A power dynamic between two characters is easy to signal visually and difficult to make credible. Costume, posture and framing establish who appears to be in charge in about a second, and a reader accepts it for about as long. What makes it hold is the writing underneath: what each character wants, what each can withhold, and what it would cost either to walk away.

This applies whether the dynamic is explicit and negotiated or entirely unspoken, and it is the same craft either way. The visual layer is the announcement; the written layer is the evidence.

Power is about what can be withheld

In practical terms, one character has power over another to the extent that she controls something the other wants. Attention, approval, information, access, decision. Naming what that thing is, specifically, is the fastest way to make a dynamic feel real rather than performed.

It also explains why apparently one-sided arrangements are frequently mutual. The person who appears to be in control may need the arrangement more, and that asymmetry underneath the visible one is where the interesting writing lives.

The visual layer is only an announcement

Framing and posture establish a dynamic immediately, as described in the piece on viewer position, where consistent high and low angles make claims about hierarchy. That is useful for establishing and useless for sustaining, because a reader stops registering a constant signal within a few pages.

Sustaining it requires the dynamic to do something: produce decisions, cause friction, get tested. A power relationship that never has consequences is set dressing, and readers treat it accordingly however emphatically it is staged.

Dynamics that never shift go flat

A fixed arrangement is stable and inert. Something has to move: a moment where the usual direction reverses, a request refused, a limit reached. The shift need not be permanent, and it works better when it is not, but it has to occur, or the reader is watching the same scene repeatedly.

This is the asymmetry engine described in the piece on writing two-handers, applied over a longer span. Each shift creates the next question and pulls a reader forward rather than presenting variations on one situation.

The ability to leave

What separates a dynamic that reads as chosen from one that reads as something else is whether leaving is visibly available. That means it should be materially possible, and ideally referenced, so a reader knows the arrangement continues because both people want it to.

This is the writing equivalent of the visible exit described in the piece on visible consent, and it does the same double duty: it is what makes the material responsible and it is what gives the situation stakes, since a choice that could be unmade is a choice worth watching.

Negotiation can be in the text

Where a dynamic is deliberate between the characters, showing them arranging it is usually more interesting than skipping to the result. Negotiation is a scene with two people wanting different things, which is exactly the material a two-hander needs, and it carries characterisation efficiently.

It also solves a legibility problem, since a reader who has seen an arrangement agreed does not have to infer it later. That distributes the signals through the work rather than concentrating them, which is the panel-travel argument made throughout these notes.

Aftermath is part of the arc

What happens afterwards tells a reader what the encounter meant. A scene that ends at its peak has an event and no consequence, which is the missing-change problem set out in the piece on what makes a scene. A few panels of return, adjustment or ordinary conversation carry more than an extended climax.

It is also where a power dynamic is most legible, because how two people behave once the arrangement is set aside says what it actually was. That contrast is unavailable to any work that stops early.

Writing a dynamic a reader believes

Name what is being withheld, treat the visuals as announcement rather than evidence, let the dynamic shift, keep leaving possible, show the negotiation, and write the aftermath. Six structural decisions, none of which are about how anything is drawn.

Characters here are fictional and unmistakably adult, and coercion is not published, as set out in the reader agreement.