Style changes via chat loops are vague, slow, and repetitive.
← Back to patterns
Direct Manipulation: A Step Beyond Programming Languages
Inline & Contextual
Tone of Voice
Direct Manipulation for Linguistic Style
Instead of iterating through "make it more professional... actually a bit more casual..." chat loops, a 2D tone pad lets users directly manipulate writing style along two axes: conciseness and formality. One drag replaces five chat turns.
Metaphor
Mixing Console
Interaction
Spatial, 2D Pad
Use case
Text Rewriting, Brand Voice, Content Adaptation
Failure mode
Weak Intervention
Control mechanisms
2D Tone Grid · Presets · Snap Points
Reference examples
Figma AI Tone · Apple Intelligence Writing Tools · Grammarly Tone Setter
Interactive Example
The live side-by-side is the comparison.
This prototype already contains the contrast between chat and the pattern response. It stands as the primary comparison surface for this page.
Stop describing the tone. Start steering it.
Tuning style through chat means serial, ambiguous loops: "a bit more professional... actually more casual..." Each turn is a guess at what the model heard, and "slightly" means something different every time. Five round-trips to land a voice you could have pointed at directly.
The Tone of Voice pad maps style onto two named axes you drag in real time — conciseness and formality — with snap-points for repeatable profiles. One gesture replaces five chat turns, the current position is always visible on the grid, and blending between tones becomes a continuous move rather than a negotiation.
Direct manipulation beats conversation whenever the thing you want to adjust lives on a continuous scale.
DOPE Lens
How this pattern improves AI interaction quality.
Observability
The current voice is a point on a labeled two-axis pad — position is the state, replacing memory of which adjectives were already tried
Directability
One drag moves both axes at once and the rewrite updates live under the cursor — five turns of adjective negotiation collapse into a single gesture
Predictability
Named presets snap to fixed coordinates, so "Executive" lands on the same voice every session — unlike "slightly more professional", which drifts per turn
Explainability
Axis labels and preset names state in plain words what each position produces, so the mapping from control to voice is inspectable, not inferred
Fit
Where it earns its place.
Use it
when
when
Users iterate on style — formality, conciseness, voice — across many short text outputs.
Avoid
when
when
The output needs editorial rationale more than exploratory tuning.
In the wild
Where this pattern already ships.
Real implementations from products users already know. Each entry shows how the pattern got applied — what shipped, where it sits in the product, what it traded off.
Figma AI Tone
Figma
Tone dial inside an axis with presets Executive / Technical / Basic / Educational. Direct-manipulation 2D pad — exactly the pattern.
Visit Figma ↗
Apple Intelligence Writing Tools
Apple
Select text → Writing Tools → Friendly / Professional / Concise preset chips with one-tap rewrite. Not a 2D pad, but the closest preset-driven tone manipulation in a major OS.
Visit Apple ↗
Grammarly Tone Setter
Grammarly
Tone detector + tone setter: the system reflects detected tone back to the user via emoji indicators before suggesting a target. Caveat — detection-plus-presets, not a continuous 2D surface.
Visit Grammarly ↗
Research Frame
Why this pattern exists, and where it breaks.
Map tone onto a direct manipulation surface with named presets.
Phrases like "slightly more professional" are ambiguous and serial.
A small set of axes can oversimplify voice and imply false precision.
Common questions
What people ask about this pattern.
When should I use the Tone of Voice pattern?
+
Users iterate on style — formality, conciseness, voice — across many short text outputs.
When should I avoid the Tone of Voice pattern?
+
The output needs editorial rationale more than exploratory tuning.
What problem does Tone of Voice solve?
+
Style changes via chat loops are vague, slow, and repetitive.
Why is chat the wrong fit for this?
+
Phrases like "slightly more professional" are ambiguous and serial.
Grounded in Research
Not just opinion.
The papers, guidelines, and usability studies this pattern rests on.
Ben Shneiderman
IEEE Computer · Vol. 16, Iss. 8 · 1983
The original case for visible, continuous, reversible direct manipulation over command-line abstraction. The patterns that move from prose-prompt to controllable surface are this argument applied to LLM I/O.
Related Patterns
Curated connections, not just neighbors.
Hand-picked relationships — what this pattern pairs with in practice, what it replaces, what it builds on, and what builds on it.
Alternative to
Parallel Alternatives
Tone of Voice tunes one draft by dragging on a 2D pad; Parallel Alternatives shows three drafts to compare, pick or blend.
Alternative to
Property Panel
The Property Panel gives each setting its own control; Tone of Voice moves formality and conciseness together in one drag.