27/07/2026
Nice work here. From the paper: "These results produce a paradox at the heart of the relationship-to-production circuit. The sequences where Indy shows genuine stress are also those where his behavior is most ethologically legible—a real dog communicating real discomfort through scientifically documented signals. The film’s welfare deficit and its representational authenticity share the same source. Documentary cinema has long faced an analogous tension: it presents itself as a faithful record of the natural world, yet the authentic emotion it puts on screen is frequently manufactured. Scenes staged through baiting or chumming, and narration that overlays anthropomorphic affect in place of accurate ethological description [14]. What audiences receive as observation is often construction, blurring the documentary’s implicit claim to show things as they are [51]. What distinguishes the animal case is the absence of consent, although recent work in animal welfare has begun to explore forms of animal assent, such as opt-in/opt-out protocols and consent training [22]. One actionable step would be to embed an independent welfare professional—a certified veterinary behaviorist or applied animal ethologist—within animal-centric productions, with the authority to pause or modify filming. Such oversight could be triggered by predefined behavioral thresholds rather than left to directorial judgement: the co-occurrence of stress indicators we observed in the shower scene (lip-licking peaking at high rate alongside avoidance) illustrates the kind of convergent micro-signal that could serve as a stop criterion, though an operational threshold would need to be established on multi-subject data rather than derived from a single case. This would shift welfare from a matter of filmmaker goodwill to a monitored, enforceable condition of production. A human subject can agree to the terms of observation; a dog cannot. This shifts the ethical burden entirely onto producers and regulators—and current frameworks remain oriented toward preventing overt physical harm rather than monitoring affective states [9,10]. But even if such frameworks were in place, Good Boy’s artisanal model—a two-person crew, a 400-day timeline, the dog’s own home—cannot scale to industrial production, and what welfare-responsible animal filmmaking looks like under commercial conditions remains an open question."
"On the other hand, the dominant mode of audience engagement remains anthropomorphic projection. Viewers read Indy’s proximity to Todd as loyalty and his behavior as devotion. The framing is emotionally powerful, but ethologically misleading. Dogs do form selective attachment bonds with their caregivers including proximity-seeking, separation distress, secure-base effects [23,24,63]. But attachment is not loyalty. Loyalty implies conscious commitment; the behaviors the film presents as devotion are better understood as products of a bonding system shaped by domestication. Framing them as heroic choice reinforces anthropomorphic readings that can distort public understanding of what dogs need [13]. Worse, this framing hides a welfare problem: a dog staying close to an unpredictable caregiver in a threatening environment is not demonstrating devotion—it is likely exhibiting coping strategies or submissive behaviors linked to significant stress [39,64]. The loyalty narrative makes this invisible.
Strikingly, not a single review in the corpus mentions nose licking, gaze aversion, or any other subtle indicator of stress. Yet these signals are visible in the film. Where the audience sees a devoted dog, behavioral analysis sees a dog in discomfort. Observers read most reliably the emotion a dog displays most overtly, and that emotion is happiness: it is correctly identified around 88% of the time [65]. Negative states fare far worse—even fear, an overt emotion, is recognized only about 45% of the time and is routinely confused with surprise [65]. Subtle, early signals of stress are missed almost entirely [66]. Nose licking, for instance, is identified as a stress indicator by fewer than 5% of owners [67]. Owners also tend to overestimate their ability to read canine behavior [65] and everyday familiarity with a pet does not, by itself, yield accurate detection of early stress signals without explicit education [67]. The film therefore does two things at once: its editing reinforces anthropomorphic projection, while its unscripted footage records authentic behavioral material that could undermine it. It does reach its audience and shift how they relate to dogs—but what they take away is filtered through anthropomorphic projection rather than ethological understanding, because most viewers cannot read what the dog is communicating."
Western societies increasingly treat dogs as sentient family members, but does this reshape cultural production, and can cultural production, in turn, reshape the relationship? We traced this circuit through Good Boy (2025), a horror film with a real, untrained dog as the sole protagonist, combining...
