One of the key differences between humans and non-human animals, it is thought, is the ability to flexibly communicate our thoughts to others. The consensus has long been that animal communication, such as the food call of a chimpanzee or the alarm call of a lemur, is the result of an automatic reflex guided primarily by the inner physiological state of the animal. Chimpanzees, for example, can't "lie" by producing a food call when there's no food around and, it is thought, they can't not emit a food call in an effort to hoard it all for themselves. By contrast, human communication via language is far more flexible and intentional.

But recent research from across the animal kingdom has cast some doubt on the idea that animal communication always operates below the level of conscious control. Male chickens, for example, call more when females are around, and male Thomas langurs (a monkey native to Indonesia) continue shrieking their alarm calls until all females in their group have responded. Similarly, vervet monkeys are more likely sound their alarm calls when their are other vervet monkeys around, and less likely when they're alone. The same goes for meerkats. And possibly chimps, as well.

Still, these sorts of "audience effects" can be explained by lower-level physiological factors. In yellow-bellied marmots, small ground squirrels native to the western US and southwestern Canada, the production of an alarm call correlates with glucocorticoid production, a physiological measurement of stress. And when researchers experimentally altered the synthesis of glucocorticoids in rhesus macaques, they found a change in the probability of alarm call production.

The wolf's howl is commonly thought of as indicating social separation. In popular culture, it is often used to reflect sadness, or loss. It was therefore not surprising that scientists at Austria's Wolf Science Center noticed that wolves howled when one of their packmates was separated from the pack.

The wolves that live at the Wolf Science Center, just forty kilometers north of Vienna, are hand-raised by humans for the first five months of their lives, before being introduced into a pack. This makes it easier for biologists and psychologists to safely interact with them once they become mature. The scientists at the Wolf Science Center are trying to understand how wolves - mythical animals who are wary of humans - transformed into the curly-tailed floppy-eared balls of fur that sleep curled up at the feet of our beds. What makes a wolf more wolf-like and a dog more dog-like? They realized that the howls of their wolves could contribute to the question of whether animal communication can be flexible or intentional.

Since the wolves live in large enclosures in packs of just 2-3 individuals, the keepers, trainers, and scientists who work with them regularly take them out on leashes for extra exercise. The remaining wolves always howled. New research published by Francesco Mazzini and colleagues in the journal Current Biology explains why.

To see whether howling could be the result of a physiological stress response associated with social separation, twenty minutes after a wolf was removed from a group, saliva was collected from its packmates. Previous research had demonstrated that the canine physiological stress response peaks roughly twenty minutes after a stressful situation. For the entire twenty minutes, the researchers also recorded all vocalizations made by the remaining wolves.

When the wolf that was removed was socially dominant, the remaining wolves howled. This isn't entirely surprising, given the centrality of social dominance to wolf life. However, when the wolf that was removed was a close friend, dominance notwithstanding, the remaining wolves howled even more.

Stress alone couldn't explain this pattern of results. While this sort of social separation was reflected in a salivary cortisol increase, the physiological stress response did not vary in sync with the wolves' howling response. While the separation was stressful in general, the howling itself was indicative of social dynamics rather than a more basic physiological reflex.

Mazzini writes that "social partner preference," or friendship, "is a more dynamic and flexible feature of wolf life [than dominance] and thus is more likely to be modulated by cognition." This explains why wolves howled more when their friends were removed than when dominant individuals, who perhaps were not their closest friends, were removed, despite the equivalent change in cortisol levels. "This provides strong support for the hypothesis that wolf howling is potentially a strategically employed vocalization with the goal of ultimately promoting contact with important individuals."

This study provides further evidence that not all animal communication is the result of automatic, inflexible physiological events, but can be intentional and voluntary. For wolves, it is important to maintain contact with allies, even if when they're out of visual range. One way they do this, apparently, is by playing a wolfy version of "Marco Polo."

Mazzini F., Townsend S., Virányi Z. & Range F. (2013). Wolf Howling Is Mediated by Relationship Quality Rather Than Underlying Emotional Stress, Current Biology, DOI:

Elsewhere on SciAm

Wolves Howl for Pals, Not Leaders - 60-Second Science Podcast

For more on animal communication:

Eavesdropping Lemurs Tune Into the Forest Soundtrack to Survive

Koalas and Bison Use the Same Rules for Choosing Mates

Singing Mice May Join Humans and Songbirds as Vocal Learners

Eavesdropping Iguanas Use Mockingbird Calls To Survive

Can You Hear Me Now? Human Noise Disrupts Blue Whale Communication

For more on wolves:

Dogs, But Not Wolves, Use Humans As Tools

Photo via Mazzini et al.