The grey wolf trio, Idaho, USA. Photo by Jim and Jamie Dutcher/National Geographic Creative.
As the heroes fled
the dark castle for the darker woods, Count Dracula’s ‘children of the
night’ began to make their ‘music’: a distant chorus of lupine howls, echoing through the Transylvanian night.
I paused the movie. ‘That’s not a European wolf, the howl’s all wrong!’
I told my long-suffering companion. ‘That wolf belongs in the backwoods
of California!’
After
hundreds of hours listening to thousands of wolves for my PhD, the
difference between howls was obvious. The voice of a Russian wolf was
nothing like that of a Canadian, and a jackal was so utterly different
again that it was like listening to Farsi and French. I believed that
there must be geographic and subspecies distinctions. Other researchers
had made this proposition before, but no one had put together a large
enough collection of howls to test it properly. A few years later, my
degree finished, I told my Dracula story to the zoologist Arik
Kershenbaum at the University of Cambridge. He promptly suggested we
explore how attuned to wolves I really am. Are there differences between
canid species and subspecies and, if so, could these reflect diverging
cultures?
When
animals call to each other, they are communicating in a single stream of
information from caller to listener. Until modern recording technology
was invented, any acoustic communication lasted only as long as the
echo. So while we can hear difference in modern human speech, with more
than 6,000 extant languages and an unknowable number of local accents
for each language, we can’t trace the origin of language from before
writing or know how ancient peoples would have sounded. Before 1860,
when de Martinville made the first acoustic recording, the world of
speech must remain silent to us, though we can sometimes hear scattered fragments of dead languages still alive in our own.
The
question of when and how language first emerged is the topic of
tremendous controversy – it has even been called ‘the hardest question
in science’. My work is on what information can be extracted from
vocalisations. It is a first step in understanding where the physical
body dictates the shape and form of the call, and where the caller has
control. For example, a piano player is limited to combinations of a
piano’s 88 keys, but a song played on a Steinway will have different
sound qualities to the same song on a bar’s upright. In addition,
different tunes can also be played. Separating the characteristics of
the instrument from the choices of the player is essential before we can
understand what meaning those choices might convey.
Behaviour
is not purely instinctual, bred in the bone and performed from birth
without thought or flexibility. It is often learned socially.
Chimpanzees are a fascinating example: their use of tools spreads from
one individual to another as they copy the successful tactics of their
troop-mates in breaking apart nuts, catching ants or cleaning their
teeth. They’re shown to prefer cooked food over raw, and are even able
to learn US sign language. All this has increased interest in how other
species develop shared cultures and knowledge. Whether it’s tool use in
birds, farming by ants, or dancing in parrots, activities that were
previously believed to be specific to humans are now being found in a
variety of species. This means that animals can be used as a model for
humans, allowing us a window into an otherwise cryptic part of our own
evolution.
Dialects,
or regional differences in the form and use of vocalisations, have been
observed in birds, bats, chimpanzees and now an increasingly long list
of other species. This has been most beautifully heard in whales, where
the songs of humpbacks are transmitted across hundreds of miles, telling
a listener which part of the ocean the whale lives in, and tracing its
family group by the influences on song formations. The bioacousticians
Katharine Payne and Roger Payne first listened to the whales on
underwater microphone recordings in the 1960s, and used musical notation
to explore the changes that occurred in each male’s song, year on year.
Whalesong, heard by humans as long ago as Aristotle, became the subject
of intense study and public interest. Their research
showed that there were geographic differences in humpback whale songs
and that we could tell apart populations just by using those songs,
which change throughout their lives. So the whales were controlling
their singing and subject to cultural influences. The Paynes had found
dialects in whale song. Would we find the same for canids?
Despite
their cultural popularity, wolf howls haven’t been the subject of
focussed research until recently. Now, following the lead of marine
biologists and ornithologists, and with improved sound recording
equipment and analysis programs, researchers can study them in depth.
The first step in understanding what animals are saying to one another
is to figure out what aspects of the voice are functional and what parts
are formed by the structure of the throat and mouth, or what is the piano and what is the tune.
Studies since the 1960s have shown that the howls that have haunted our
dreams for centuries can tell us a lot about the particular wolf
vocalising. Like humans, each wolf has its own voice. Each pack also
shares howl similarities, making different families sound distinct from
each other (wolves respond more favourably to familiar howls). This much
we knew. What we didn’t know was whether the differences seen between
packs were true of subspecies or of species, and if an Indian wolf howl
would be distinct from a Canadian one.
More
questions follow. If howls from different subspecies are different, do
the howls convey the same message? Is there a shared culture of
howl-meanings, where an aggressive howl from a European wolf means the
same thing as an aggressive howl of a Himalayan? And can a coyote
differentiate between a red wolf howling with aggressive intent and one
advertising the desire to mate? Even without grammar or syntax, howls
can convey intent, and if the shape of the howl changes enough while the
intent remains constant, the foundations of distinctive culture can
begin to appear.
The wolf species were like music bands with preferred styles of playing: riff-filled like jazz or the pure tones of classical
To
explore this, Kershenbaum brought together a group of researchers to
share data and ideas. We compared howls across 13 different subspecies
and species of coyotes, dogs, wolves and jackals (collectively known as
canids). The howls came from my own recordings on chilly evenings in
Poland and Russia; from UK zoos and conservation sites (where I howled
at the bemused wolves and listened in awe as they responded); from our
co-authors in the US, Spain and India; from historic recordings taken
across continents and time; from the public, in the form of hundreds of
YouTube videos of howling pets. Our canine voice collection represented was one of the most comprehensive ever.
We
then stretched all the howls to the same length, using a process called
dynamic time warping, to compare the changes in the tune without
including the tempo it was played at. We found that each species had its
own favourite howl shape, a preferred set of changes to their howls to
raise and drop the pitch, but that they also used howl shapes preferred
by other species, and varied the shapes as they pleased. The species
were like music bands with preferred styles of playing, whether
riff-filled like jazz or the pure tones of classical, but were flexible
in what they actually played at any given time. So while they had a
favourite style, the tune itself varied.
Like
musicians, the wolves were influenced by their forebears in the genre,
and species shared traits with other canids that were closer to them
geographically and genetically. An Eastern grey wolf, recorded in the
US, sounded more like a North Carolinian red wolf than a European wolf,
and an African jackal sounded quite different again. Small and delicate
compared with their cousins the European wolves, golden jackals have
high, rising howls, running up and down the scales in bravura
performances of control and speed, but with less variation in overall
shape, whereas the European wolves used a slower style of deep and
steady long notes ending in falls that seem to drift away into the
night. New Guinea singing dogs earned their names with a large vocal
repertoire and a wide selection of howl shapes. While sometimes the
different species achieved crossovers to other shapes, most had a style
that dominated their repertoires.
If
these differences across species sound familiar, they should. We’ve
known for thousands of years that birdsong is distinct to each species,
and sometimes even populations, with a nuthatch’s wha-wha-wha
very different to a robin’s whistling call. We’ve seen that birds have
adaptable repertoires, literally changing their tunes as new sounds
become popular and spread through populations. Humpback whales sing new
songs when they hear them, collecting new patterns of song throughout
their lives and passing them on to others in their population. Our canid
study
showed that they had different howls for their species, but we have yet
to answer whether they can change their howls with time or exposure to
different howl patterns. Now that we’ve seen there are differences, the
next question to answer is whether they are innate or learned, and how
far a wolf can change its howl.
To
understand how well wolves maintain their own styles when exposed to
others, we did a smaller analysis of red wolves, Eastern grey wolves and
coyotes. Wolves across the world are subject to conflict with humans,
and red wolves are critically endangered with numbers that were once as
low as 20 individuals in the 1970s. Huge effort has been put into saving
the red wolf and returning it to its old hunting ranges of the
southeastern US. But these wolves have a worrying propensity to mate
with coyotes and interbreed with Eastern wolves, resulting in fertile
hybrids. We hoped that the red wolf would have its own robust howl
shape, separate from its neighbours’ and as distinct as the blues are
from pop, which might mean that wolves could create their own
communicative barrier to reproduction.
Unfortunately,
that wasn’t the case. Like any good musician, the red wolves were not
averse to using novel styles, and we found that they fitted at a midway
point between coyote and Eastern wolf howls, showing that the howls
wouldn’t form a barrier to their breeding. As the red wolves we studied
were as genetically pure as possible, this suggests that they might be
influenced by both coyotes and Eastern wolves, copying the sounds they
hear around them. Or perhaps they never had a distinctive, red wolf howl
shape and simply shared that of their near relatives. We will never
know if there were other howl shapes of red wolves that were overwhelmed
by hearing coyote and Eastern wolf howls too often – as lost today as
early delta blues recordings – but the modern differences aren’t
stopping cross-pollination for both sound and puppy creation.
The
range of howl shapes we found is enough to encompass a large number of
possible meanings. However, a howl is by its nature a public shout, not a
private whisper, and that limits what is likely to be communicated.
We can guess, but don’t currently know, what these meanings are.
Perhaps there are subtle differences between ‘All pack come here’ and
‘Brother wolf here’, and potential vast differences between ‘Stay away,
stranger’ and ‘Good food here, sister’. Beyond the interests in
language, this matters to conservationists, some of whom have tried to
protect livestock by playing recordings of wolf howls to scare away
local wolves, but they might be using howls that ring the dinner bell
instead of sending a keep out signal.
is the territorial ‘get off my forest’ howl different to the lonely hearts ad of a lone wolf?
Some
of these howl shapes were shared across species, while others were
distinctive. While some of these differences between species might be
the result of genetic drift, random mutations leading to changes that
spread throughout a population, they might also develop in response to
need, or to the shaping influence of the animal’s habitat. Where calls
are flexible and shift with new experiences, as in the humpback whale
song, they can be used to illuminate the possible evolution of language.
Human accents and word usage shift as the speakers come into contact
with new ideas and they adopt new sounds, sometimes a little mangled,
into their own vocabulary.
The
next step is to show whether there are behavioural contexts to howls,
demonstrating if wolves, like other animals, can convey emotional states
and information to their listeners, whether a territorial ‘get off my
forest’ howl is different to a lonely hearts ad of a lone wolf searching
for a new mate, whether a call to the hunt differs from the
advertisement of a successful kill, and whether pups must learn these
shapes or know them by instinct. We’ll also explore how long these
differences persist through time and why some howl shapes are shared.
The
howl isn’t an example of true language as humans understand it, with
syntax and grammar where a single additional sound or stress can change a
phrase’s entire meaning. However, if we can show that howls are not an
instinctual but a learnt display, it will give us another potential
piece of the puzzle of language evolution. Wolves and humans both act
cooperatively, living in complex societies with multiple individuals in
close association and where there are great advantages to communicating
complex meaning. Humans use language to convey meaning through a huge
variety of sounds, but proto-words probably developed in association
with very basic ideas and intents that then evolved towards complexity.
More closely related languages tend to be more similar, with French and
Italian sharing enough words to be intelligible, but incomprehensible to
speakers of Hindi, yet they can still all be linked back to a
proto-Indo-European language.
Isolation
and geographic distance have meant that human language has diverged
multiple times, creating thousands and thousands of dialects, many with
words distinct to the environment in which they arose. Yet certain words
are so basic that they have barely changed over thousands of years, eg
the word mother, which is ‘matar’ in Sanskrit, ‘mater’ in Latin
and ‘meter’ in Ancient Greek, and ‘mzaa’ in Swahili. The word shows its
original roots in a possibly universal proto-language even today, while
the words for more complex ideas are more typically unrelated. Perhaps
the wolves mirror us, the shared howl shapes representing a similarly
universal concept as mother, with the more divergent howls
relating to local concepts. By exploring this, we can explore the first
steps towards true language.
Have
wolves evolved to convey meaning in their calls? I don’t know for sure,
but I think so. To my ears, a happy wolf surrounded by a pack has a
very different howl to a lone voice crying in the wilderness, and a love
duet with a mate does not match the chorus howl made with the new pups
raising their tiny heads to the sky, but all the sounds are beautiful.
Perhaps one day we will even understand their meaning. For now, I can
only do as the wildlife ecologist Durward L Allen described, and listen
‘for a voice crying in the wilderness, and [hear] the jubilation of the
wolves!’
No comments:
Post a Comment