara requests departure. The inspector sets a condition. Ivo claims to have supplied the chart; the inspector distinguishes the old chart from the one needed; Ivo offers to fetch the new one. His empty hands connect an action to his line. They do not establish whether his earlier delivery claim is accurate. Attribution clarifies who says what without verifying their claims.

Separately, notice the punctuation in this US-style example. Can we leave keeps its question mark inside the quotation marks before Mara asked. The chart statement uses a comma inside the closing quotation mark before the inspector said. Ivo raised his empty hands is a complete action sentence, followed by a separately quoted statement. New paragraphs mark new speakers. These are conventions of this style, not a rule that every scene needs identical tags.

For a clarity pass, cover the names and ask where two people could plausibly say a line. Restore an identifier at those points. Then check again with the actual names visible. You may remove a redundant tag when context really makes the next speaker clear, but do not trade comprehension for a lower tag count. Distinct voices help, yet a new reader should not need to decode a verbal fingerprint on every turn.

The goal is effortless orientation, not maximum labeling. Preserve a longer speech in one speaker's paragraph, and use the conventions required by your editorial style; other styles can differ. Mark speaker changes where ambiguity arises. Ask a reader to identify who makes the request and who promises the next action. Next, we will decide whether a dialogue tag adds useful information or merely draws attention to itself.