How the editorial groups were assigned

CrowLingo has not run NatureLM-audio[6], , or on this corpus. The group labels are editorial, informed by descriptive vocabulary in the literature, and recording assignments use source metadata. A measured research map would require audio embeddings, documented clustering choices, a source and listening review, and independent behavioral validation. That proposed workflow is described in the pipeline pages; it is not the provenance of the current atlas.

Long-duration caws emitted from a perch, typically by a single bird, faced outward across a territory boundary.

Territorial caw

Long-duration caws emitted from a , typically by a single bird, faced outward across a territory boundary. The acoustic profile is one of declaration: long durations, regular spacing, low spectral roughness. Carries caller identity, sex, and approximate intent. This is the cluster that most directly maps onto the canonical caw the broader public hears in their neighborhoods. The atlas position does not measure its distance from mobbing calls.

Mobbing alarm

Compressed, urgent caws delivered in rapid sequences, often by recruited pairs or groups targeting an aerial predator. Spectrally rough, tightly packed in time. Kevin McGowan's Cornell field studies catalogued the recruitment chains: one bird sees, calls; three arrive; ten more follow. Some flocks remember and re-mob the same individual predator for years. Its position in this atlas is illustrative and does not establish acoustic separation by call rate.

Assembly calls

This editorial group covers source recordings described as gatherings or roosts. Their placement in the atlas does not show that a model detected a shared acoustic structure or a particular behavioral function. Read each recording’s source attribution for the available context.

Rattle

The rattle is the strangest sound in the American crow repertoire — mechanical, almost prehistoric, weakly harmonic. Rattles appear in affiliative, recruitment, and occasionally territorial contexts, with high individual variation. Pairs and family groups seem to use them to signal something closer to mood than position. No comparative analysis underlies this group’s position in the atlas.

Juvenile begging

Higher-frequency, narrower-band calls from juveniles soliciting feeding. Diagnostic spectral signature: tightly clustered, easy to identify. The Marzluff[4] lab's parent-offspring work showed begging calls are surprisingly individual; chicks recognize their own parents' return calls within weeks of fledging. Its proximity to other groups in the current atlas does not imply a measured acoustic relationship.

Companion calls

The companion group is an editorial home for archival examples assigned from source metadata. Neither caller relationships nor communicative function have been independently validated by CrowLingo. A Wikipedia article-introduction file previously included here has been excluded from the active corpus pending source and listening verification.

Quiet grunts and loud grunts

These two editorial groups currently have no field recordings in the active corpus. Their pages retain written introductions, but their atlas positions and context weights are illustrative. The labels do not demonstrate that a model discovered a quiet-versus-loud separation.

The exceptional cluster

The exceptional group is an editorial category for atypical source descriptions. It is not a model-generated outlier class, a demonstrated dialect, or evidence of a new call type. Additional recordings and independent review could change how these examples are organized.

What these groups do and do not establish

The nine groups are a presentation choice, not a scientific contribution establishing new vocal categories. Their relationship to published taxonomies needs source-by-source review; this site has not demonstrated that the grunt labels or exceptional group were recovered by an AI model. Use the archival recordings to explore the material, and the cited papers to assess research claims.