There’s something very cool about listening to a wolf’s howl and seeing it at the same time. That’s what spectrograms do. They are graphic representations of sounds that literally visualize pitch, tone, and loudness as you listen. Until I started watching a few, I never realized that a single wolf can howl in harmonics, making one wolf sound like two or three.

Bioacoustics is the art of recording the sounds of living things. Those who do this work have significantly enlarged our understanding of the natural world since the 1970s through recording the songs of whales and birds, the clicks of dolphins, and the infrasonic calls of elephants. Now AI is being used to search through massive amounts of recordings, including those of wolf vocalizations, to find patterns and sequences that may help us better understand what the animals are saying.

The Crying Wolf Project, the team behind this leap in cataloguing wolf vocalizations, is studying the sounds of wild wolves — some of the most famous wolves of Yellowstone National Park, where they were exterminated in the 1920s and successfully reintroduced in 1995 — using highly sensitive microphones, 24-hour recorders, spectrograms, and AI.

At first glance, it doesn’t seem like they have gleaned a whole lot of new information. When a wolf howls, or a chorus of wolves howl together, there’s a whole lot going on we still don’t understand. We don’t know why they repeat sounds, or if these “phrases” are made up of “words.” We don’t know if their vocal chords might be different from our own, or if they can hear things we can’t.

But the Crying Wolf Project’s recordings in Yellowstone indicate the wolves are communicating something more complex than simply being aware of each other. The howls, moans, barks, yips and growls—a total of 20 different vocalizations—indicate a whole lot of social bonding is going on within the pack and between packs. This ability to communicate is why they have survived and they demonstrate their skills in communication not just vocally, but also through posture and scent. Taken together, wolves share a common language, one that software engineer Jeff Reed of the Crying Wolf Project calls “Wolfish.”

In a 2024 video, Reed says that to assume wolves talk anything like humans “is a disservice to the animals themselves” and adds, “… But I do think we can get better at speaking Wolfish. All we have to do is look to the wisdom of ancestors.” His reference is to the many Indigenous people past and present who have mastered the ability to mimic their sounds to improve hunting and as a sacred part of songs and ceremonies.

We live in a time when wild animals make up just 4% of the world’s mammals. Listening to them is getting harder and harder to do. In this spectrogram below, the Crying Wolf team has recorded the sounds that took place one evening not long ago at a wolf den site in Yellowstone National Park. The recording offers a deeper, more immersive experience of what biodiversity means in a way that words fail us.

Enough said.