The No Modes pattern solves User Action Congestion by removing the need for the user to choose what mode they are in before acting.

The problem: User Action Congestion

User Action Congestion happens when too many possible actions compete for the same UI space or mental space.

For example, in Seed a user may be looking at a document and want to:

read

edit

comment

cite

copy link

select blocks

move blocks

open version history

open people

reply to a conversation

create a child document

inspect references

If the interface exposes all of these as buttons, tabs, menus, toolbars, sidebars, and modes, the user has to constantly ask:

“What mode am I in?”
“Where is the action I need?”
“Do I need to select first?”
“Am I commenting, editing, navigating, or organizing?”

That creates congestion.

How No Modes helps

The No Modes pattern says:

The object should remain the same. The user action should emerge from context.

Instead of switching between separate modes like:

Read mode

Edit mode

Comment mode

Select mode

Move mode

Cite mode

Organize mode

…the user interacts directly with the document, block, comment, card, or selection, and the relevant actions appear only when needed.

So the UI changes from:

“Choose a mode, then act.”

to:

“Point at the thing, then act.”

Example in Seed

Bad modal design:

Click “Comment mode”

Select text/block

Write comment

Exit comment mode

Enter edit mode

Edit block

Enter citation mode

Add citation

No Modes design:

Click into text → edit

Select text → actions appear: comment, cite, copy link

Hover block → block handle appears: move, copy, reference

Open comment thread → reply box appears inline

Drag document/card → organize it

Type / → create structure or insert object

The user never thinks “which mode am I in?” The object itself reveals possible actions.

Why it reduces congestion

No Modes reduces congestion in three ways.

First, it hides irrelevant actions until the user has given context. A blank document surface does not need to show every possible action. Once the user selects text, only text-relevant actions appear.

Second, it localizes actions near the object. Actions belong to the thing being acted on: block actions near blocks, comment actions near comments, document actions near documents.

Third, it turns global controls into contextual affordances. Instead of one giant toolbar containing everything, each object exposes a small set of meaningful actions.

The deeper principle

The interface should not be organized around application modes.

It should be organized around objects and intentions.

In Seed terms:

Document → read, edit, publish, cite, version

Block → select, move, comment, reference, transclude

Comment → reply, resolve, quote, link

Person → view contributions, mention, follow

Citation → inspect source, navigate, quote

Board card → move, prioritize, open document

Each object carries its own action surface.

The tradeoff

No Modes does not mean “no state.” The system still has state: selection, focus, hover, active thread, current document, current draft.

But the state should be lightweight, visible, and reversible.

Bad mode:

“You are now in Comment Mode.”

Good contextual state:

“You selected this sentence, so Comment is available.”

Product sentence

You could describe it like this:

The No Modes pattern reduces User Action Congestion by making actions contextual to the object under focus, instead of forcing the user to switch between global modes. The interface stays centered on the document, while editing, commenting, citing, linking, and organizing emerge from selection, focus, and intent.

For Seed, this is especially important because the product is not just an editor. It is a relational workspace. Without No Modes, the number of possible actions will explode. With No Modes, the document remains calm, and the network reveals itself only when the user touches the relevant object.