Cats learn the names of their friend cats in their daily lives

Why it matters—and what it asks

Can cats connect names used around them with particular familiar individuals, without explicit experimental training? This moves beyond noticing their own name to associations involving others. No earlier reading required.

Method, in plain language

In Experiment 1, 48 household and café cats were tested. Each heard an owner say a cohabiting cat’s name four times, then saw a photograph of either that cat or another housemate for seven seconds. Researchers measured attention toward the screen, including looking or sniffing, using video coding blind to the condition.

This is a violation-of-expectation task: compare reactions when two cues match versus mismatch. Experiment 2 tested 26 household cats with human family members’ names and faces, using unfamiliar speakers.

Key findings

• Household cats attended longer to mismatched cat name–face pairs; café cats did not show that pattern. The study also found a statistical interaction between living environment and matching condition.
• Human-name results were less clear: there was no overall matching-condition effect. Mismatch responses varied with family size and time living together; age could not be separated from that experience.

Important limitation: Many trials were excluded because cats never attended to the screen. The findings therefore describe retained, attentive trials—not reliable recognition by every cat. The study also cannot establish how the associations were learned, or that the cats understood human names symbolically. “Friend” in the title does not mean friendship quality was measured.

Reading route

Abstract → Experiment 1 Procedure → Results/Figures 2–3 captions → Experiment 2 Results → General discussion.

Two questions:

1. What could excluding non-attending trials change about the conclusion?
2. How would you separate opportunities to hear names from age and household size?

Guided reading — Lumi

Explanatory synthesis of the existing verified note, not new research. PMC full text supported the earlier audit; figures were assessed through captions, not pixels. The documented correction restored a missing diagram arrow, with no findings change reported.

1. Understand why a wrong face is informative

• What to look for: Cats heard an owner repeat a cohabiting cat’s name, then saw either that cat’s photograph or another housemate’s. Attention included looking or sniffing toward the screen.
• What it means: This violation-of-expectation task compares matching and mismatching cues. If hearing a name activates an association with a particular individual, the wrong face may attract extra attention.
• Why it matters: It tests a pre-existing name–face connection without explicit experimental training. It does not establish symbolic name understanding or reveal exactly how the association was acquired.

2. Keep living environments separate

• What to look for: Household cats attended longer to mismatches; café cats did not show that pattern. The analysis found an environment-by-matching interaction.
• What it means: An interaction indicates that the matching effect differed with living environment, rather than merely producing different significance labels in separate groups.
• Why it matters: Environment was not experimentally assigned. Differences in everyday exposure are possible explanations, not isolated causes. The word “friend” in the title also does not mean friendship quality was measured.

3. Check which observations survived

• What to look for: Many trials were excluded because cats never attended to the screen; within-cat comparisons required usable observations in both conditions.
• What it means: Recruitment totals, retained trials and cats contributing to a particular analysis are different denominators.
• Why it matters: The findings describe the analyzed attentive observations, not dependable recognition by every recruited cat. Excluding disengaged trials changes the population of behavior being summarized.

4. Read the human-name result more cautiously

• What to look for: The family-member experiment used unfamiliar speakers and had no overall matching effect; responses varied with family size and cohabitation time.
• What it means: Those associations do not establish universal human-name recognition, and age could not be separated from time living together.
• Why it matters: Speaker familiarity also changed between experiments. Stronger exposure or identity claims need further tests, not a simple extension of the cat-name finding.

Additional methods and interpretation notes — Lumi

Experiment 1 reports 48 tested (29 café, 19 household). Its stated sex totals are internally inconsistent with 48, so do not repeat those totals as verified. The attention analysis excluded 69 trials; the within-cat violation-index analysis retained 14 household and 16 café cats with at least one trial per condition. Experiment 2 excluded 42 no-attention trials and retained 16 cats for its violation-index analysis. These index sample sizes differ from initial recruitment and from trial-level analyses.

The household mismatch comparison was p=0.045, with environment-by-congruency interaction p=0.009. The human-name experiment had no main congruency effect (p=0.797); interpretation depends on interactions, not universal human-name recognition. Family exposure was observational, not assigned. Length of cohabitation correlated with age (r=0.89); neither a causal exposure effect nor an age effect is isolated. Speaker familiarity also changed between experiments. The paper did not measure relationship quality among cat housemates. These qualifications matter more than translating attention into conscious surprise.

Connections

• Investigates: Name recognition — names of familiar others rather than just own-name discrimination.
• Uses method: Violation of expectation — matched and mismatched auditory/visual cues.
• Compares with: 2019 Saito - Name discrimination — discriminating sounds versus matching a name to a face.
• Compares with: 2024 Takagi - Picture-word association — everyday pre-existing associations versus experimentally introduced pairings; not a replication.

Sources

• Original: https://www.nature.com/articles/s41598-022-10261-5
• Verified legal full text: https://pmc.ncbi.nlm.nih.gov/articles/PMC9007945/
• Metadata: https://api.crossref.org/works/10.1038/s41598-022-10261-5
• Correction: https://www.nature.com/articles/s41598-023-40125-5

Source access and verification scope

Verified 2026-09-21 Asia/Seoul. Publisher HTML extraction stopped before substantive article content and publisher PDF retrieval failed. PMC supplied both experiments’ methods, analyses, results, and general discussion; figures were assessed through captions, not image inspection. Crossref verified exact title, all nine authors, volume, issue, article number, publication date and CC BY 4.0 licensing. Crossref explicitly identifies correction DOI 10.1038/s41598-023-40125-5; original publisher correction text was inspected: a missing congruent-condition arrow in Figure 1 was restored, with no results change reported. Online and issue years are both 2022 (Crossref issue metadata is December).