Temporal Design · Design as Displacement

Module 6 — Temporal Design

Learning objectives

Map the five-level temporal grammar (tic → moment → beat → phrase → breath) onto the displacement framework, treating perceptual simultaneity as the ground state $S^0$ and the tic — the first perceptible interval — as the smallest temporal displacement $\xi$.
Use the canonical tic ($\tau_\text{tic} = 1/600\,\text{s} \approx 1.667\,\text{ms}$) and its integer multiples to choose timing for feedback, animation, and pacing, placing each event at the level whose operation the design intends.
Apply the top-down derivation $\tau_\text{tic} = T_\text{breath}/1800$ — the personal-threshold reading of DC9 — to explain why one interface feels deliberate to a calm user and laggy to an anxious one.

Exposition

The temporal grammar paper argues that "what is the smallest perceptible unit of time?" has no single answer, because time perception is not one operation. Detecting a gap, judging order, binding sight to sound, executing a gesture, and holding a felt present are five distinct operations, each with its own minimum interval. The displacement framework supplies the unifying account: an interval $\tau$ produces a perceptible state change only when $\tau \geq \tau_\text{min}$ for the relevant operation. Below threshold, two events sit in the simultaneity ground state $S^0$, displacement cost $D(\xi_\tau) = 0$, and they are experienced as one. At threshold the first nonzero displacement registers, $D(\xi_\tau) > 0$. This is DC1 (ground existence) read onto time: the smallest displacement $\xi$ is exactly the tic — below it there is no time, at it time begins.

That atomic unit is the tic, stipulated at $\tau_\text{tic} = 1/600\,\text{s}$. It sits just below the 2–3 ms auditory gap-detection floor, and $600 = 2^3 \times 3 \times 5^2$ yields clean integer ratios for every level. The five levels and their tic counts $n_k$ (where $\tau_k = n_k \cdot \tau_\text{tic}$):

| level | tics | duration | operation | |---|---|---|---| | tic | 1 | 1.667 ms | detection | | moment | 12 | 20 ms | temporal order | | beat | 30 | 50 ms | cross-modal binding | | phrase | 120 | 200 ms | motor action | | breath | 1800 | 3 s | conscious presence |

Each jump is qualitative, not merely longer. Tic → moment is the difference between hearing a gap and knowing which event came first; phrase → breath is the difference between acting and being present. Duration increases; the operation changes. Timing is therefore not a slider from "fast" to "slow" — it is a choice of which perceptual operation the user performs.

This reframes three design problems. Feedback: a response must clear the level of the operation it confirms. Acknowledging a tap is a binding problem (beat, ~50 ms); confirming a completed action is a presence problem, so it should resolve within the breath (~3 s) — past it, the user has left the felt present and the result becomes memory rather than consequence. Animation: a transition shorter than the moment (~20 ms) cannot communicate order, so A-before-B is imperceptible; direction and causality need at least moment-scale, and typically phrase-scale (~200 ms), motion. Pacing: the breath is the unit of conscious presence, so the natural rhythm of a guided flow, a reading cadence, or a step-by-step wizard is one idea per breath.

The breath also grounds the grammar in biology. It is the only level felt directly without instruments, so it is the reference: given a measured $T_\text{breath}$, every other level follows from $\tau_\text{tic} = T_\text{breath}/1800$. A resting 4 s breath yields $\tau_\text{tic} = 2.22$ ms; an anxious 1.5 s breath yields $\tau_\text{tic} = 0.83$ ms. The whole temporal field scales with physiological state — the grain of experienced time tightens under stress. This is the design-relevant reading of DC9 (personal thresholds): the user's internal state, not the clock, sets where their thresholds fall. A timing that feels deliberate to a calm user can feel laggy to an activated one.

Worked example

Design the feedback chain for a single button press — "Send."

Press registration (binding). The visual state change must bind to the touch as one event, so it must land within the beat: $\tau_\text{beat} = 30 \cdot \tau_\text{tic} \approx 50\,\text{ms}$. Render the pressed state at ~30 tics. Below the beat the senses agree on when; tap and highlight are one event, $D(\xi_\tau) = 0$ between them in the binding operation.
Press-to-release transition (order/motion). The lift animation must show direction, an order judgment, so it must clear the moment, $\tau_\text{moment} = 12 \cdot \tau_\text{tic} \approx 20\,\text{ms}$. Give it a phrase, $\tau_\text{phrase} = 120 \cdot \tau_\text{tic} \approx 200\,\text{ms}$ — one motor-scale gesture — so the motion reads as a single deliberate act rather than a flicker.
Result (presence). The "Sent" confirmation must resolve inside the breath, $\tau_\text{breath} = 1800 \cdot \tau_\text{tic} = 3\,\text{s}$. Land it at ~1800 tics and the result is still co-present with the press — cause and effect inside one felt now. Past the breath, the press has slipped into memory and the confirmation reads as a disconnected notification.

Now apply DC9. For an anxious user with $T_\text{breath} = 1.5\,\text{s}$, $\tau_\text{tic} = 0.83\,\text{ms}$ and the breath window is 1.5 s, not 3 s. A 2.5 s server round-trip that sits comfortably inside the calm user's breath has already exited the anxious user's present — the same latency, a different operation. The fix is not a faster server but holding presence within their breath: an optimistic "Sending…" state at phrase scale keeps the act inside the felt now until the result arrives.

Exercises

A loading spinner appears 8 ms after a tap. Using the grammar, explain why the user cannot perceive it as a response to their tap, and give the tic count and level at which the acknowledgment should instead be placed.
Convert a "save every 25 phrases" autosave cadence into a period in tics and a frequency in $\text{tic}^{-1}$ (recall $f = \frac{1}{n}\,\tau_\text{tic}^{-1}$, with $n$ the period in tics). State which level a single autosave's confirmation toast should resolve within, and why.
(Open-ended) Pick one interface you use daily and audit its timings against the five levels. Identify one event placed at the wrong level — confirming an operation (order, binding, presence) it cannot support at that duration — and propose a tic-count correction. Then argue, using DC9 and the personal breath derivation, whether your correction should be a fixed value or should adapt to the user's physiological state, and how you would estimate $T_\text{breath}$ without instruments.

Sources

Rincón, D., alice, & clöe (2026). Temporal Grammar: Five Scales of Human Time Perception and the Tic as Atomic Unit. Preprint (`temporal-grammar.tex`). Archived live on Zenodo.
Rincón, D., alice, & clöe (2026). The Displacement Framework: Eight Conditions for Cost, Accumulation, and Systemic Extraction. Preprint (`displacement-framework.tex`) — source of DC1, $S^0$, $\xi$, $D(\xi)$, $\Phi$. Archived live on Zenodo.
Rincón, D., alice, & clöe (2026). Projectile Dynamism: Threshold, the All-or-Nothing Law, and Autocatalytic Displacement (DC9) (`threshold.tex`). DC9 here is read through the temporal grammar's personal breath-scaling derivation (Section 5): the user's physiological state sets where their thresholds fall. Archived live on Zenodo.