Brains show signs of two bilingual roads.

Unlike people who become bilingual after childhood, those who learn a second language at an early age rely on the same critical patch of brain tissue when speaking either tongue, according to a new study.

Adult learners of language apparently recruit nearby groups of brain cells, suggest neuroscientist Joy Hirsch of Memorial Sloan-Kettering Cancer Center in New York and her colleagues.

"On the basis of our findings, the distinction between native and second languages may be less for [people who had] younger ages of exposure to a second language," Hirsch holds.

According to her study, bilingual individuals who acquired a second tongue during childhood display elevated activity in the same part of Broca's area--a frontal lobe structure considered crucial for language use--regardless of which language they use. In contrast, people employing a second language acquired later exhibit neuronal bustle in another segment of Broca's area, the researchers report in the July 12 NATURE.

Wernicke's area, located in the temporal lobe and also known to perform language functions, displayed comparable responses in both groups.

The researchers relied on a noninvasive technique known as functional magnetic resonance imaging (fMRO to study changes in blood flow in the brains of 12 bilingual adults. Half of the group had learned a second language starting in infancy, while the remainder attained fluency as teenagers.

Together, the volunteers speak 10 native and second languages, including English, French, and Turkish. The two groups reported roughly equal fluency and frequency of use for their second tongues.

Researchers obtained brain scans as participants silently recited, first in one language and then the other, brief descriptions of an event from the previous day.

The findings may reflect either the sensitivity of part of Broca's area to language exposure during childhood or the existence of marked differences in the ways that children and adults learn languages, Hirsch says.

"These new results are interesting but inconclusive," comments neuroscientist Robert J. Zatorre of the Montreal Neurological Hospital. "It's devilishly difficult to study naturalistic types of language in a well-controlled way."

For instance, the short descriptions of personal events offered by volunteers in Hirsch's study allow for large individual differences in the amount of mental imagery generated during the task and the extent to which events sparked emotional reactions. Such differences may have influenced language-related brain activity, Zatorre contends.

Unpublished fMRI data obtained from bilingual speakers as they name various objects, a more restricted verbal task, yields the same disparity regarding age of learning a language, Hirsch responds.