Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 57 of the 57 references it cites were found in a registry or an index; 51 of the 106 sources its claims cite were checked for support.

Cleaner fish recognize self in a mirror via self-face recognition like humans

Open access

Abstract

Some animals have the remarkable capacity for mirror self-recognition (MSR), yet any implications for self-awareness remain uncertain and controversial. This is largely because explicit tests of the two potential mechanisms underlying MSR are still lacking: mental image of the self and kinesthetic visual matching. Here, we test the hypothesis that MSR ability in cleaner fish, Labroides dimidiatus , is associated with a mental image of the self, in particular the self-face, like in humans . Mirror-naive fish initially attacked photograph models of both themselves and unfamiliar strangers. In contrast, after all fish had passed the mirror mark test, fish did not attack their own (motionless) images, but still frequently attacked those of unfamiliar individuals. When fish were exposed to composite photographs, the self-face/unfamiliar body were not attacked, but photographs of unfamiliar face/self-body were attacked, demonstrating that cleaner fish with MSR capacity recognize their own facial characteristics in photographs. Additionally, when presented with self-photographs with a mark placed on the throat, unmarked mirror-experienced cleaner fish demonstrated throat-scraping behaviors. When combined, our results provide clear evidence that cleaner fish recognize themselves in photographs and that the likely mechanism for MSR is associated with a mental image of the self-face, not a kinesthetic visual-matching model. Humans are also capable of having a mental image of the self-face, which is considered an example of private self-awareness. We demonstrate that combining mirror test experiments with photographs has enormous potential to further our understanding of the evolution of cognitive processes and private self-awareness across nonhuman animals.

Overview

CoChat Summary

TL;DR

Cleaner fish that passed the mirror mark test recognized themselves in motionless photographs via self-face recognition, not kinesthetic visual matching. After mirror self-recognition, they reduced aggression toward self-face images but continued attacking unfamiliar faces, and most scraped their throat when shown a marked self-photograph.

Key findings

  • All focal cleaner fish (n = 10) passed the mirror mark test: throat-scraping occurred at 2.85 ± 0.40 SEM/h when marked with the mirror visible, versus 0.00 ± 0.00 in both control conditions (Friedman test, χ² = 20.00, P < 0.0001, Kendall W = 1.000).
  • Before mirror self-recognition, mirror-naïve fish attacked their own and unfamiliar photographs similarly: self-photograph 24.5 ± 9.5 SEM times/5 min; unfamiliar photograph 26.6 ± 11.5.
  • After passing the mark test, aggression declined toward self-face/self-body photographs (1.4 ± 0.3) and self-face/unfamiliar-body composites (2.1 ± 0.4), but remained high toward unfamiliar-face/unfamiliar-body (20.2 ± 4.6) and unfamiliar-face/self-body (19.0 ± 4.7) photographs.
  • In the photograph mark test, 6 of 8 mark-naïve fish (75%) scraped their throat after viewing a marked self-photograph; no fish scraped when viewing an unmarked self-photograph or a marked familiar-individual photograph (total throat scrapes: 22 for marked self, 0 for both controls).

Methods

  • Wild-captured cleaner fish, *Labroides dimidiatus* (26 fish total; 61–78 mm TL), were housed individually in laboratory tanks. Mirror mark tests used n = 10 fish; Experiment 1 used n = 10; Experiment 2 used n = 8; Experiment 3 used n = 8.
  • Fish were exposed to mirrors and tested with size-matched composite photographs: self-face/self-body (SS), unfamiliar-face/unfamiliar-body (UU), self-face/unfamiliar-body (SU), and unfamiliar-face/self-body (US). The mirror mark test used a brown visual implant elastomer mark on the throat; the photograph mark test placed a brown mark on a self-photograph.
  • Aggression and throat-scraping were video-recorded and analyzed with ELAN. Statistical tests included Friedman tests, exact Wilcoxon signed-rank tests with sequential Bonferroni correction, and a linear mixed model for swimming speed; a subset of videos was analyzed blind to assess observer bias.

Limitations

  • The paper states that advanced cognitive abilities and self-awareness in nonprimate animals, particularly fish, remain controversial; that the original mirror mark test was devised for primates and may yield false-negative results in other taxa; and that it remains untested how quickly the association between visual matching and a mental image of the self-face forms, and whether MSR is possible in the visual-matching phase or requires a mental image of self-face.
  • The paper acknowledges a possible sequence effect in Experiment 2, though it reports no significant order effect in Experiment 1 (negative binomial GLMM: treatment × order, χ² = 10.64, df = 10, P = 0.39; order, χ² = 6.29, df = 4, P = 0.18).
  • Evident from the text: sample sizes are small (n = 10, n = 8, n = 8; Experiment 3 had 2 of 8 fish classified as “unclear”), and the study uses captive, wild-caught fish under laboratory conditions. The claims about private self-awareness and mental states are inferential and described as speculation rather than directly measured.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1073/pnas.2208420120

Cites

57 of 57 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Fig. 1. Results of mirror mark test. The number of throat-scraping behaviors by cleaner fish in E1: period before treatment, E2: fish marked and mirror not visible, and E3: fish marked and mirror visible. ×3 = 3 individuals. Friedman test, χ52 = 20.00, P < 0.0001, and Kendall W (effect size) = 1.000. A and B show statistically significant differences by exact Wilcoxon signed-rank tests with sequential Bonferroni adjustment.
Fig. 1. Results of mirror mark test. The number of throat-scraping behaviors by cleaner fish in E1: period before treatment, E2: fish marked and mirror not visible, and E3: fish marked and mirror visible. ×3 = 3 individuals. Friedman test, χ52 = 20.00, P < 0.0001, and Kendall W (effect size) = 1.000. A and B show statistically significant differences by exact Wilcoxon signed-rank tests with sequential Bonferroni adjustment.CC BY · PubMed Central
Fig. 2. Cleaner fish photograph models used in experiment 1. (Cleaner fish photograph models used in experiment 1. (A) Faces of focal fish, and (B) an example of the photographic models used in the experiment. SS, self; UU, unknown fish; self-face/unknown body, SU; and unknown face/self-body, US.
Fig. 2. Cleaner fish photograph models used in experiment 1. (Cleaner fish photograph models used in experiment 1. (A) Faces of focal fish, and (B) an example of the photographic models used in the experiment. SS, self; UU, unknown fish; self-face/unknown body, SU; and unknown face/self-body, US.CC BY · PubMed Central
Fig. 3. Frequency of aggressive behavior directed toward photographic models in experiment 3. Self (SS) and unfamiliar fish (UU) models and composite self-face/unfamiliar body (SU) and unfamiliar face/self-body (US) models. Before MSR = before mirror presentation, after MSR = after passing the mark test. Means and SEM. (A) Total aggression. Friedman test, χ52 = 36.51, n = 10, P < 0.0001, and Kendall W (effect size) = 0.730. a and b show statistically significant differences by exact Wilcoxon signed-rank tests with sequential Bonferroni adjustments. (B) Aggressive behaviors: rushing (white), bumping (light gray), and biting (dark gray). Friedman test, χ52 = 3.75, n = 10, P = 0.59, and Kendall W (effect size) = 0.075; χ52 = 27.59, n = 10, P < 0.0001, and Kendall W (effect size) = 0.552; and χ52 = 35.05, n = 10, P < 0.0001, and Kendall W (effect size) = 0.701, respectively. (C) Changes in aggression toward photograph models over the 5 min experimental period.
Fig. 3. Frequency of aggressive behavior directed toward photographic models in experiment 3. Self (SS) and unfamiliar fish (UU) models and composite self-face/unfamiliar body (SU) and unfamiliar face/self-body (US) models. Before MSR = before mirror presentation, after MSR = after passing the mark test. Means and SEM. (A) Total aggression. Friedman test, χ52 = 36.51, n = 10, P < 0.0001, and Kendall W (effect size) = 0.730. a and b show statistically significant differences by exact Wilcoxon signed-rank tests with sequential Bonferroni adjustments. (B) Aggressive behaviors: rushing (white), bumping (light gray), and biting (dark gray). Friedman test, χ52 = 3.75, n = 10, P = 0.59, and Kendall W (effect size) = 0.075; χ52 = 27.59, n = 10, P < 0.0001, and Kendall W (effect size) = 0.552; and χ52 = 35.05, n = 10, P < 0.0001, and Kendall W (effect size) = 0.701, respectively. (C) Changes in aggression toward photograph models over the 5 min experimental period.CC BY · PubMed Central
Fig. 4. Frequency of aggressive behavior toward three composite photographs of self-face, familiar neighbor face, and unfamiliar stranger face, all placed on the body of another unfamiliar individual. Aggression toward self and neighbor photograph models after a 1 d absence from the focal fish’s territory. Means and SEM. (A) Total aggressions. Friedman test, χ22 = 16.00, n = 8, P < 0.001, and Kendall W (effect size) = 1.000. A–C show statistically significant differences by exact Wilcoxon signed-rank tests with sequential Bonferroni adjustments. (B) Aggressive behaviors: rushing (white), bumping (light gray), and biting (dark gray). Friedman test, χ52 = 3.90, n = 8, P = 0.14, and Kendall W (effect size) = 0.244; χ52 = 13.04, n = 8, P < 0.01, and Kendall W (effect size) = 0.815; and χ52 = 14.21, n = 8, P < 0.001, and Kendall W (effect size) = 0.888, respectively. (C) Changes in aggression toward photograph models over the 5 min experimental period.
Fig. 4. Frequency of aggressive behavior toward three composite photographs of self-face, familiar neighbor face, and unfamiliar stranger face, all placed on the body of another unfamiliar individual. Aggression toward self and neighbor photograph models after a 1 d absence from the focal fish’s territory. Means and SEM. (A) Total aggressions. Friedman test, χ22 = 16.00, n = 8, P < 0.001, and Kendall W (effect size) = 1.000. A–C show statistically significant differences by exact Wilcoxon signed-rank tests with sequential Bonferroni adjustments. (B) Aggressive behaviors: rushing (white), bumping (light gray), and biting (dark gray). Friedman test, χ52 = 3.90, n = 8, P = 0.14, and Kendall W (effect size) = 0.244; χ52 = 13.04, n = 8, P < 0.01, and Kendall W (effect size) = 0.815; and χ52 = 14.21, n = 8, P < 0.001, and Kendall W (effect size) = 0.888, respectively. (C) Changes in aggression toward photograph models over the 5 min experimental period.CC BY · PubMed Central

Verification

Claims the paper makes

23 Valid · 5 Not valid · 11 Not checked · 46 Own results and interpretations

34 more claims · 5 with their cited sources checked

Create a free account to see all 85 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 20
  • It may state two separate claims as one. 34

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

2 of 2

Citations

Cites

Show all 56 references

Research works printed but not linked: 1

  1. G. S. Losey G. L. Rosenquist J. L. Mahon J. Zamzow “Cleaning symbiosis” in A review in Behaviour and Conservation of Littoral Fishes Grutter (ISPA Lisboa Portugal ed. A 1999) pp. 379–395.

Cited by

See all 90

Metadata

Authors
Kohda, M., Bshary, R., Kubo, N., Awata, S., Sowersby, W., Kawasaka, K., Kobayashi, T., & Sogawa, S.
Affiliations
  • Osaka City University (JP)
  • University of Neuchâtel (CH)
Corresponding author
  • Kohda, Masanori
  • Sogawa, Shumpei
Venue
Proceedings of the National Academy of Sciences
Published
2023-02-06
Volume
120
Issue
7
Pages
e2208420120-e2208420120
ISSN
  • 0027-8424 (ISSN-L)
  • 1091-6490
OpenAlex
W4319303766
Cited by
90
References
56
Citations by year
2026: 8 · 2025: 45 · 2024: 21 · 2023: 16
FWCI
13.89
Type
journal article
Language
en
Open access
Locations
  • Proceedings of the National Academy of Sciences (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
Funders
  • MEXT | Japan Society for the Promotion of Science (20K20630, 19H03306, 17K18712)
  • Swiss Science Foundation (310030_192673 / 1)
Topics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Face Recognition and Perception
  • Life Sciences › Agricultural and Biological Sciences › Ecology, Evolution, Behavior and Systematics › Primate Behavior and Ecology
  • Social Sciences › Psychology › Developmental and Educational Psychology › Child and Animal Learning Development
Keywords
mirror self-recognition, self-face recognition, private self-awareness, cleaner fish, Labroides dimidiatus, animal cognition, cognitive evolution
MeSH terms
Facial Recognition, Animals, Behavior, Animal, Fishes, Humans, Self Concept, Recognition, Psychology

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-01, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-01 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Cleaner fish recognize self in a mirror via self-face recognition like humans. https://app.cochat.ai/papers/10.1073/pnas.2208420120
The paper
Kohda, M., Bshary, R., Kubo, N., Awata, S., Sowersby, W., Kawasaka, K., Kobayashi, T., & Sogawa, S. (2023). Cleaner fish recognize self in a mirror via self-face recognition like humans. Proceedings of the National Academy of Sciences. https://doi.org/10.1073/pnas.2208420120

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.