Skip Navigation

Maths skills survive linguistic damage

February 14, 2005 By Philip Ball This article courtesy of Nature News.

Different processes underpin the grammars of numbers and language.

An inability to process language needn't stop you from doing maths, UK researchers have found.

They say that three men with severe aphasia, a linguistic impairment, can understand 'grammatical' rules in mathematics even though they cannot handle analogous rules in language. Aphasia leaves people unable to use or comprehend words, and is often triggered by stroke or other brain injuries.

The discovery challenges a commonly held view that linguistic and mathematical mental processing draw on the same cognitive resources. "Our findings very strongly turn that idea on its head," says Rosemary Varley, a cognitive neuroscientist at the University of Sheffield, UK. The research is published in Proceedings of the National Academy of Sciences1.

Their natural language grammar is blown to pieces.
Rosemary Varley
University of Sheffield, UK
According to the view of cognition developed by linguist Noam Chomsky, language processing is a fundamental skill that is used for related grammatical tasks in the brain, such as certain mathematical ones. Previous studies of the relationship between linguistic and mathematical ability have lent some support to this notion.

For example, brain-imaging studies have shown that some areas of the brain involved in language processing also become active when people are performing mathematical tasks. But it has been unclear whether this use of neural language centres is essential for maths: there are also indications that the two mental functions can work independently.

Blown to pieces

Varley and colleagues set out to clarify the picture by asking patients with aphasia to carry out some mathematical calculations. All three of their patients were aged between 50 and 60 and were well educated; one had been a university professor before incurring a brain lesion.

The three men were almost entirely unable to communicate verbally or in writing. "Their natural language grammar was blown to pieces," says Varley. But they clearly retained much of their higher mental functioning and could communicate, for example, through "highly expressive" drawings, she says.

Their impairments left them unable to decode the grammatical relationships of simple sentences. For example, they had great difficulty distinguishing subject and object in the phrase "The boy chased the girl," which they were unable to differentiate from "The girl chased the boy."

Analogous object-relation problems are posed by mathematical expressions such as 90 ÷ 30 and 30 ÷ 90. More complex expressions might involve problems of nesting, such as the use of brackets (90 - [(3 + 17) x 3]), which mirror linguistic sub-clauses and embedded relations, such as in the phrase "The man who killed the lion was angry."

Mathematical grammar

Although the patients were unable to decode such linguistic expressions, they were all able to perform the mathematical calculations accurately with pen and paper. They could interpret Arabic numerals correctly even though they struggled to understand number words such as 'three' or 'ninety', when spoken or written.

Varley speculates that it might be possible to use this capacity for mathematical grammar to help the patients find a way to interpret speech and the written word.

But even if that remains a remote possibility, the study had a much more immediate benefit for the patients. "They really enjoyed doing these experiments", says Varley. "It was a joy to them to demonstrate competence."


  1. Varley R. A., Klessinger N. J. C., Romanowski C. A. J. & Siegal M. Proc. Nat. Acad. Sci. USA published online. doi:10.1073/pnas.0407470102 (2005).


Need Assistance?

If you need help or have a question please use the links below to help resolve your problem.