Overview
- Researchers scanned awake, trained pet dogs in fMRI machines while the animals viewed photos of unfamiliar people showing happiness, anger, fear, and sadness.
- Smiling faces produced stronger activity in a temporal cortex–caudate network, a circuit linked to higher visual processing and reward, suggesting smiles have positive value for dogs.
- Whole‑brain machine‑learning analyses found distinct neural patterns that separated fear from anger and sadness but did not reliably separate anger from sadness.
- The study used small samples of companion animals (eight dogs in the first experiment, 12 in the second) with many border collies and relied on static, unfamiliar photos, so the authors warn results may not generalize to other breeds or real‑world, multimodal cues.
- Authors frame the findings as consistent with domestication shaping dogs' sensitivity to human social signals and say follow‑ups will test multimodal stimuli, broader dog populations, and direct human–dog brain comparisons.