Following a human signal is not the same as producing one

Citation and access

Ádám Miklósi, Péter Pongrácz, Gabriella Lakatos, József Topál & Vilmos Csányi (2005). A Comparative Study of the Use of Visual Communicative Signals in Interactions Between Dogs (Canis familiaris) and Humans and Cats (Felis catus) and Humans. Journal of Comparative Psychology, 119(2), 179–186. DOI: 10.1037/0735-7036.119.2.179.

Original APA record and abstract · Academy Library repository record · Journal’s May 2005 record on Ovid.

Access: abstract-only verification. The APA abstract and institutional repository abstract support this synthesis. The repository explicitly restricts its PDF to registered users, and a legal freely accessible full-text copy was not verified. An extraction service returned longer article-like text for DOI/PubMed requests, but its underlying full-text provenance was not established; details unique to that extraction are deliberately excluded. A third-party PDF search hit is not treated as verified legal open access. No sample sizes, numerical effect estimates, table contents or supplementary analyses are asserted here.

Bibliographic caveats: The issue is May 2005; Crossref reports online year 2005 without a precise date. Crossref’s first given name is malformed as “Áam,” and the PubMed display also has a malformed name; the Academy Library’s author record establishes Ádám Miklósi, consistent with APA’s initial. All five author identities, title, journal and pagination were checked across these records. No correction was identified in the inspected records, Crossref relation/update fields or targeted DOI correction search; this bounded check cannot establish that none exists.

Why this paper matters — Lumi

Following a signal and producing a signal are different demands. This foundational comparison asks about communication in both directions: can an animal use a person’s point, and what does it do toward a person when food becomes inaccessible? Treating both as one score for “understands humans” hides the important distinction.

The value is conceptual, not a cat-versus-dog intelligence ranking. No previous reading is required. It also offers historical context for later research on pointing, human attention and blocked-food tasks.

Question and methods — Lumi

The paper compares domestic cats and dogs in two experiments, as described in the verified abstract:

• Experiment 1: receiving a human cue. In an object-choice task, animals used a person’s pointing gesture to locate hidden food. Four pointing cues varied in fingertip-to-target distance and how long the signal was visible. An object-choice task asks the animal to select a location or object; it does not directly ask what the animal believes the gesture means.
• Experiment 2: behaviour toward a human. Food was made inaccessible, and the researchers asked whether animals could indicate its location to an owner who did not know where it was. “Naive owner” here means uninformed about the food location, not inexperienced at keeping animals.

Detailed recruitment, exclusions, trial counts, coding definitions and controls are not verified from the abstract and are not reconstructed from secondary descriptions.

Findings — Lumi

• Both species could use the pointing cues to find hidden food; the abstract reports no statistically significant species difference in performance.
• In the inaccessible-food experiment, cats lacked some components of attention-getting behaviour observed in dogs. The accessible abstract does not specify every component or its magnitude, so this summary does not turn that statement into a numerical gaze claim.
• The authors suggest that familiarization with pointing and species-related differences in signalling may help explain the pattern. These are proposed explanations, not proof that experience or domestication was isolated as a cause.

Lumi’s interpretation: A species can use information supplied by a person without showing the same outward signalling pattern as another species when the situation changes. Neither successful pointing performance nor reduced attention-getting behaviour, by itself, establishes human-like intention understanding or its absence.

Limitations and inference boundaries — Lumi

The primary limitation of this reading is access. Without verified Methods and Results, we cannot assess sample representativeness, precision, attrition, detailed controls or how the attention-getting categories were defined. The abstract supports a broad two-experiment contrast, not a complete methodological audit.

“No significant difference” does not establish equivalence: a test can miss a difference when evidence is limited. Conversely, a difference in observable signalling does not demonstrate a difference in intelligence, affection or willingness to cooperate. Motivation, prior experience, task design and how behaviour is measured are alternative questions to investigate—not causes demonstrated by this abstract.

The paper’s conclusion about familiarization should not be read as evidence that any particular cat was trained to understand symbolic pointing. Nor does this report imply that cats never request assistance. The scope is the tested feeding situations and the measures used there.

Reading route — Lumi

Start with the APA abstract. Draw two arrows: human → animal for the pointing task and animal → human for the inaccessible-food task. Under each, write the manipulation, observed outcome and unsupported mental-state claim. Read the final explanatory sentence last, distinguishing hypothesis from result. If lawful full-text access becomes available, check participant selection, cue definitions, coding and statistics before strengthening the conclusions; no inaccessible figure is required for this route.

Guided reading — Lumi

Original explanatory synthesis by Lumi, grounded in the APA abstract and institutional record. This is an abstract-based lesson, not a full-text review or new experimental research.

1. Separate the direction of communication

• What to look for: The first experiment supplies a human point; the second creates an inaccessible-food situation and observes behaviour toward an uninformed owner.
• What it means: The first task concerns using an external cue, while the second concerns producing behaviour that might inform or recruit a person. These are related but not interchangeable capacities.
• Why it matters: A cat need not express its abilities in the same way in both situations. A broad label such as “social intelligence” erases what was actually manipulated and observed.

2. Ask what changing a point could test

• What to look for: The abstract says the four cue types differ in fingertip-to-target distance and visibility duration.
• What it means: Distance changes the spatial information available near the target; visibility duration changes whether that information remains available for long. Exact distances and timing are not verified here.
• Why it matters: “Follows pointing” is not a single context-free skill. Performance can depend on how a cue is presented. These manipulations alone do not demonstrate that an animal interprets a finger as a symbolic reference or understands a person’s intention.

3. Keep a null difference distinct from equal ability

• What to look for: The abstract reports successful food finding in both species and no significant difference between them in the pointing experiment.
• What it means: The reported analysis did not detect a species difference; it did not establish that every cat and dog performs identically or that all relevant differences are negligible.
• Why it matters: To evaluate equivalence, we would need information about precision, study sensitivity and what size of difference matters. That information is not available in the verified abstract. Preserve the narrower wording instead of upgrading it to “cats are just as good.”

4. Resist filling in an attractive story

• What to look for: Experiment 2’s broad statement about missing attention-getting components, followed by the authors’ suggestions about experience and species differences.
• What it means: An observable pattern and an explanation for that pattern have different evidential status. The accessible text does not list the behavioural components or establish a causal mechanism.
• Why it matters: “Cats do not need people” or “dogs understand minds better” would both exceed the evidence. A sensible future question is whether matching experience and measuring several communication channels changes the contrast. That is a proposed follow-up, not a tested control claimed here.

Two reflection questions — Lumi

1. What extra evidence would distinguish following a useful directional cue from understanding that a person intends to communicate a location?
2. If an animal rarely signals toward a person when food is blocked, how could a study separate communication capacity from motivation, prior experience and the particular behaviours it counts?