{"brand":"CrowLingo","publisher":"Kymata Labs","url":"https://crowlingo.org","contact":"contact@kymatalabs.com","description":"AI-powered animal language processing for the American crow — methods, decoding, ethics, and the practical pipeline from caw to cluster.","keywords":["American crow","Corvus brachyrhynchos","crow vocalization","crow communication","crow cognition","crow tool use","crow face recognition","Marzluff mask experiments","corvid intelligence","corvid behavior","cooperative breeding crows","syrinx","crow syrinx anatomy","how crows make sound","bird two-source vocal production","crow frequency range","crow caw frequency","crow harmonic structure","bioacoustics","computational bioacoustics","deep learning bioacoustics","BirdNET","BirdNET embeddings","Perch model bioacoustics","PaSST audio embedding","PANNs audio","AudioMAE","self-supervised audio","SSL audio model","masked spectrogram prediction","audio foundation model","NatureLM-audio","Earth Species Project","ICLR 2025","BEANS-Zero benchmark","audio-language foundation model","latent space","audio embedding","UMAP","PaCMAP","t-SNE","HDBSCAN","dimensionality reduction audio","vocal map UMAP","crow vocal atlas","graded calls","repertoire mapping","what AI can decode about crow calls","caller identity inference","individual signature crow","crow dialect","group-level acoustic centroid","contextual clustering crow","behavioral context audio","crow syntax","combinatorial crow communication","crow language","field recording crow","crow audio recording rig","Sennheiser ME66 crow","Zoom H1 H5 field recorder","behavior log synchronization","clap sync field audio","48kHz 24-bit audio capture","bandpass filter bioacoustics","preprocessing crow audio","noisereduce librosa","audio embedding extraction","from caw to cluster","crow translation pipeline","crow playback ethics","bioacoustic playback rules","animal welfare audio research","no playback near nests","alarm call playback ethics","six rules listening back","responsible animal communication research","demonstrated vs aspirational ALP","animal language processing 2026","real-time bidirectional crow dialogue","compositional decoding animal","wearable audio logger crows","Demartsev 2026 carrion crow","carrion crow repertoire mapping","is there a crow translator app","can AI talk to crows","do crows have language","do crows have grammar","do crows have dialect","what do crow calls mean","how do AI models understand animal sounds"],"sections":[{"name":"The Crow","type":"Pillar","path":"/the-crow","full_url":"https://crowlingo.org/the-crow","description":"Why the American crow as a model species: cognition, sociality, vocal anatomy, and a repertoire dense enough to warrant a map."},{"name":"Types of crow vocalizations","type":"Catalog","path":"/the-crow/vocalization-types","full_url":"https://crowlingo.org/the-crow/vocalization-types","description":"American crows produce nine acoustically-distinct call types: territorial caws, mobbing alarm, assembly calls, rattle, juvenile begging, companion calls, quiet grunts, loud grunts, and an exceptional category for atypical vocalizations. Each one defined, with real audio."},{"name":"Vocal anatomy","type":"Sub-page","path":"/the-crow/vocal-anatomy","full_url":"https://crowlingo.org/the-crow/vocal-anatomy","description":"How a crow makes sound — the syrinx, two independent sound sources, the 200 Hz – 8 kHz frequency window, and how it differs from a human larynx."},{"name":"Repertoire Atlas","type":"Interactive","path":"/the-crow/repertoire-atlas","full_url":"https://crowlingo.org/the-crow/repertoire-atlas","description":"An interactive 2-D map of crow vocalizations. ~800 seeded points across nine clusters; click any point to see cluster context, spectrogram, and behavioral probabilities."},{"name":"Cognition & society","type":"Sub-page","path":"/the-crow/cognition-and-society","full_url":"https://crowlingo.org/the-crow/cognition-and-society","description":"What makes the American crow worth taking seriously as a communicative animal — tool use, face recognition, family-group sociality, intergenerational learning."},{"name":"Methods","type":"Pillar","path":"/methods","full_url":"https://crowlingo.org/methods","description":"The new generation of AI audio methods — self-supervised learning, latent spaces, NatureLM-audio — and what they enable for crows specifically."},{"name":"Self-supervised audio","type":"Sub-page","path":"/methods/self-supervised-audio","full_url":"https://crowlingo.org/methods/self-supervised-audio","description":"How self-supervised learning trains audio models without labels — masked prediction, what the model actually learns, why it works for bioacoustics."},{"name":"Latent space 101","type":"Primer","path":"/methods/latent-space-101","full_url":"https://crowlingo.org/methods/latent-space-101","description":"Embeddings, latent spaces, and dimensionality reduction — the minimum mental model for reading a vocal atlas."},{"name":"NatureLM-audio","type":"Reference","path":"/methods/naturelm-audio","full_url":"https://crowlingo.org/methods/naturelm-audio","description":"Earth Species Project's audio-language foundation model for bioacoustics. ICLR 2025. What it does, what it doesn't, how it changed the workflow."},{"name":"Traditional vs ALP","type":"Sub-page","path":"/methods/traditional-vs-alp","full_url":"https://crowlingo.org/methods/traditional-vs-alp","description":"The fifty-year hand-labeling regime versus the new map-based regime. What the field gained; what it gave up."},{"name":"BirdNET vs Perch 2.0 vs NatureLM-audio","type":"Compare","path":"/methods/birdnet-vs-perch-vs-naturelm","full_url":"https://crowlingo.org/methods/birdnet-vs-perch-vs-naturelm","description":"Side-by-side comparison of the three audio foundation models that matter for bird and animal vocalization analysis in 2026: BirdNET (Cornell), Perch 2.0 (Google), and NatureLM-audio (Earth Species Project). Architecture, training, output, license, when to use which."},{"name":"Decoding","type":"Pillar","path":"/decoding","full_url":"https://crowlingo.org/decoding","description":"What we can now see in crow vocalizations that we couldn't see before — repertoire mapping, contextual clustering, individuality, combinatorial evidence."},{"name":"What we can decode now","type":"Flagship","path":"/decoding/what-we-can-decode-now","full_url":"https://crowlingo.org/decoding/what-we-can-decode-now","description":"The four features a self-supervised model extracts from one half-second of crow voice, and what each tells us — pitch contour, harmonic emphasis, duration, spectral grain."},{"name":"Contextual clustering","type":"Analysis","path":"/decoding/contextual-clustering","full_url":"https://crowlingo.org/decoding/contextual-clustering","description":"How latent coordinates correlate with behavior. The Demartsev 2026 carrion-crow preprint as the cleanest current example."},{"name":"Individuality & dialect","type":"Sub-page","path":"/decoding/individuality-and-dialect","full_url":"https://crowlingo.org/decoding/individuality-and-dialect","description":"Caller identity from harmonic signature, group-level acoustic centroids, and how seriously to take the dialect hypothesis."},{"name":"Combinatorial evidence","type":"Sub-page","path":"/decoding/combinatorial-evidence","full_url":"https://crowlingo.org/decoding/combinatorial-evidence","description":"Sequence-level statistical regularities in crow vocalizations and the open question of crow 'syntax'. Honest about behavioral evidence."},{"name":"Pipeline","type":"Centerpiece","path":"/pipeline","full_url":"https://crowlingo.org/pipeline","description":"Eight stages from a phone recording to an interpretable vocal map: capture, detect, preprocess, embed, project & cluster, contextualize, inspect, respond."},{"name":"Record","type":"Stage 1","path":"/pipeline/record","full_url":"https://crowlingo.org/pipeline/record","description":"Field-recording specifics for crow audio: microphone choice, sample rate, mono vs stereo, behavior-log synchronization, ethical floor."},{"name":"Preprocess","type":"Stage 3","path":"/pipeline/preprocess","full_url":"https://crowlingo.org/pipeline/preprocess","description":"Bandpass, peak-normalize, light spectral denoise. The minimum that helps without distorting what the model needs to read."},{"name":"Embed","type":"Stage 4","path":"/pipeline/embed","full_url":"https://crowlingo.org/pipeline/embed","description":"Pick your encoder honestly: BirdNET embeddings, Perch, CLAP, NatureLM-audio. Each is its own space. Disclose which."},{"name":"Cluster & label","type":"Stages 5–7","path":"/pipeline/cluster-and-label","full_url":"https://crowlingo.org/pipeline/cluster-and-label","description":"Project to 2-D for inspection, cluster on the full embeddings, label clusters by exemplars, join to behavior context."},{"name":"Respond","type":"Stage 8","path":"/pipeline/respond","full_url":"https://crowlingo.org/pipeline/respond","description":"How to run a playback session as data collection, not a stunt: pre-registered protocol, observer, time-bounded, halt on distress."},{"name":"Frontier","type":"Pillar","path":"/frontier","full_url":"https://crowlingo.org/frontier","description":"The honest state of the field: what's demonstrated, what's emerging, what's not yet science. Ethics. Open dataset. How to contribute."},{"name":"Current vs aspirational","type":"Sub-page","path":"/frontier/current-vs-aspirational","full_url":"https://crowlingo.org/frontier/current-vs-aspirational","description":"Demonstrated, emerging, and not-yet-science capabilities in animal-language processing for crows. A clean three-bucket framing."},{"name":"Open dataset","type":"Reference","path":"/frontier/open-dataset","full_url":"https://crowlingo.org/frontier/open-dataset","description":"10k+ labeled crow calls planned for v2 release on Hugging Face, CC-BY-NC. v0 placeholder; honest about the timeline."},{"name":"Contribute","type":"Submission","path":"/frontier/contribute","full_url":"https://crowlingo.org/frontier/contribute","description":"How to record crows well, and how to submit your recordings. v0: email + Google Form. v3 ships the proper upload pipeline."},{"name":"Library","type":"Reading list","path":"/library","full_url":"https://crowlingo.org/library","description":"Reading list for crow vocal communication and animal language processing — papers, books, primary sources, organized by constellation."},{"name":"Journal","type":"Long-form","path":"/journal","full_url":"https://crowlingo.org/journal","description":"Editorial deep-dives on what AI audio foundation models reveal about American crow vocalizations: methods, evidence, ethics, and open questions. Citation-driven, anti-hype, written for non-engineers but careful enough for practitioners."},{"name":"FAQ","type":"FAQ","path":"/faq","full_url":"https://crowlingo.org/faq","description":"Every common question about American crow vocalizations, AI audio foundation models, the CrowLingo vocal atlas, and the methods behind it — answered with quotable, citation-grounded responses."},{"name":"Glossary","type":"Glossary","path":"/glossary","full_url":"https://crowlingo.org/glossary","description":"Definitions of the technical vocabulary used across CrowLingo: embedding, latent space, UMAP, HDBSCAN, self-supervised learning, NatureLM-audio, BirdNET, Perch, syrinx, dialect, syntax. One sentence each, with deep links to the methods page that explains it."},{"name":"Territorial caw","type":"Cluster","path":"/the-crow/cluster/territorial","full_url":"https://crowlingo.org/the-crow/cluster/territorial","description":"Long-duration caws emitted from a perch, often paired and faced outward across a territory boundary. Carries caller identity, sex, and appro (3 CC-licensed recordings in v1.)"},{"name":"Mobbing alarm","type":"Cluster","path":"/the-crow/cluster/mobbing","full_url":"https://crowlingo.org/the-crow/cluster/mobbing","description":"Compressed, urgent caws delivered in rapid sequences, often by recruited pairs or groups targeting an aerial predator. Spectrally rough; tig (1 CC-licensed recording in v1.)"},{"name":"Assembly","type":"Cluster","path":"/the-crow/cluster/assembly","full_url":"https://crowlingo.org/the-crow/cluster/assembly","description":"Loud, far-carrying calls that summon group members to a roost or food source. Acoustically distinct from territorial in rate and inter-call  (4 CC-licensed recordings in v1.)"},{"name":"Rattle complex","type":"Cluster","path":"/the-crow/cluster/rattle","full_url":"https://crowlingo.org/the-crow/cluster/rattle","description":"Mechanical rattling vocalizations with weak harmonic structure. Frequent in affiliative and recruitment contexts. Highly individual. (2 CC-licensed recordings in v1.)"},{"name":"Quiet grunts","type":"Cluster","path":"/the-crow/cluster/quiet-grunts","full_url":"https://crowlingo.org/the-crow/cluster/quiet-grunts","description":"Low-amplitude, close-range grunts used in affiliative and parent-offspring contexts. Subtle; only the new wearable-logger studies recover th (exemplars coming.)"},{"name":"Loud grunts","type":"Cluster","path":"/the-crow/cluster/loud-grunts","full_url":"https://crowlingo.org/the-crow/cluster/loud-grunts","description":"Higher-amplitude grunts during foraging and recruitment. Often serial and quickly answered by group members. (exemplars coming.)"},{"name":"Begging","type":"Cluster","path":"/the-crow/cluster/begging","full_url":"https://crowlingo.org/the-crow/cluster/begging","description":"Higher-frequency, narrower-band calls from juveniles soliciting feeding. Diagnostic spectral signature; tightly clustered. (1 CC-licensed recording in v1.)"},{"name":"Exceptional","type":"Cluster","path":"/the-crow/cluster/exceptional","full_url":"https://crowlingo.org/the-crow/cluster/exceptional","description":"Rare, atypical, or unusual vocalizations that don't fit the named categories. Where future repertoire expansion lives. (1 CC-licensed recording in v1.)"},{"name":"Companion","type":"Cluster","path":"/the-crow/cluster/companion","full_url":"https://crowlingo.org/the-crow/cluster/companion","description":"Soft contact calls between paired adults. Heavily individual and pair-specific; acoustic signatures distinguish couples. (4 CC-licensed recordings in v1.)"},{"name":"How AI is decoding crow vocalizations in 2026","type":"Article","path":"/journal/ai-decoding-crow-vocalizations-2026","full_url":"https://crowlingo.org/journal/ai-decoding-crow-vocalizations-2026","description":"Until about 2022, every published account of American crow vocal repertoires was a hand-curated taxonomy: someone listened to hundreds of recordings, picked categories, drew boundaries. The new generation of audio foundation models flipped "},{"name":"Do crows have dialects? What the evidence actually shows","type":"Article","path":"/journal/do-crows-have-dialects","full_url":"https://crowlingo.org/journal/do-crows-have-dialects","description":"The short answer is yes, in a careful sense. The longer answer is what the careful sense means. Family groups of American crows carry measurable acoustic centroids — average vectors of their shared call types — that differ between geographi"},{"name":"Why crows recognize human faces","type":"Article","path":"/journal/marzluff-mask-experiments-explained","full_url":"https://crowlingo.org/journal/marzluff-mask-experiments-explained","description":"Between 2006 and 2014, researchers at the University of Washington put on caveman masks, trapped crows on campus, released them, and waited. What they found settled an old argument about corvid cognition and opened a new one about how commu"},{"name":"The nine emergent clusters of American crow communication","type":"Article","path":"/journal/nine-clusters-of-crow-communication","full_url":"https://crowlingo.org/journal/nine-clusters-of-crow-communication","description":"When researchers stopped naming crow vocalizations in advance and let a self-supervised audio model embed them instead, nine dense regions emerged in the high-dimensional space. The names — territorial, mobbing, assembly, rattle, begging, c"},{"name":"Self-supervised audio learning, explained for non-engineers","type":"Article","path":"/journal/self-supervised-audio-for-non-engineers","full_url":"https://crowlingo.org/journal/self-supervised-audio-for-non-engineers","description":"Until self-supervised learning crossed from natural-language processing into audio in roughly 2022, every bioacoustic AI model had to be told what to look for. Someone built a labeled training set — these clips are robins, these are sparrow"},{"name":"BirdNET vs Perch 2.0 vs NatureLM-audio: the practical 2026 guide","type":"Article","path":"/journal/birdnet-vs-perch-practical-guide","full_url":"https://crowlingo.org/journal/birdnet-vs-perch-practical-guide","description":"Three models dominate practical bioacoustics in 2026: BirdNET from Cornell, Perch 2.0 from Google Research, and NatureLM-audio from Earth Species Project. They are not interchangeable. Each is optimized for a different job, runs on differen"},{"name":"Wearable bioacoustic loggers and the Demartsev revolution","type":"Article","path":"/journal/wearable-bioacoustic-loggers-demartsev","full_url":"https://crowlingo.org/journal/wearable-bioacoustic-loggers-demartsev","description":"For decades the rate-limiting step in wild bird vocal research was synchronization. You could record audio. You could record behavior. Joining them at the resolution needed to ask 'what was this individual doing when it made this call' requ"},{"name":"The ethics of playback experiments in bioacoustic research","type":"Article","path":"/journal/ethics-of-playback-experiments","full_url":"https://crowlingo.org/journal/ethics-of-playback-experiments","description":"Playback experiments — playing recorded vocalizations back to wild animals to test how they respond — are the methodological backbone of how field biologists test claims about meaning in animal communication. They are also the source of nea"},{"name":"From caw to cluster: the eight-stage pipeline explained","type":"Article","path":"/journal/from-caw-to-cluster-pipeline","full_url":"https://crowlingo.org/journal/from-caw-to-cluster-pipeline","description":"Eight stages, in order: capture, detect, preprocess, embed, project, cluster, contextualize, respond. That is the modern bioacoustic pipeline in its simplest form. Each stage adds value; each stage has a characteristic failure mode that, if"},{"name":"How to record crows: a field guide for citizen scientists","type":"Article","path":"/journal/field-recording-guide-citizen-scientists","full_url":"https://crowlingo.org/journal/field-recording-guide-citizen-scientists","description":"Useful crow recordings start with one decision that every other choice cascades from: you're recording for archival quality, not for a one-time listen. Archival means the file will be reusable, attributable, license-compatible, and acoustic"},{"name":"What 'translating' animal language would actually require","type":"Article","path":"/journal/what-translating-animal-language-requires","full_url":"https://crowlingo.org/journal/what-translating-animal-language-requires","description":"Every six months a popular outlet runs the same headline: AI is about to translate animal language. The piece quotes a researcher, names a foundation model, sketches a hopeful timeline, and ends with a quote about how this will change how w"},{"name":"Why the American crow is the model species for AI bioacoustics","type":"Article","path":"/journal/why-american-crow-model-species","full_url":"https://crowlingo.org/journal/why-american-crow-model-species","description":"Choosing a species for animal-language research is not neutral. The species shapes which questions are tractable, which methods are applicable, which findings are interpretable, and which generalizations the field can defend. The American c"},{"name":"Why animal-language AI is harder than human-language AI","type":"Article","path":"/journal/why-animal-language-ai-is-hard","full_url":"https://crowlingo.org/journal/why-animal-language-ai-is-hard","description":"Both human-language AI and animal-language AI start with the same architecture — a transformer pretrained on large unlabeled corpora via self-supervised objectives. Both produce embeddings. Both downstream-classify, retrieve, generate. The "},{"name":"The Mates 2014 paper and why it still matters","type":"Article","path":"/journal/mates-2014-why-it-matters","full_url":"https://crowlingo.org/journal/mates-2014-why-it-matters","description":"Eva Mates, R.R. Tarter, James Ha, Anne Clark, and Kevin McGowan published 'Acoustic profiling in a complexly social species, the American crow' in the journal Bioacoustics in 2014. The paper is short by today's standards, cited unevenly out"},{"name":"Corvid cognition: fifty years of breakthroughs","type":"Article","path":"/journal/corvid-cognition-fifty-years","full_url":"https://crowlingo.org/journal/corvid-cognition-fifty-years","description":"The contemporary AI bioacoustics work on American crows would not be possible — or interesting — without fifty years of cognitive research that established what kind of mind sits behind the calls. The through-line from the 1970s pioneers to"},{"name":"What we can and cannot decode about crow communication, today","type":"Article","path":"/journal/what-we-can-and-cannot-decode","full_url":"https://crowlingo.org/journal/what-we-can-and-cannot-decode","description":"Pop science coverage of AI bioacoustics oscillates between two extremes. One: we've already cracked the code — crow language is being translated. Two: it's all hype — AI tells us nothing useful about wild animals. The truth is more interest"},{"name":"Crow funerals and crow grief: separating evidence from anthropomorphism","type":"Article","path":"/journal/crow-funerals-and-grief","full_url":"https://crowlingo.org/journal/crow-funerals-and-grief","description":"Search 'crow funerals' and you'll find hundreds of articles claiming that crows hold mourning rituals for their dead. The framing is irresistible. It is also more than the science supports — and the part of it that science does support is i"},{"name":"How AI changed birdwatching: BirdNET, Merlin, and the citizen-science boom","type":"Article","path":"/journal/how-ai-changed-birdwatching","full_url":"https://crowlingo.org/journal/how-ai-changed-birdwatching","description":"Ask a research bioacoustician which AI bioacoustics deployment has had the biggest real-world impact in the past five years and they probably won't name a peer-reviewed paper. They'll name an app. Specifically, they'll name Merlin Sound ID,"},{"name":"Project CETI and what we learned from sperm whales","type":"Article","path":"/journal/project-ceti-and-cetacean-language","full_url":"https://crowlingo.org/journal/project-ceti-and-cetacean-language","description":"The Cetacean Translation Initiative — CETI — is the most ambitious public animal-language AI project of the past decade. Multi-million-dollar funding from the Audacious Project, a multidisciplinary team led by Shafi Goldwasser, David Gruber"},{"name":"Reading the vocal atlas: a beginner's guide to UMAP for bird sounds","type":"Article","path":"/journal/reading-the-vocal-atlas","full_url":"https://crowlingo.org/journal/reading-the-vocal-atlas","description":"The first time you look at a UMAP scatter of bird vocalizations — including the one at the center of CrowLingo's atlas — there's a moment of confusion. What are the axes? Why are there clusters? What does it mean when two dots are close tog"},{"name":"Crow tools and the cognitive ceiling","type":"Article","path":"/journal/crow-tools-cognitive-ceiling","full_url":"https://crowlingo.org/journal/crow-tools-cognitive-ceiling","description":"If you've seen a viral video of a crow shaping a hooked tool from a twig and using it to fish a grub out of a log, you've seen a New Caledonian crow — not the American crow CrowLingo is built around. The two species are corvid cousins, but "},{"name":"Bernd Heinrich, Mind of the Raven, and the slow-naturalist tradition","type":"Article","path":"/journal/heinrich-mind-of-raven-naturalist-tradition","full_url":"https://crowlingo.org/journal/heinrich-mind-of-raven-naturalist-tradition","description":"Bernd Heinrich's 1999 book Mind of the Raven is the kind of contemporary classic that doesn't fit neatly into the AI-bioacoustics conversation but quietly anchors it. Heinrich spent decades observing common ravens in the Maine and Vermont w"},{"name":"Spectrograms decoded: what those squiggles actually show","type":"Article","path":"/journal/spectrograms-decoded","full_url":"https://crowlingo.org/journal/spectrograms-decoded","description":"A spectrogram is a picture of sound. Time runs across the horizontal axis, frequency runs up the vertical, and brightness or color shows how much acoustic energy is present at each time-frequency point. Once you understand that one sentence"},{"name":"Passive acoustic monitoring: what AI ears reveal about wild populations","type":"Article","path":"/journal/acoustic-monitoring-for-conservation","full_url":"https://crowlingo.org/journal/acoustic-monitoring-for-conservation","description":"While most of the AI-bioacoustics public coverage focuses on the headline question of whether we can talk to animals, the quietly bigger story is what AI ears are doing for conservation biology. Inexpensive devices the size of a deck of car"},{"name":"What you'll hear in your backyard: a beginner's guide to crow vocalizations","type":"Article","path":"/journal/backyard-crow-sounds-guide","full_url":"https://crowlingo.org/journal/backyard-crow-sounds-guide","description":"If you have any tree taller than ten feet within a city block of your home in North America, you have crows. They're probably the loudest birds in your neighborhood, and the most acoustically interesting. Most people tune them out as backgr"},{"name":"How to tell a crow from a raven by sound","type":"Article","path":"/journal/crow-vs-raven-by-sound","full_url":"https://crowlingo.org/journal/crow-vs-raven-by-sound","description":"American crows and common ravens are the two corvids most likely to confuse a casual observer in North America. The visual differences — raven is bigger, has a heavier bill, a diamond-shaped tail, shaggy throat feathers — are real but easy "},{"name":"Why crows gather at dusk: the roost-formation phenomenon","type":"Article","path":"/journal/why-crows-gather-at-dusk","full_url":"https://crowlingo.org/journal/why-crows-gather-at-dusk","description":"If you've ever stood under a tree at sunset and watched what felt like every crow in the city descend onto it, you've witnessed roost formation. It's not a rare event. In many North American cities, the winter roost of American crows number"},{"name":"The cooperative-breeding family structure of American crows","type":"Article","path":"/journal/cooperative-breeding-crow-family-structure","full_url":"https://crowlingo.org/journal/cooperative-breeding-crow-family-structure","description":"Most North American songbirds follow a similar life pattern: hatch, fledge, disperse, find a mate, breed independently. American crows do not. Many young crows stay with their natal family for two to seven years before establishing their ow"},{"name":"Do crows mate for life? What the long-term pair-bond research shows","type":"Article","path":"/journal/do-crows-mate-for-life","full_url":"https://crowlingo.org/journal/do-crows-mate-for-life","description":"Search 'do crows mate for life' and you'll get a confident yes from most popular sources. The careful answer is: mostly yes, often yes, more often than for most bird species, with documented exceptions and a few empirical caveats. The reaso"},{"name":"Voxaboxen and the open infrastructure of modern bioacoustics","type":"Article","path":"/journal/voxaboxen-and-open-bioacoustic-infrastructure","full_url":"https://crowlingo.org/journal/voxaboxen-and-open-bioacoustic-infrastructure","description":"The audio foundation models that map crow vocal repertoires — Perch 2.0, NatureLM-audio, BirdNET — are the headline. Behind them sits a layer of less-glamorous infrastructure that does the unsexy work of finding the vocalizations in long re"},{"name":"What the Wright laboratory is up to: contemporary corvid vocal research","type":"Article","path":"/journal/wright-laboratory-corvid-vocal-research","full_url":"https://crowlingo.org/journal/wright-laboratory-corvid-vocal-research","description":"The contemporary research landscape in corvid vocal communication includes plenty of AI-driven groups producing foundation-model results. It also includes labs doing patient behavioral work that doesn't try to leverage scale, but produces f"},{"name":"Acoustic indices: measuring ecosystem health through sound","type":"Article","path":"/journal/acoustic-indices-ecosystem-health","full_url":"https://crowlingo.org/journal/acoustic-indices-ecosystem-health","description":"Counting birds is expensive. Counting every bird, every frog, every insect, every mammal in a habitat is impossible. So a parallel research program in soundscape ecology has been asking: what if we don't try to identify every vocalizing org"},{"name":"How smart are crows? A measured assessment","type":"Article","path":"/journal/how-smart-are-crows","full_url":"https://crowlingo.org/journal/how-smart-are-crows","description":"Search 'how smart are crows' and you'll find rankings comparing them to chimps, dolphins, dogs, and seven-year-old children. The rankings vary because the question varies. There is no single 'intelligence' that can be ranked across species."},{"name":"What we don't know about crow communication","type":"Article","path":"/journal/what-we-dont-know-about-crows","full_url":"https://crowlingo.org/journal/what-we-dont-know-about-crows","description":"Most science writing emphasizes what's known. Done right, naming what isn't known is more useful — it shows where the work that matters is going to happen, helps readers calibrate confidence in popular claims, and reminds the field of its o"},{"name":"The unexpected richness of crow play behavior","type":"Article","path":"/journal/crow-play-behavior","full_url":"https://crowlingo.org/journal/crow-play-behavior","description":"Marc Bekoff's standard definition of animal play behavior has four parts: it has to be voluntary, intrinsically motivated, structurally varied, and observed in safe contexts. By this definition, crows play. Not occasionally — frequently, ac"},{"name":"What corvid researchers wish you understood about AI bioacoustics","type":"Article","path":"/journal/what-corvid-researchers-wish-you-understood","full_url":"https://crowlingo.org/journal/what-corvid-researchers-wish-you-understood","description":"If you sat down with a corvid researcher who's been in the field for fifteen years and asked them what they wished the public understood about AI bioacoustics, the answer would not be 'we're closer to translation than you think.' It would b"},{"name":"Do crows really bring gifts? Untangling the evidence","type":"Article","path":"/journal/do-crows-bring-gifts","full_url":"https://crowlingo.org/journal/do-crows-bring-gifts","description":"In 2015 the BBC ran a story about an eight-year-old in Seattle named Gabi Mann who had been feeding crows in her backyard and started receiving small objects from them — beads, buttons, a piece of metal, a tiny piece of porcelain. The story"},{"name":"How long do crows live? The longevity question, answered carefully","type":"Article","path":"/journal/crow-lifespan-longevity","full_url":"https://crowlingo.org/journal/crow-lifespan-longevity","description":"Search 'how long do crows live' and you'll find numbers ranging from seven years to fifty-nine years, often without context. The wide range isn't a measurement problem; it reflects two very different statistical realities — what an average "},{"name":"Predator-specific alarm calls: do crows have words for hawks?","type":"Article","path":"/journal/predator-specific-alarm-calls","full_url":"https://crowlingo.org/journal/predator-specific-alarm-calls","description":"The most-cited finding in animal-communication research is probably Robert Seyfarth and Dorothy Cheney's 1980 paper showing that vervet monkeys produce distinct alarm calls for leopards, eagles, and snakes — calls that elicit predator-type-"},{"name":"Crow population trends in North America: what acoustic monitoring shows","type":"Article","path":"/journal/crow-population-trends-acoustic-monitoring","full_url":"https://crowlingo.org/journal/crow-population-trends-acoustic-monitoring","description":"American crow populations across North America have moved in different directions over the past two decades. Some regions have lost substantial population; others have gained. The classical visual-survey methods of the Breeding Bird Survey "},{"name":"Carrion crow versus American crow: a comparative species note","type":"Article","path":"/journal/carrion-crow-vs-american-crow","full_url":"https://crowlingo.org/journal/carrion-crow-vs-american-crow","description":"Two species of crow occupy approximately equivalent ecological niches on opposite sides of the Atlantic: the American crow (Corvus brachyrhynchos) in North America, the carrion crow (Corvus corone) across western Europe and beyond. They loo"},{"name":"Honeybee waggle dance: the gold standard for animal communication decoding","type":"Article","path":"/journal/honeybee-waggle-dance-gold-standard","full_url":"https://crowlingo.org/journal/honeybee-waggle-dance-gold-standard","description":"Karl von Frisch won the 1973 Nobel Prize in Physiology or Medicine for cracking the honeybee waggle dance. To this day it remains the single most rigorously decoded non-human communication system on the planet. Understanding what made the d"},{"name":"Why corvids and parrots both vocal-learn: a convergent evolutionary puzzle","type":"Article","path":"/journal/vocal-learning-in-corvids-and-parrots","full_url":"https://crowlingo.org/journal/vocal-learning-in-corvids-and-parrots","description":"Vocal learning — the ability to produce vocalizations based on auditory experience rather than purely innate templates — is rare across animals. Most species produce stereotyped vocalizations from birth. The few that learn their vocalizatio"},{"name":"What spectrograms can't tell you","type":"Article","path":"/journal/what-spectrograms-cant-tell-you","full_url":"https://crowlingo.org/journal/what-spectrograms-cant-tell-you","description":"Spectrograms are the standard 2D visualization for sound. They're useful enough that every bioacoustic paper includes them, every audio-analysis tool produces them, every AI bioacoustics model operates on them. They're also misleading in sp"},{"name":"Urban versus rural crow behavior: what the acoustic record shows","type":"Article","path":"/journal/urban-versus-rural-crow-behavior","full_url":"https://crowlingo.org/journal/urban-versus-rural-crow-behavior","description":"An American crow in downtown Seattle and an American crow on a Wenatchee orchard are the same species. They share genetics, anatomy, vocal-learning machinery, and the deep behavioral template that defines the species. They also behave diffe"},{"name":"Kaeli Swift and the crow funeral question","type":"Article","path":"/journal/kaeli-swift-funeral-research","full_url":"https://crowlingo.org/journal/kaeli-swift-funeral-research","description":"The 'crow funeral' phenomenon — groups of crows gathering noisily around a dead conspecific — became internet famous around 2015 and has been a recurrent topic since. The systematic research behind the phenomenon is mostly the work of Kaeli"},{"name":"The Macaulay Library and the open-research tension","type":"Article","path":"/journal/macaulay-library-and-open-research-tension","full_url":"https://crowlingo.org/journal/macaulay-library-and-open-research-tension","description":"The Macaulay Library at the Cornell Lab of Ornithology holds approximately 1.3 million wildlife audio recordings as of 2026, the largest single archive of wildlife sound on the planet. Its data is partially open (preview clips, individual r"},{"name":"West Nile virus and the American crow population collapse","type":"Article","path":"/journal/west-nile-virus-and-american-crow-collapse","full_url":"https://crowlingo.org/journal/west-nile-virus-and-american-crow-collapse","description":"West Nile virus, a mosquito-borne pathogen originating in sub-Saharan Africa, was first detected in the Western Hemisphere in New York City in 1999. Within five years it had spread across most of North America, and along the way it killed m"},{"name":"Crow olfaction: the quiet sensory mode","type":"Article","path":"/journal/crow-olfaction-the-quiet-sensory-mode","full_url":"https://crowlingo.org/journal/crow-olfaction-the-quiet-sensory-mode","description":"For most of the twentieth century, ornithology textbooks taught that birds have a poor sense of smell. The conclusion was based partly on small olfactory bulbs relative to mammalian brains and partly on a lack of obvious smelling behavior. "},{"name":"Climate change and corvid range shifts: what the data shows","type":"Article","path":"/journal/climate-change-and-corvid-range-shifts","full_url":"https://crowlingo.org/journal/climate-change-and-corvid-range-shifts","description":"Bird ranges across North America have shifted measurably with warming since the 1970s, and the shifts are accelerating. American crow is among the species that has expanded its range — northward and into higher elevations — over this period"},{"name":"How eBird became a bioacoustic engine","type":"Article","path":"/journal/ebird-as-bioacoustic-engine","full_url":"https://crowlingo.org/journal/ebird-as-bioacoustic-engine","description":"eBird, launched by the Cornell Lab of Ornithology and the Audubon Society in 2002, started as a checklist platform for bird sightings. Submit a list of birds you saw in a particular location at a particular time; the data feeds into populat"},{"name":"Acoustic monitoring as a conservation tool","type":"Article","path":"/journal/acoustic-monitoring-as-conservation-tool","full_url":"https://crowlingo.org/journal/acoustic-monitoring-as-conservation-tool","description":"Passive acoustic monitoring (PAM) — deploying autonomous recording devices in the field to capture continuous or scheduled audio from a study area — has matured into a standard conservation tool over the past decade. It changes what conserv"},{"name":"Alex the parrot and the comparative-cognition tradition","type":"Article","path":"/journal/alex-the-parrot-and-comparative-bird-cognition","full_url":"https://crowlingo.org/journal/alex-the-parrot-and-comparative-bird-cognition","description":"Irene Pepperberg's thirty-year study of an African grey parrot named Alex remains the most-cited animal-cognition research in popular science. Alex died in 2007 at age 31. The research produced careful documentation of capacities most peopl"},{"name":"Project CETI and the sperm whale codas","type":"Article","path":"/journal/project-ceti-and-sperm-whales","full_url":"https://crowlingo.org/journal/project-ceti-and-sperm-whales","description":"The Cetacean Translation Initiative (Project CETI), launched in 2020, is the most-ambitious effort to date to apply modern AI to the communication system of a non-human species. The target is the sperm whale, the species's communication uni"},{"name":"How birds actually produce sound: the syrinx briefly explained","type":"Article","path":"/journal/how-birds-actually-produce-sound","full_url":"https://crowlingo.org/journal/how-birds-actually-produce-sound","description":"Birds don't have a larynx like mammals. They have a syrinx, a sound-producing organ unique to birds, located at the bottom of the trachea where the airway splits into the two bronchi. The structural differences between syrinx and larynx sha"},{"name":"The replication problem in animal cognition","type":"Article","path":"/journal/replication-crisis-in-animal-cognition","full_url":"https://crowlingo.org/journal/replication-crisis-in-animal-cognition","description":"The replication crisis that hit psychology hard in the 2010s also touched animal cognition research, though less attention has been paid to it. Several high-profile findings in the field haven't replicated as well as the original publicatio"},{"name":"John Marzluff and the Seattle crow lab","type":"Article","path":"/journal/john-marzluff-and-the-seattle-crow-lab","full_url":"https://crowlingo.org/journal/john-marzluff-and-the-seattle-crow-lab","description":"Most of what the modern public knows about American crow cognition — face recognition, social learning of threat, brain-imaging correlates of fear response — traces back to a single research program at the University of Washington. John Mar"},{"name":"Bernd Heinrich and the ravens of winter","type":"Article","path":"/journal/bernd-heinrich-and-ravens-of-winter","full_url":"https://crowlingo.org/journal/bernd-heinrich-and-ravens-of-winter","description":"Before contemporary corvid cognition became a research field with brain-imaging and AI-augmented analysis, Bernd Heinrich spent winters in the Maine woods watching what ravens actually do. The methodology is mostly older than the instrument"},{"name":"Nicola Clayton and the scrub jay memory studies","type":"Article","path":"/journal/nicola-clayton-and-scrub-jay-memory","full_url":"https://crowlingo.org/journal/nicola-clayton-and-scrub-jay-memory","description":"Episodic memory — the capacity to recall specific events along with where and when they occurred — was considered uniquely human (or at least uniquely primate) until Nicola Clayton's research on western scrub-jays demonstrated it in birds. "},{"name":"Auguste von Bayern and the New Caledonian crow tool program","type":"Article","path":"/journal/auguste-von-bayern-and-new-caledonian-crow-tools","full_url":"https://crowlingo.org/journal/auguste-von-bayern-and-new-caledonian-crow-tools","description":"New Caledonian crows make tools. They modify the tools. They use sequences of tools where solving a problem requires combining the tools they've built. The systematic research documenting this — and increasingly the experimental work on wha"},{"name":"The avian pallium and the bird brain revolution","type":"Article","path":"/journal/avian-pallium-and-the-bird-brain-revolution","full_url":"https://crowlingo.org/journal/avian-pallium-and-the-bird-brain-revolution","description":"For most of the twentieth century, neuroscience taught that bird brains lacked the cortical architecture that supports mammalian cognition. The avian forebrain was characterized as 'paleocortex' or 'basal ganglia,' the implication being tha"},{"name":"The history of bird recording technology","type":"Article","path":"/journal/the-history-of-bird-recording-technology","full_url":"https://crowlingo.org/journal/the-history-of-bird-recording-technology","description":"Wildlife audio recording is now cheap, automatic, and ubiquitous. A four-figure AudioMoth deployment can monitor a forest for weeks. A free Cornell Merlin Sound ID install on a smartphone identifies birds in real time. Getting there took a "},{"name":"The next decade of bioacoustic AI: a careful forecast","type":"Article","path":"/journal/the-next-decade-of-bioacoustic-ai","full_url":"https://crowlingo.org/journal/the-next-decade-of-bioacoustic-ai","description":"Predicting where AI bioacoustics is going over the next decade is a fraught exercise. Predictions in adjacent fields (language models, image generation, robotics) have repeatedly been wrong in both directions. The honest version of a 2026-t"},{"name":"What good bioacoustic citizen science looks like","type":"Article","path":"/journal/what-good-bioacoustic-citizen-science-looks-like","full_url":"https://crowlingo.org/journal/what-good-bioacoustic-citizen-science-looks-like","description":"Citizen science contributes most of the bioacoustic data the modern AI revolution runs on. The eBird-Macaulay pipeline, the iNaturalist audio integration, and a handful of similar platforms now feed millions of recordings annually into the "},{"name":"How to read a bioacoustics paper","type":"Article","path":"/journal/how-to-read-a-bioacoustics-paper","full_url":"https://crowlingo.org/journal/how-to-read-a-bioacoustics-paper","description":"Bioacoustics papers are mostly written for specialists. The structure is consistent enough across the field that non-specialists can read them productively with a few orientation tips. The honest version of reading a research paper is not '"},{"name":"Convergent intelligence across distant lineages","type":"Article","path":"/journal/convergent-intelligence-across-distant-lineages","full_url":"https://crowlingo.org/journal/convergent-intelligence-across-distant-lineages","description":"Crows, dolphins, octopuses, elephants, and humans all show signs of high intelligence on multiple cognitive measures. Their last common ancestor lived hundreds of millions of years ago, in some cases over half a billion years ago. The indep"},{"name":"What machine learning probably can't decode, no matter how much audio we feed it","type":"Article","path":"/journal/what-cant-be-decoded-by-machine-learning","full_url":"https://crowlingo.org/journal/what-cant-be-decoded-by-machine-learning","description":"Some claims about AI animal-communication decoding will probably never be supported, regardless of how much data we have, how much compute we apply, or how sophisticated our models become. The limits aren't always about technology; they are"},{"name":"Bird taxonomy in flux: what counts as a species","type":"Article","path":"/journal/bird-taxonomy-in-flux","full_url":"https://crowlingo.org/journal/bird-taxonomy-in-flux","description":"Bird species lists are being revised constantly as DNA evidence reshapes the field's understanding of species boundaries. Within the corvid genus alone, the carrion crow and hooded crow have been split as separate species; the northwestern "},{"name":"Listening to your local crows: a practical guide","type":"Article","path":"/journal/listening-to-your-local-crows","full_url":"https://crowlingo.org/journal/listening-to-your-local-crows","description":"The crows in your neighborhood are accessible bioacoustic subjects. They're easy to find, they're vocal, they're patient with respectful observers, and they have rich enough vocal repertoires to reward sustained attention. With the CrowLing"},{"name":"Crows in human cultures: a brief survey","type":"Article","path":"/journal/crows-in-human-cultures","full_url":"https://crowlingo.org/journal/crows-in-human-cultures","description":"Crows show up in human cultural traditions across nearly every continent. The cultural significance is older than scientific ornithology by thousands of years and informs how people perceive the species today, sometimes constructively and s"},{"name":"The Bothell roost and other mega-aggregations","type":"Article","path":"/journal/the-bothell-roost-and-other-mega-aggregations","full_url":"https://crowlingo.org/journal/the-bothell-roost-and-other-mega-aggregations","description":"Some American crow communal roosts contain tens of thousands of individuals overnighting together. The University of Washington Bothell campus, outside Seattle, hosts a winter roost of approximately 15,000-20,000 American crows that has bee"},{"name":"What the CrowLingo atlas is not","type":"Article","path":"/journal/what-the-crowlingo-atlas-is-not","full_url":"https://crowlingo.org/journal/what-the-crowlingo-atlas-is-not","description":"Boundary-setting matters for any reference work, and being honest about what the CrowLingo atlas is not is part of being credible about what it is. The temptation in a content-rich, well-designed reference is to claim more than the underlyi"},{"name":"Crow play behavior: what it shows about cognition","type":"Article","path":"/journal/crow-play-behavior","full_url":"https://crowlingo.org/journal/crow-play-behavior","description":"Crows engage in behavior that looks like play to human observers — sliding down snowy roofs, dropping and catching objects, tugging on the tails of cats and dogs, riding wind currents apparently for the experience. The play interpretation h"},{"name":"Captive corvid research: what the ethical debate involves","type":"Article","path":"/journal/captive-corvid-research-ethics","full_url":"https://crowlingo.org/journal/captive-corvid-research-ethics","description":"Most of what's been established about corvid cognition comes from captive studies. The Pepperberg Alex work was captive. The Clayton scrub-jay episodic-memory work was captive. The von Bayern New Caledonian crow tool work is captive. The et"},{"name":"Crow sleep and roost acoustics","type":"Article","path":"/journal/crow-sleep-and-roost-acoustics","full_url":"https://crowlingo.org/journal/crow-sleep-and-roost-acoustics","description":"When the day's territorial calls and mobbing alarms end, when the family group has settled into the communal roost for the night, the acoustic record doesn't fully stop. Crows sleep in patterns that include occasional nocturnal vocalization"},{"name":"Indigenous knowledge and Western corvid science","type":"Article","path":"/journal/indigenous-knowledge-and-western-corvid-science","full_url":"https://crowlingo.org/journal/indigenous-knowledge-and-western-corvid-science","description":"Pacific Northwest Indigenous communities have observed and integrated knowledge of raven and crow behavior for thousands of years. Their understanding of the species — encoded in oral traditions, behavioral expectations, ecological knowledg"},{"name":"Vervet monkey alarm calls: the other gold-standard study","type":"Article","path":"/journal/vervet-monkey-alarm-calls","full_url":"https://crowlingo.org/journal/vervet-monkey-alarm-calls","description":"Before the AI revolution, the Cheney-Seyfarth vervet monkey work was the closest thing animal communication research had to demonstrated referential semantic content in a non-human species. Vervets in the Amboseli area of Kenya produce thre"},{"name":"What 'open source' actually means in bioacoustic AI","type":"Article","path":"/journal/what-open-source-means-in-bioacoustic-ai","full_url":"https://crowlingo.org/journal/what-open-source-means-in-bioacoustic-ai","description":"BirdNET is described as open source. Perch 2.0 is described as open source. NatureLM-audio is described as open source. The phrase 'open source' covers significantly more variation in practice than the marketing implies, and the variation m"},{"name":"Crows and humans: a coevolution sketch","type":"Article","path":"/journal/crows-and-humans-a-coevolution-sketch","full_url":"https://crowlingo.org/journal/crows-and-humans-a-coevolution-sketch","description":"American crow's ecological history with humans goes back further than most people appreciate. The species's relationship with human-modified landscapes — agricultural fields, settlements, refuse sites — predates European colonization of Nor"},{"name":"Stowell 2022: the deep-learning bioacoustics review that grounds the field","type":"Article","path":"/journal/stowell-2022-deep-learning-bioacoustics-review","full_url":"https://crowlingo.org/journal/stowell-2022-deep-learning-bioacoustics-review","description":"Dan Stowell's 2022 review paper in PeerJ Computer Science, titled 'Computational bioacoustics with deep learning: a review and roadmap,' is the most-cited methodological reference in modern bioacoustic AI work. The paper synthesized the sta"}],"related_tools":[{"name":"Kymata Labs","url":"https://kymatalabs.com","description":"Parent publishing entity. Editorial intelligence + applied AI projects.","category":"Publisher"}],"trust_and_methodology":[{"name":"Ethics","path":"/frontier/ethics","full_url":"https://crowlingo.org/frontier/ethics"},{"name":"About","path":"/about","full_url":"https://crowlingo.org/about"},{"name":"Privacy","path":"/privacy","full_url":"https://crowlingo.org/privacy"},{"name":"Terms","path":"/terms","full_url":"https://crowlingo.org/terms"},{"name":"Disclaimer","path":"/disclaimer","full_url":"https://crowlingo.org/disclaimer"},{"name":"AI & Developer Access","path":"/ai","full_url":"https://crowlingo.org/ai"}],"machine_readable_endpoints":{"llms_txt":"https://crowlingo.org/llms.txt","llms_full_txt":"https://crowlingo.org/llms-full.txt","ai_json":"https://crowlingo.org/ai.json","ai_json_well_known":"https://crowlingo.org/.well-known/ai.json","ai_plugin_json":"https://crowlingo.org/.well-known/ai-plugin.json","openapi_yaml":"https://crowlingo.org/.well-known/openapi.yaml","sitemap_xml":"https://crowlingo.org/sitemap.xml","feed_json":"https://crowlingo.org/feed.json","schema_feed_json":"https://crowlingo.org/schema-feed.json","nlweb_ask":"https://crowlingo.org/api/ask","nlweb_mcp":"https://crowlingo.org/api/mcp"},"updated_at":"2026-05-20T13:19:08.550Z"}