Feedback cost determines the form and frequency of confirmation

What it is

Grounding costs are the joint effort of presenting, receiving, accepting, and repairing a contribution. A medium's support for co-presence, visibility, audibility, cotemporality, simultaneity, sequentiality, reviewability, and revisability changes which evidence is inexpensive. A face-to-face nod barely interrupts speech; “received” can consume a turn in text conferencing; email lacks immediate feedback but permits review and revision.

Why it happens

People choose evidence by balancing the grounding criterion against combined effort, captured by the principle of least collaborative effort. When a medium lacks one constraint, another action compensates: write a location when pointing is invisible, quote a message when interruption is unavailable, or repeat crucial content when it cannot be reviewed. A channel therefore does not simply produce more or less confirmation; it redistributes formulation, production, comprehension, delay, and repair costs between participants.

Studying it

Dyads can perform the same referential task face to face, by voice, in synchronous text, and in asynchronous text. Code acknowledgment forms and intervals, utterance length, repair turns, errors, and both partners' time. A comparison must specify actual constraints—whether video includes the shared object, text streams character by character, or messages are editable—because channel labels alone do not explain outcomes. Joint cost must include both partners rather than treating effort shifted to the addressee as a gain.

Where it stops holding

The cheapest evidence may not satisfy a high-risk task, and organizational rules may require redundant readback. Available capabilities are not necessarily usable: an enabled camera may not show the referent, and editable messages may weaken record stability. Language proficiency, accessibility needs, and power relations redistribute effort, so mean completion time can hide an unequal burden.

Applying it

• Inventory the medium constraints required by critical exchanges instead of substituting synchronous/asynchronous labels for actual capabilities.
• Provide channel-appropriate evidence: lightweight reactions, precise quotation, object anchors, readback structures, and visible revision history.
• Raise the evidence criterion for consequential content without forcing explicit acknowledgment of every low-risk message.
• Compare joint time, repair turns, and errors on the same task, and inspect sender and addressee effort separately.

Related

• Same group: V1.03.1 Common ground is mutually available knowledge · V1.03.2 Grounding requires reciprocal evidence · V1.03.4 Insufficient common ground produces divergent interpretations · V1.03.5 Experts' omissions do not work for newcomers
• Adjacent: V5.01 Media richness · V5.08 Text communication
• Search terms: grounding costs · least collaborative effort · media constraints