DEV Community

Cover image for Interpreting silence and distance without deciding for someone else
LOGIHEART
LOGIHEART

Posted on

Interpreting silence and distance without deciding for someone else

Social capabilities across human development | Part 3

People communicate through gestures, face direction, tone of voice, and where they stand. Yet a single signal cannot reveal everything about another person's mind. Nonverbal social capability includes noticing signals, checking interpretations, and prioritizing the person's stated preference.

Should you approach someone who steps back?

Two colleagues are examining a document. As one moves closer while explaining, the other steps back. The speaker starts to follow, then stops and asks, "Is this distance comfortable?" The colleague nods, and they continue from there.

This is a fictional scene. Stepping back may reflect a preferred distance, a bag, visibility, or conditions in the walkway. There is no need to decide that the person dislikes you, or to keep approaching because they are "only being polite." An interpretation can remain provisional while you check with the person.

Research and the range of interpretation

Blais and colleagues reported differences associated with cultural background in how participants looked at faces. That finding concerns face viewing in an experiment; it does not show that nationality predicts someone's preferred everyday distance. See Culture shapes how we look at faces.

Barrett and colleagues' review examines variation and context when inferring emotion from facial movements. It provides a basis for questioning fixed meanings assigned to smiles, gaze, and silence. See Emotional Expressions Reconsidered.

These sources do not justify prescriptions about how many centimeters someone from a particular culture should stand away. Learning through development and new relationships that one's habits may not suit another person is a broader question. The specific adjustment described here is this article's analysis of an everyday scene.

Before filling a silence

Silence may reflect agreement, confusion, thinking, or a wish to end the conversation. A missing reply should not be treated as permission to expand the conversation or touch someone. Distinguish situations requiring a brief clarification from those calling for time to think.

What can be observed is an action: stepping back, raising a hand, or replying. Separating observation from inference avoids overriding the person's account with "You are angry." If they say the interpretation is wrong, reflect that correction in the next approach or distance.

A design hypothesis for physical AI

For embodied AI, an incorrect distance can lead to an obstructed walkway or unwanted approach, as well as conversational discomfort. One possible design conditions movement on both physical safety and the person's expressed preference.

An inferred emotion should not itself authorize contact or approach. Even if a model classifies someone as friendly, do not follow when they step back. Update the current action when their distance preference changes. When clarification is unavailable, maintain distance within safe conditions.

Evaluation could include identical expressions with different preferences, different expressions with the same preference, corrections, and changed distance requests. The proposed focus is whether unwanted approaches stop, alongside facial classification accuracy.

A decision in four panels

Fictional four-panel example: a colleague approaches too closely, notices the other person stepping back, stops, and continues the discussion at a comfortable distance.

  1. Someone moves too close while explaining a document.
  2. The other person steps back.
  3. The speaker stops and restores space.
  4. They continue with a comfortable gap.

This fictional comic shows adjustment in response to another person's signal, rather than one universally correct distance.

References

  1. Culture shapes how we look at faces
  2. Emotional Expressions Reconsidered

About this series

This series uses published literature to examine everyday situations. Its 7 domains and 22 capabilities are an editorial framework, not a standard academic classification. Differences in age or culture are not rankings of people's social worth or capability. The proposed AI designs do not establish humanlike emotional experience or demonstrated product performance. The text was created with AI assistance; the comics are AI-generated fictional examples.

LOGIHEART's research and development vision

LOGIHEART is researching how to implement social capabilities in physical AI that interacts with people.

Explore the research and development vision at the LOGIHEART website and in the English introduction video.

This post was drafted with AI assistance.

Top comments (0)