Journal: PLOS Biology

Article Title: Dog–human vocal interactions match dogs’ sensory-motor tuning

doi: 10.1371/journal.pbio.3002789

Figure Lengend Snippet: Word stream stimuli were first constructed by recording dog-specific command words (mostly disyllabic and monosyllabic, cf. small insert) that were appended into a 5-word stream with approximately 300 ± 50 ms silence intervals. These word streams were altered with regards to (1) speech type: by either removing content (reversed words) or prosodic information (flattened pitch modulation and reversed energy contour); and (2) speech rate: compression by a factor of 2 or 4; amounting to 9 word-stream conditions in total. The behavioural experiment consisted of an intelligibility scoring task for humans who listened to the full word stream, and of a playback task for dogs, who heard each word command separately (45 in total) a maximum of 3 times each, while the experimenter and the owner agreed on a behavioural response score. For the EEG experiment, dogs were first fitted with 1 to 4 electrodes covered by a headband and linked to an amplifier strapped on their back (photo inserts). They were then instructed to lie down and passively listen to an audio track (broadcasted via a speaker) containing 40 repetitions of each word stream condition. For comparability purposes, human EEG recordings were made under the same experimental conditions (see also ). Photo credit: E. Déaux. EEG, electroencephalography.

Article Snippet: EEG listening task : On the day of the EEG test, dogs were fitted with 1 to 4 golden cup electrodes (at least Cz and if possible, C3, C4, and POz) using gel and a conductive paste, and connected to a g.Nautilus amplifier (g.tec medical engineering GmbH, Austria) secured on the back of the dog, which wirelessly transmitted data to a receiver connected to a recording DELL laptop.

Techniques: Construct