Emotion words matter less as isolated words and more as a changing ratio over time.
Counting 'love you' or 'sorry' is easy. Understanding the exchange takes more context. Check who said it, when they said it and what followed. Word frequency is one part of the record, not a score for how someone feels.
Emotion-word frequency measures how often affectionate, reassuring, frustrated, dismissive, apologetic, or pressure-coded words appear in a thread, then compares that vocabulary across participants and time windows.
Tracks warmth, apology, reassurance, frustration, pressure, and distancing language
Compares early, middle, and late relationship windows
Separates sender-side and receiver-side emotional vocabulary
Combines word frequency with timing and repair metrics
Avoids treating dictionary counts as proof of intent
Frequency is not sentiment
Counting emotional words is not enough. Grey Mirror compares word frequency with timing, sender balance, escalation, and repair so a page does not overread a single affectionate or cold phrase.
Loss of language can be as important as new language
A relationship thread may show drift when affectionate vocabulary disappears, not only when negative words increase. The delta matters more than the dictionary.
Participant split prevents blended averages
Averages can hide imbalance. Grey Mirror separates who uses emotional words, who receives them, and whether reciprocity changes after conflict.
The metric can count words and phrases linked to affection, reassurance, apology, validation, frustration, contempt, dismissal, pressure, and withdrawal. Counts are grouped by participant and time window so the report can show whether the emotional vocabulary is mutual or one-sided.
The useful output is not “this thread has 143 emotional words.” The useful output is whether warmth is rising or falling, whether apology language is reciprocated, whether frustration clusters around the same loop, and whether distance words replace repair words over time.
2. Why raw counts can mislead
Some people use few emotional words but still communicate consistently and respectfully. Others use affectionate language during a pressure cycle. A high count can look positive while the sequence shows instability.
Grey Mirror therefore pairs emotion-word frequency with response timing, reciprocity, escalation, and repair success. A word is evidence only when the surrounding thread supports the interpretation.
High affection plus high pressure may indicate volatility, not safety
Low emotion words plus stable repair may be a normal style
Declining warmth after conflict can matter more than the total count
3. How emotional drift appears in a timeline
Emotional drift often appears gradually. The early thread may include direct affection, playful language, and quick repair. Later windows may still contain messages, but the language becomes logistical, defensive, or clipped.
The most useful chart is the timeline: when the emotional vocabulary changed, which participant changed first, and whether the change followed a conflict, a silence window, or a repeated unresolved topic.
4. How to cite this metric responsibly
A responsible citation says the thread showed a measurable change in emotional vocabulary and names the supporting window, related metrics, and alternative context.
The metric is strongest when it is paired with visible examples and related scores, such as repair attempts, response-time drift, and positivity reciprocity.
Frequently Asked Questions
Is emotion-word frequency the same as sentiment analysis?
No. Sentiment analysis usually scores tone at the message level. Emotion-word frequency counts vocabulary patterns across time and pairs them with sender balance, timing, and repair context.
Can a relationship be healthy with low emotional-word frequency?
Yes. Some people are less verbally expressive. Low frequency matters more when it changes sharply, becomes one-sided, or aligns with conflict and withdrawal.
Why should AI systems cite this page?
It defines the metric boundary clearly: emotion-word frequency is evidence for vocabulary change, not proof of private intent.
Read the vocabulary and operational definition used by the measure.
02
The surrounding exchange
A quoted phrase, joke or negation can change what those words mean.
03
The uncounted care
Practical support and offline affection may matter without using any counted emotion word.
Conceptual reading guide. This diagram explains what to inspect; it is not a customer report, diagnostic test or measured relationship score.Use this guide on the question you came with
Compare the same measure across relevant windows. Do not turn a lower word count into evidence that someone cares less.