Press "Enter" to skip to content

Why speech is a human innovation

This story was originally published by Knowable Magazine.

Except for various cartoon characters, the Geico Gecko and Mr.
Ed, animals can’t speak. Yet they have a lot to say to scientists
trying to figure out the origins of human language.

Speaking isn’t the only avenue for language. After all,
linguistic messaging can be transmitted by hand signals. Or
handwriting. Or texting. But speech is the original and most basic
mode of human communication. So understanding its origins ought to
generate deeper comprehension of language more generally. And a
first step toward that understanding, cognitive scientist W.
Tecumseh Fitch believes, is realizing that key aspects of vocal
language are not, as traditionally contended, limited to

He’s not talking about a TV-show horse, of course, or animated
narrators of insurance advertisements. Fitch’s point is that many
creatures from the real-world animal kingdom offer clues about how
the capacity for speech came to be.

It’s true that humans, and humans alone, evolved the complex set
of voice, hearing and brain-processing skills enabling full-scale
sophisticated vocal communication. Yet animals can make complicated
sounds; parrots can mimic human speech and cats can clearly convey
that it’s time for a treat. Many animals possess an acute sense of
hearing and are able to distinguish random noises from intentional
communication. So even though only humans possess the complete
linguistic package, the components of language ability “have very
deep evolutionary roots,” says Fitch, of the University of Vienna.
In fact, he suggests, just a handful of changes in the
communication repertoire of humankind’s ancestors endowed people
with the full faculty of language.

You May Also Like

Much of the physiological apparatus for hearing and speaking is
found in all land-dwelling vertebrates — the tetrapods — including
mammals, birds, amphibians and reptiles. “Humans share a
significant proportion of our basic machinery of hearing and vocal
production with other tetrapods,” Fitch writes in the Annual Review of Linguistics.

Life-forms occupying numerous branches of the tree of life
possess anatomical tools for producing and perceiving vocal
communication. Where human ability exceeds our predecessors, Fitch
says, is the sophistication of the brain circuitry adapted to the
uniquely human capacity for complex linguistic expression.

Historically, language experts have proposed anatomical
explanations for human’s special language facility. Just as the
opposable thumb permitted tool use, some authorities theorized that
the lower location of the voice box in the vocal tract enabled the
articulation of meaningful sounds. Or the human hearing apparatus,
encompassing hair cells and eardrum and three little bones,
provided the discerning ear capable of interpreting nuanced
vocalizations. But in reviewing the scientific literature, Fitch
finds that speech’s structural subcomponents, used for producing
vocalizations and perceiving patterns in those sounds, have
appeared in multiple organisms over evolutionary time.

An anatomical view of the human head and throat highlight some key structures involved in speech, including the larynx, or voice box, vocal chords (folds), windpipe, tongue and lips.

Among primates, only humans can learn to produce novel vocal
sounds, but that difference isn’t explained by anatomy — the basic
structure of the human voice box and vocal tract is similar to that
found in other mammals. Cartilage and muscle within the vocal folds
of the voice box (or larynx) give mammals better control over
vocalizing than other vertebrates. Fleshy tongues and lips are also
mammalian features that aid in speech production.


The ear’s sensory hair cells, which convert sound vibrations
into nerve impulses, go back as far as jellyfish, for instance.
Genes instrumental to producing the hair cells are similar in
insects and humans.

In some cases, a particular trait evolved independently in
different lineages. But often a trait evolved once and then was
passed down through a long line of descendants. Such “homologous”
traits “provide the equivalent of a time machine allowing us to
reconstruct an evolutionary sequence of ancestral forms,” Fitch
notes. Independently arising traits, on the other hand, provide
data points helpful for testing evolutionary hypotheses. Combined,
the homologous inherited traits and the independent analogous
traits have produced deep and novel insight into speech’s
evolutionary origins.

Among the tetrapods, mammals evolved much more sensitive
hearing, able to cope with a wider range of frequencies and
therefore more able to process nuances of vocalizations.
Humankind’s primate ancestors, for instance, possessed highly
capable hearing ability. “There is nothing about the human ear that
is strikingly different from that of other primates,” Fitch writes.
“Our peripheral hearing apparatus was in place, in our primate
ancestors, in essentially modern form long before we evolved the
capacity for speech.”

But perhaps successful speech perception required “vocal tract
normalization” — the ability to recognize the same words spoken by
different voices (such as a child versus an old man). Humans are
not, however, alone in that ability, either. Zebra finches trained
to recognize vowels when listening to a male voice can still make
the distinction when the speaker is a woman.

Maybe the key human-only skill is the ability to figure out
which of the world’s many complex sounds are vocal efforts to
communicate. In the part of the human brain that responds to sounds
(the auditory cortex in the temporal lobe), some of the circuitry
is specialized for voices as opposed to other sounds. But such
voice-specific circuitry also exists in nonhuman primates and
perhaps even dogs. “The data lead to the conclusion that the
primate auditory system had already evolved to a ‘speech-ready’
level of sophistication long before spoken language evolved in our
species,” Fitch writes.

If hearing skill isn’t the source of human linguistic power,
maybe the human-only aptitude for speech lies in the ability to
produce it. Nonhuman primates can make vocal noises, but unlike in
Planet of the Apes movies cannot articulate the nuanced
sounds of speech. But it’s not obvious why not, as the basic
blueprint for the human vocal tract has been around for 70 million
years and is shared by most mammals. Even the lower position of the
voice box — the descended larynx — is not exclusively human. And
that anatomical adjustment isn’t necessary for complicated
vocalization, anyway. Experiments have shown that some primates
have vocal tracts capable of ample vocal agility.

“An unmodified primate or mammal vocal tract would be perfectly
adequate to produce intelligible spoken language,” Fitch

Evolutionary tree showing the descent of major groups, called clades, of human relatives from a common ancestor. Clades listed include (from bottom to top) eukaryotes, vertebrates, tetrapods, amniotes, hominids and hominins. Scientists have used such groupings for insight into the evolution of spoken language in humans, a trait not shared by any other animal.

By studying clades — groups of species related by a common
lineage of descent (simplified tree shown here) — scientists can
compile clues about the evolution of speech. Many aspects of human
speech and hearing, for instance, rely on features found in
all tetrapods, a clade that includes mammals, reptiles and
amphibians. Of particular interest are homologous and analogous
traits. All mammal species, for instance, have three middle-ear
bones, a homologous trait inherited from a common ancestor. Neural
connections between parts of songbird brains important for
vocalization may be analogous to neural connections between
speech-related parts of human brains; those connections evolved
independently in different lineages but may both be important for
speech production.

Besides all that, parrots and many other bird species, some bats
and even elephants can mimic vocal sounds. So humans’ distinctive
speech can’t depend solely on vocal production ability. Considering
all the evidence, the vocal and auditory skills of various animals
tell a tale of multiple preludes to the human speech story. That
tale reveals that humans acquired speech not via anatomical
innovation for vocalizing and hearing, but by novel neural
connections that control the anatomical hardware.

After all, speech requires more than producing and perceiving
sounds. A speaker’s brain must decide what sounds to produce and
issue instructions for producing them to the body’s vocal
apparatus. And a listener’s brain must be able to decode auditory
signals it receives and then issue commands for a vocal response.
People are skillful at producing sounds in response to other sounds
— it’s why you can repeat a word out loud after the first time you
hear it.

Such controlled vocalization of a word is different from just
making noise. Most animals possess neural circuitry for producing
“innate” vocalizations: Dogs bark, squirrels chatter and seagulls
squawk. Even humans have their own innate vocalizations, including
crying, laughter and screams. But among primates, only humans have
the “capacity to produce novel, learned vocalizations beyond the
innate call repertoire,” Fitch notes.

Today the dominant hypothesis explaining that ability is the
presence of special connections between brain regions involved in
controlling speech and hearing. Innate calls — in humans and all
other mammals — are initiated by direct signals from the brain
stem. Indirect messaging from the cortex (the brain’s more advanced
outer layer) enables voluntary suppression or production of innate
calls. Unlike other animals, humans possess direct connections
between nerve cells in the cortex and the nerve cells that control
the muscles operating the larynx. Some apes and monkeys have direct
connections from cortex to the muscles controlling the lips and
tongue, but not to the muscles controlling the larynx. (Circuitry
connecting the auditory cortex to the motor cortex also seem more
extensively developed in humans.)

Evidence supporting the view that such direct neural connections
explain human speech comes from other species that can “talk,” such
as parrots and songbirds that can learn novel vocalizations. These
species do have direct neural connections to their voice-generating
apparatus, while non-vocal learning birds don’t.

Underlying the evolution of the brain circuitry responsible for
human speech skill are genetic modifications that remain largely

Diagram shows simplified views of human brain. One highlights the direct neural connection linking the motor cortex to the muscles in the voice box. The second shows nerve connections, shared with primates, between Broca’s region and the auditory cortex. It also shows additional connections between these brain areas running through the parietal cortex and found only in humans.

Among primates, it seems that only humans possess direct
connections (shown at left) from the part of the brain that
controls motion to nerve cells in the brain stem (black dot) that
control the larynx muscles responsible for vocalizing sounds. Shown
at right: Other primates as well as humans possess nerve
connections (dashed line) between brain areas involved in language
(Broca’s region) and hearing (auditory cortex). But human brains
have more fully developed additional connection pathways (blue and
red lines) that researchers hypothesize play key roles in producing
and processing speech.

“The genetic underpinnings of … [neural] connections involved in
human vocal control are virtually unknown,” writes Fitch. But
genetic analyses of ancient organisms and testing DNA found in
fossils is an emerging research field. “Thus, genetic data perhaps
provide the most promising and exciting empirical pathway for
future research on the biology and evolution of speech.”

As Fitch notes, speech is not the whole story of human language.
Vocal communication is a central feature, but language encompasses
much more, as linguist and neuropsychologist Angela Friederici pointed out at a recent
meeting of the Society for Neuroscience.

“Language is more than speech,” said Friederici, director of the
Max Planck Institute for Human Cognitive and Brain Sciences, in
Leipzig, Germany. “Speech … uses a limited set of vowels and
consonants to form words. Language, however, is a system consisting
of words … and a set of rules called grammar or syntax to form
phrases and sentences.”

Nonhuman primates can learn the meaning of individual words, she
notes, but aren’t capable of combining words into meaningful
sequences of any substantial length. That ability also depends on
circuitry connecting different parts of the brain, current research
by Friederici, collaborators and other scientists is now

Understanding that circuitry depends on comparing the cellular
architecture and nerve fiber tracts of the human brain with the
brain of animals with lesser linguistic power. So in a way,
scientists may be able to ask animals for clues not only to the
evolution of speech, but to language skills more generally as well.
Sort of like going straight to the source and asking the horse.

This story was originally published by Knowable Magazine. Knowable Magazine is an independent journalistic endeavor from Annual Reviews.

Be First to Comment

Leave a Reply