Blog / Cognitive and Affective Empathy: Understanding, Resonance, and Response Are Different Evidence

Cognitive and Affective Empathy: Understanding, Resonance, and Response Are Different Evidence

Inferring another person’s feelings, sharing an emotion, and responding helpfully require separate records.

October 6, 2026 · 6 min read · By Cepingku Editorial Team

‘I understand why my colleague is upset’ and ‘I feel upset when I hear about it’ are often placed under the one label empathy. They are related possibilities, not the same evidence. Kogler and colleagues’ neuroimaging meta-analyses operationalised cognitive empathy as taking another person’s perspective in emotional or painful scenarios and affective empathy as self-referentially feeling another person’s emotion or pain. Those definitions identify what the analyses aimed to distinguish. They do not provide a universal ruler for deciding that a particular person simply does or does not have empathy.

Someone may make a careful inference about another person’s concern without immediately sharing that emotion. Someone else may feel intense distress while still misunderstanding what the other person actually needs. Neither an inference nor emotional resonance is a record of helpful behaviour. Keeping these claims apart gives a more accurate account of an interaction and prevents a broad personality label from replacing the details.

Record to keep separateWhat it actually shows
Understanding or inferenceHow a person explains another person’s possible feeling, situation, or concern.
Emotional resonanceWhether the observer has their own emotional response to another person’s emotion.
Actual responseWhat was said, asked, or done in this particular situation.

A fictional team scene can produce three different records

This is a fully fictional team scene, not a test item, brain experiment, or conduct standard. After a weekly meeting, Lin’s project proposal is redirected at the last minute. She says, ‘Fine, I will revise it tonight,’ in a clipped voice, then closes her laptop without asking her usual follow-up questions. Another teammate, Arun, notices the change.

Arun’s understanding can be written as a revisable hypothesis: ‘Lin may be worried that revision will disrupt tomorrow’s client presentation, or that the reason for the change is still unclear. I do not yet know which concern is closer to her view.’ This is not mind-reading. It is an inference from limited cues that calls for a check: ‘How does this change affect tomorrow’s presentation or your schedule? Should we confirm the priority first?’ A more accurate inference requires Lin’s opportunity to correct it.

Arun may also experience resonance: Lin’s tone might leave him feeling tense, frustrated, or uneasy. That is Arun’s emotion, not proof that he has accurately identified Lin’s experience. Conversely, he may feel little emotional spillover and still pause, ask a clear question, restate the conflict he sees, and leave Lin room to decide whether to discuss it. Emotional resonance and understanding may coexist, but the scene does not let one certify the other.

Only then comes the response record. Did Arun stop packing up, ask a question, create unhurried time, help reorder the work, or merely say ‘that sounds hard’ and leave? Whether a response was helpful depends on Lin’s actual need, the team’s authority, and what happened next. It cannot be inferred automatically from the fact that Arun felt concern.

Why a self-report cannot become proof of a live interaction

A questionnaire item such as ‘I usually understand how other people feel’ first collects a respondent’s description of their own usual tendency. It cannot demonstrate that Arun correctly inferred Lin’s concern in this meeting, nor can it document whether he asked, listened, or helped. A high answer may be sincere and still be shaped by the item wording, the time frame the respondent imagines, and their self-view at the moment of answering.

The cited neuroimaging meta-analyses synthesise findings from prior studies. Those results are research background, not evidence that a web self-report produces a particular pattern of brain activity. They likewise do not transfer findings from the studies included in the meta-analyses, neural findings, or claims about helping behaviour to this site’s Emotional Intelligence Assessment.

Replace one broad label with three narrow questions

For a later reflection, ask three separate questions. Which observable cues led me to infer that the other person might care about something? What emotional response did I have myself? What did I actually do, and how did the other person respond? The answers may line up, or they may reveal a useful mismatch. Each result points to a different next step: clarify an assumption, regulate one’s own reaction, or change the practical response.

Suppose Arun later learns that Lin’s main worry was not evening work but unstable requirements. That does not prove that he lacks empathy. It shows that his earlier hypothesis missed the concern and that a more specific question would help next time. The Emotional Intelligence Assessment can prompt reflection on the self-reported answers given here. It should not be interpreted as proof that someone has read another person accurately, produced a specific brain response, or demonstrated helpful behaviour.

Want to see where you land?

Take the Emotional Intelligence Assessment and get your results in minutes.

Take the Emotional Intelligence Assessment →

← Back to the blog