WASHINGTON, D.C. — Shhhh. Listen closely — that’s the sound of people counting without using or even thinking about number words.
English speakers can identify small numbers of items placed in front of them, even as they perform a verbal task that interferes with the ability to count, according to a study presented on July 25 at the annual meeting of the Cognitive Science Society.
The finding adds to evidence that language is not required for thinking about numbers of objects, said study coauthor Michael Frank of the Massachusetts Institute of Technology. Instead, number words are abstract-thinking tools that allow people to manipulate and remember exact quantities with greater efficiency by building on basic, nonverbal number knowledge, Frank proposed.
Science News headlines, in your inbox
Headlines and summaries of the latest Science News articles, delivered to your email inbox every Thursday.
Thank you for signing up!
There was a problem signing you up.
MIT cognitive scientist Edward Gibson directed the new investigation.
Subscribe to Science News
Get great science journalism, from the most trusted source, delivered to your doorstep.
“When prevented from using numbers, people’s underlying numerical competence remains unchanged,” Frank said.
Language may expand on basic, nonverbal knowledge in other areas, such as color perception, spatial navigation and inferences about others’ beliefs and desires, he added.
Gibson’s team already reported that the Pirahã, an Amazonian hunter-gatherer group, lack number words but accurately placed small numbers of uninflated balloons next to corresponding small numbers of spools (
SN: 7/19/08, p. 5
). Tasks that required remembering more than about four items eluded Pirahã volunteers. In other words, they know that adding or subtracting one object from a group makes a difference, but have no linguistic way to keep track of accumulating quantities, which requires abstract thinking.
English speakers deprived of the ability to count performed much as the Pirahã did on the same tasks, although U.S. participants did better at remembering quantities greater than four, Frank explained. The scientists recruited 20 Boston-area participants, ages 18 to 50.
Participants listened to BBC Radio clips and repeated what the announcer said as quickly as possible while matching numbers of balloons to numbers of spools placed on a table by an experimenter. The verbal tasks sapped mental resources needed to count verbally.
In some trials, an experimenter placed varying numbers of spools, from one to 10, in a right-to-left line in front of participants. In other trials, spools ran vertically away from participants, spools were hidden from sight after being set down or spools were dropped one at a time into a can.
Like the Pirahã, U.S. volunteers made few errors when matching corresponding quantities of balloons to spools in plain view. On the other tasks, which required remembering quantities of spools, English speakers made relatively more errors but still outperformed the Pirahã.
Given their linguistic experience with numbers, English speakers may have used nonverbal mental strategies on difficult tasks that would simply not occur to people who use no number words, Frank suggested. It’s possible that U.S. volunteers perceived three or four items at a time — a rapid, nonverbal process called subitizing — in longer strings of items and then combined subitized sets.
Core, nonverbal numerical abilities operate in two systems that earlier studies have identified in people and in many nonhuman animals, Frank noted. One system visually tracks the identity of no more than three or four discrete items. The other system estimates large, approximate quantities of items.
The question of whether language orchestrates thought or thought exists outside of language has sparked intense controversy for more than 100 years. Frank said he hopes his team’s findings will lead to a “middle-ground” approach that grants substantial influence both to underlying nonverbal capacities and to related linguistic concepts.
Language may build on nonverbal numerical systems as Frank proposes, but language may influence other forms of thinking without having to exploit a bed of nonverbal knowledge, remarked StanfordUniversity psycholinguist Lera Boroditsky. She studies how spatial representations of time — whether it’s assumed to move forward or upward — differ in certain languages, such as English and Mandarin Chinese.