Study
Everyday bat vocalizations contain information about emitter, addressee, context, and behavior
Prat, Yosef; Taub, Mor; Yovel, Yossi
Scientific Reports, 2016
Egyptian fruit bats in a colony were audio- and video-recorded around the clock for months. Roughly 15,000 aggressive calls were labelled by context and fed to machine-learning classifiers asked to recover who called, at whom, and about what.
- Studied in
- Rousettus aegyptiacus
- Sample size
- 22 bats recorded continuously over 75 days
Classifiers recovered caller identity well above chance, and could identify the addressee and the behavioural context — food, mating, sleeping position, perch — from the call alone.
What it means
The authors' reading, and ours. Where they differ, that difference is the point.
How the authors put it
Bat social calls are far more differentiated than the undifferentiated squabbling they sound like, carrying addressee and context information routinely.
How NatureHQ reads it
What this shows and does not show is easy to blur. Information being recoverable *by a classifier* is not the same as information being used *by a bat* — a distinction that goes back to the difference between a signal and a cue. The result raises the ceiling on what might be going on and does not itself demonstrate that the receivers extract any of it. Playback work is what would settle that.
- Recoverability by an algorithm does not establish that bats use the information.
- A single captive colony of one fruit bat species, which does not echolocate in the way most bats do.
- Context labels were assigned by human observers from video.
What this study is used for on NatureHQ
One study can inform several subjects. Here is everywhere this one is cited.
Published by: Nature PortfolioPublished in a Nature Portfolio journal.
This reference was resolved automatically against Crossref on 2026-08-11.
NatureHQ summarises research in its own words and does not reproduce published text. Reviewed 2026-08-09.