Distributing Narrative Authority

TabletopInteractive FictionLive Game

The rules that decide whose turn it is to say what is true.

At any table telling a shared story, someone is always deciding what happens next. In a traditional roleplaying game that someone is the GM, and the arrangement is so familiar it looks like the natural state of things rather than a design choice. Take the GM away and the choice becomes visible. Now the right to frame a scene, to voice the world’s resistance, to declare that the door is locked must come from somewhere. A game that does not say where will hand it by default to whoever talks loudest.

This article is about the mechanisms that hand authority around instead. A coin pushed across the table to buy an interruption. A card drawn that says you speak for this week and no one else does. A token earned by putting your character in danger. A role that rotates clockwise so the lens through which the table sees the story keeps changing hands. Each is a small machine for answering one question: by what rule does the right to author the fiction move from one person to the next.

The well-known games are exemplars, not the subject. Microscope, Fiasco, The Quiet Year and the rest are worth knowing because each isolates a different mechanism cleanly. What carries across them, and out into interactive fiction and large live games, is the underlying move: rules that take the authority a single narrator would otherwise hold and parcel it out on purpose. No one ends up owning the whole truth, so the surprise can land at your own table.

The Mechanism (How it works)

Authority is not one thing

The first useful move is to stop treating narrative authority as a single lump that one person either holds or does not. It is a bundle of separable rights, and most of the craft is in distributing them independently. A traditional GM happens to hold all of them at once, so the bundle reads as one solid thing until you try to share it.

Four rights matter most. The framing right initiates a scene. It fixes the place and the moment and then says what stands in front of everyone. The interrupt right modifies a scene already running and bends its shape without starting a new one. The override right cancels or rewinds something already established. The ending right decides when the whole story stops. A GMless game is, at bottom, a scheme for routing each of these four to the right person at the right moment, and a design that leaves any one of them unassigned discovers an authority vacuum exactly there.

The reason to be explicit is not tidiness. Every table distributes these rights somehow, even a traditional one. The only question is whether the distribution lives in the rules or in the personalities. Held in the rules, the distribution can be taught to a newcomer and put right when it drifts. Held by personality alone, it quietly favours whoever is most confident and offers no way to appeal. Making it explicit is what lets a game run without a designated authority figure at all. The price is procedural weight: every licensed moment of narration is a moment when everyone else is waiting, and the gaps between licensed moments are the central pacing problem of the form.

Token and wager economies

The oldest formal mechanism treats authority as a finite currency handed out at the start. The Extraordinary Adventures of Baron Munchausen gives each player coins. While another player tells their story, you push a coin forward to force a complication into it. They either take the coin and the complication, or refuse and insult you. From there you can escalate by adding more coins. The coin is not a resource that buys a good outcome. It is the price of interruption and the cost of social friction made literal, so that redirecting someone else’s story always costs you something.

Belonging Outside Belonging, first seen in Dream Askew, inverts the politics of that economy. Weak moves, where you put your character at risk or accept a hard position, earn you a token. Strong moves, where you reach out and reshape the fiction, spend one. Authority is coupled directly to exposure: you buy the right to author by first making yourself vulnerable. That is a different claim about who deserves to speak than Munchausen’s, where authority is hoarded and spent like capital. Here it is earned by going first into difficulty.

Fiasco uses dice as the control resource rather than coins. Players build a shared pool at setup and take turns drafting dice against the playset’s tables to fix the relationships and needs that bind their characters together. During play the dice migrate between a white pool and a black pool as scenes end well or badly, and the balance at the act break sets the Tilt that tips the back half of the game. The dice are content generator and authority marker at once: they record whose story has been going well and hand out the closing narration to match.

Cards, scene-framing rights and the draw as transfer

Cards make authority visible and portable, because the license to speak is a physical thing you are holding. In Once Upon a Time players hold story-element cards and a single Ending card. Whoever is narrating can be interrupted by any player who holds a card matching something just said, and playing it seizes the narrator role. Authority is not assigned by turn order at all. It is claimed by relevance, and the cards in your hand are the relevances you are allowed to claim.

The Quiet Year tightens this into one of the cleanest demonstrations of a separated framing right. A standard deck is dealt out across a year, each card a week. The player who draws reads its prompt and frames that week, naming the project that begins or the crisis that arrives. Everyone else may add to the shared map between draws. Only the draw confers the right to frame. So the framing right and the contribution right come cleanly apart, and the draw event is the one visible moment authority changes hands. This is the same separation that runs underneath any branching collaborative story, where the question of who gets to open a new thread is distinct from who fills it in; the structural version of that is Parallel Threads & Convergence.

Fiasco makes the framing right an explicit trade each turn. On your turn you choose to either frame your scene yourself and cede its outcome to the group, or let the group frame it and keep the outcome for yourself. You can author the setup or the resolution but never both, so no single player owns a scene from end to end. That trade is one instance of a deeper rule the form keeps rediscovering.

Oracles: handing authority to chance

A different family routes authority not to another player but to the dice acting as an absent GM. Mythic introduced the modern oracle: you pose a yes or no question to the fiction and roll against a probability ladder weighted by a chaos factor. The result is a verdict. The dice do not narrate. They pick an outcome and leave you to author what it means. Authority is split down the middle. The oracle holds the verdict while the player supplies the interpretation.

This is the mechanism that carries solo play, where there is no second person to hand the framing right to and the oracle stands in for the whole apparatus of an opposing will. The craft of building and consulting one is its own subject, Oracle & Solo-Play Resolution; here it matters as one more answer to the routing question, and as the answer with a distinctive failure. An oracle is delegated authority that the player is meant to keep overriding whenever the fiction demands it. Players who instead defer to it completely stop authoring altogether. The authority formally given to chance is then surrendered to it for real, and the dice end up holding what the table was supposed to keep.

Rotating roles and the fractured lens

The plainest distribution is to rotate the authoring role on a fixed schedule. Microscope hands one player the Lens each round; the Lens sets the Focus the round will explore and narrates first within it, then passes clockwise. The role is temporary and total: while you hold it you are the GM, and when you pass it you are not. Kingdom splits the role itself into three that run in parallel. The Power decides what the community does, the Perspective narrates the consequences and is proven right or wrong, the Touchstone voices how the community feels about it. No single player authors the whole shape of any event, because the decision, its outcome and its reception are held by three different people at once.

Rotation has its own characteristic failure, and it is the opposite of concentration. Pass the framing right too often, every couple of minutes, and no scene lives long enough to gain traction; the story fragments into a string of openings nobody develops. The cures are all forms of slowing the rotation down. A minimum scene length helps, as does the right to close your own scene, or the one-action-per-week discipline that The Quiet Year builds into its structure. The lens can fracture usefully, but only if each holder gets long enough to focus it.

The move that has to separate: adversity from response

Underneath the trades and rotations sits a single principle that explains most of them. The Czege Principle, named for the designer Paul Czege, holds that it is not satisfying for one person to author both a character’s trouble and the resolution of that trouble. A traditional game separates the two structurally. The GM is the source of adversity; the players are the source of the response. Remove the GM and that separation has to be rebuilt for every scene, by hand, inside the mechanism itself.

That constraint explains why the recurring shapes look the way they do. Role separation puts the trouble and the response in different mouths, as Polaris does with its ritual phrases (“And so it was”, “But only if…”) through which a protagonist’s player and an antagonist’s player negotiate each outcome in a fixed grammar everyone can audit. Oracle randomness sources the adversity from the dice so no player chose it. Token-driven weak moves let you author your own vulnerability while the table authors what comes of it. Fiasco’s random setup makes the binding relationships nobody’s particular agenda. The practical test for any new design is one question asked of every scene: is the same person authoring the difficulty and the way out of it? If yes, the design needs a mechanism to pull them apart.

The Specimen

The Quiet Year runs the whole authority question on a 52-card deck and almost nothing else. Each card is a week; whoever draws frames that week from the card’s prompt and may not be vetoed, while everyone else adds to the map but cannot reframe. The draw is the entire transfer mechanism, which makes the framing right unusually legible: at any moment you can point to who holds it. The game also carries a Contempt Token with no mechanical effect at all. You take one when you feel your input was ignored, and its only function is to make that social friction visible at the table rather than letting it fester in silence. It is a sharp reminder that a token need not be an economy. Sometimes it is just a signal the rules give players permission to send.

Veto, the override, and the safety layer

A tempting design instinct is to let players reject each other’s contributions outright. Almost every game in the tradition refuses, because a free veto produces consensus paralysis. If anything you add can be struck down, you start offering only what feels safe. The fiction goes bland. Microscope states the alternative as a rule, no veto. If a contribution bores you, you simply build nothing on it, and that social cold shoulder is trusted to discipline the table without a procedural rejection right. Contributions stick by default; the fiction keeps moving because nothing stops it without cause.

The one override the tradition does grant universally is the safety tool. The X-Card and its descendants let any player, at any moment, pause, rewind or exit, and that right outranks scene-framing, turn order and any token economy in play. It is the single effective veto in GMless design, and it is deliberately scoped to comfort rather than narrative taste. You cannot strike a plot turn for being dull. You can always stop one that has turned harmful. That careful scoping is what keeps the no-veto default from being dangerous: an override held by everyone, yet pointed only at safety. The wider treatment of these tools as a system layer is Safety Systems as a Mechanical Layer.

A map of the mechanisms

The families above sort cleanly by what carries the authority and which of the four rights each one chiefly moves. The data-dense version is a table.

Mechanism	Carrier	Right it chiefly moves	Representative games	Characteristic failure
Token / wager economy	Coins or move-tokens	Interrupt	Baron Munchausen, Belonging Outside Belonging	Concentration in a strategic player
Dice-as-marker	A shifting dice pool	Ending / weighting	Fiasco	Setup luck overshadows play
Card-as-license	Cards in hand	Framing / interrupt	Once Upon a Time, The Quiet Year	Hand starvation stalls a player
Oracle	Dice or a deck	Outcome verdict	Mythic, solo engines	Passive deference, players transcribe
Role rotation	A passed token or seat	Framing (whole)	Microscope, Kingdom	Rotation too fast, scenes fragment
Role separation	Fixed ritual roles	Adversity vs response	Polaris	Rigid phrasing slows the table
Override / safety	A card or gesture	Override	X-Card, Script Change	Under-used through social pressure

Read down the failure column and a pattern shows: every mechanism fails either by concentrating authority back into one person or by diffusing it until no one carries a scene. Those are the two attractors any distribution scheme is steering between, and naming which one a given table is drifting toward usually points straight at the fix.

The same question in other media

Interactive fiction asks the authority question in a sharper form, because one side of the table is a fixed author who wrote everything in advance and the other is a player meeting it live. The whole design problem is the balance between them. Give the player total control and the authored shape dissolves into aimlessness; give them none and they feel toyed with, an audience pretending to act. A piece of IF is tuned somewhere on that line, and the parser-and-world-model architecture is the mechanism that meters it: the world model fixes what is true and the parser decides which of the player’s intentions the truth will honour. That is the same split an oracle makes. The author now stands where the dice once stood and holds the verdict, while the player authors the meaning of their own attempt. The procedural-rhetoric reading of how those authored rules argue a position is IF Mechanics & Procedural Rhetoric.

Large live games answer the question at the opposite scale, with information rather than tokens. A megagame distributes authority by distributing knowledge: each faction is briefed on its own objectives and its own intelligence. The careful part is what it is not told, the gap in its knowledge that stops it seeing the board in full. Watch the Skies runs a double-blind master map held by Control and forces factions to spend actions to learn anything beyond their own assets. The right that circulates is not framing but the right to know. Information becomes the currency that authority is traded in. Whoever holds a fact others lack must decide whether to share it or hoard it, and that share-or-hoard tension is what gives the form its charge. Holding that asymmetry together against the table’s natural urge to compare notes is its own craft, Information Asymmetry: Hidden Information vs Hidden Roles.

Where this fits

The bigger picture. Distributing narrative authority is the structural question sitting underneath every game that tells a story without a single boss. It is the layer that decides who the other mechanisms serve. A resolution method only matters once you know who is allowed to call for the roll, which is Mechanical Resolution Types; a PbtA Move frames a moment of fiction but says nothing about who gets to frame the scene around it; a playbook bundles a character’s authority over their own corner of the story. Where a game sits on the spectrum from one concentrated narrator to fully distributed authorship is one of the deepest things its format taxonomy records, and reading a design as a distribution scheme often explains its feel faster than reading its dice: that wider map is TTRPG Format Taxonomy.

Next steps. Once you have decided how authority moves, the next questions are about the people moving it. The matter of how a facilitator who keeps some authority should use it falls to GM Role Models & Stances, which lays out the stances open to them. Where authority is held by table consent rather than by rule, and how a single arbiter’s calls stay legitimate, is Rulings vs Rules & the Authority Contract. When the distribution has to survive a lone player with no second voice to hand the framing right to, the mechanism that stands in for the absent table is Oracle & Solo-Play Resolution.