Living in a world of ones and twos
A tribe without numbers
What if our language did not have words to describe numbers? Without the
words, would we know how to count?
The Piraha tribe lives along the Maici River in the Amazon region of
Brazil. They have words for one and two. For more than two, there is only a
word for "many." The Piraha language does not have "number" in its
vocabulary, and it does not have comparisons like "more sticks in this pile
than in that one."
Peter Gordon, a Columbia University professor of bio-behavioral science,
tested the Piraha's ability to count objects. Gordon thinks that his
results, which were published in the journal Science in August, support
Benjamin Lee Whorf's theory that the structure of our language affects the
way we think.
In the 1930s, Whorf compared Hopi with English by studying his New York
City neighbor, a Hopi speaker. In his book "Language, Thought and Reality,"
Whorf wrote that "all observers are not led by the same physical evidence
to the same picture of the universe." Whorf said we divide and organize the
natural world into concepts that are based on the words and patterns of our
language. He took anthropologist Franz Boas' idea, that language reflects
culture, one step further: Language can influence culture and our
understanding of the world.
Working with a group of seven Piraha men, Gordon laid out a row of
batteries, nuts or sticks in front of one person. There were never more
than 10 objects. Each person had to make another row containing the same
number of things. Gordon found that they had no problems with three or
fewer items. But as the numbers got closer to 10, they made more errors.
Gordon also observed the tribesmen's ability to compare quantities, using
two boxes that had different numbers of fish drawn on them. Those tested
had trouble distinguishing between groups of objects that were close in
"I think they could distinguish five versus eight," Gordon said. "Maybe not
five and six, though."
Babies and some animals (such as monkeys) have an innate ability to count
small numbers of things and make comparisons of large numbers in groups.
Gordon said the existence of numbers in English and other languages helps a
child develop skills beyond the limited counting ability that he is born
But whereas English allows us to exactly count beyond three, the Piraha
language does not. Gordon concludes that the Piraha people have
difficulties recognizing specific quantities greater than five because of
what's missing from the grammar and syntax of their language.
"If you don't have the counting system, it doesn't allow you to exactly
identify the numerical size of a set beyond 1 - 2 - 3," he said.
Gordon said his results prove Whorf's linguistic determinism theory, at
least in one case. Not having the words in your language to describe a
concept prevents you from knowing it.
"I want to stress that this is not about inferiority or superiority,"
Gordon said. "I do point out in the article that the Piraha [language] is
incredibly complex. It actually has things like Navajo, a very complex verb
Indeed, the Piraha have optimized the art of tonal communication for the
rivers and jungles where they live. They use cries, whistles, and "eating
speech." When in canoes spread over a river, they talk across the large
distances with cries. Whistle tones are the "code talk" of the Piraha, used
on jungle expeditions when they don't want their voices to alert animals or
be understood by humans outside their group. Eating speech, as the name
implies, lets them converse while they have food in their mouths.
The Piraha language also has an intricate naming system based on their
spiritual beliefs. While a baby is still in the womb, he receives a name to
ensure the development of his body. Throughout their lives, the Piraha
receive names from the gods which create their souls and their destinies.
"They have their way of doing it - maybe it wasn't necessary to count,"
said Ellavina Tsosie Perkins, a Navajo linguistics expert. "But there are
other means how they can take care of those things."
She said when you study another language, it's easy to jump to conclusions.
The Navajo language doesn't have words for "I'm sorry." But that doesn't
mean Navajo people never express remorse.
"You can't say 'I'm sorry'," Perkins said, "so we laugh and say 'Tso nadish
nish neh' - that's what you might hear from someone - 'I won't do it
"I can't see that Gordon has shown that the language is preventing speakers
from understanding any concepts," said Ted Fernald, a professor of
linguistics at Swarthmore College. "He has shown that you can't count
Fernald thinks that keeping track of exact numbers isn't very useful in the
Piraha culture, and therefore they haven't practiced enough to be good at
Gordon stresses that his study does not reveal any universal truth about
language. "It's completely existential," he said. "There exists at least
one case that follows Whorf, but that's it. I don't want to generalize
anything beyond that."
Ironically, Whorf's hypothesis came from his incorrect deductions about
Hopi grammar. He claimed that the Hopi people think about time differently
because they don't use past, present or future tense in their verbs. The
Hopi language actually does express time through constructs different from
the tenses used in European languages.
Did the Piraha have problems with counting because a lack of tools
(language) kept them from comprehending numbers, or did they perform poorly
because the absence of number words kept them from practicing enumeration?
Gordon's study suggests that language is an essential part of understanding
numbering. But linguists specializing in American Indian languages doubt
that the Piraha experiment, or any study, can ever define the relationship
between language and thought.