Multilingual Miracles
In Swathi Kiran’s Aphasia Research Laboratory last summer, ten people learned to speak again—in two languages. A speech pathology professor, Kiran is developing treatment programs for bilingual patients who have aphasia caused by stroke.
Swathi Kiran
Aphasia is the loss of language—an impaired ability to understand or produce words or sentences. Among the approximately 20 percent of stroke victims afflicted with aphasia, some have difficulty in reading or writing, others in finding the words to identify objects or ideas, and still others in putting sentences together. A few lose the ability to understand language entirely.
Though no single agreed-upon treatment program currently exists for aphasia, therapy for the disorder has traditionally paralleled the progression of normal language learning: patients are retaught language beginning with simple words and concepts and progressing through levels of increasing complexity. But Kiran has found that aphasic patients progress most efficiently when their course of treatment begins with more complex concepts.
In reteaching an aphasic patient the words for various foods, for example, rather than begin with the simplest, or most typical, examples—carrot, cucumber—Kiran introduces the words for less commonly used items within the category, such as parsley and pumpkin. Patients taught in this way improve not only in identifying the foods whose names they relearn, but also in naming the more familiar ones that are not taught, the carrots and cucumbers.
Starting with more difficult tasks may seem counterintuitive, but Kiran explains that aphasic patients—who once were fluent in the language or languages they are being taught—are in a very different situation from people learning a language for the first time.
“Following the stroke, these individuals have lost access to certain aspects of their brain—to certain networks of information,” she says. “But it doesn’t make sense to treat them like blank slates, to teach them as though they’re children learning a new language. At a fundamental level, the goal is not to reteach but to facilitate reorganization and reaccess.”
Kiran is now extending her research to aphasic patients who were English-Spanish bilingual before their strokes. In the United States, bilingual aphasia patients are frequently treated in English, regardless of their pre-stroke fluency, simply because the available clinicians tend to speak English. Kiran’s work seeks to determine if this is the best way to help these patients relearn language.
Her experiences with a stroke patient named Ricardo in 2002 suggest that it may not be. Kiran began providing therapy in English, Ricardo’s stronger language, but found that his progress was slow and he failed to improve in Spanish, a language he had spoken fluently as a child, at all. “When we realized he wasn’t improving in Spanish, we switched to providing therapy in Spanish instead of English,” Kiran recalls. “And we found that as he improved in Spanish, he also improved in English—in fact, he improved more than he had during the English treatment.”
The parallel to Kiran’s previous semantic-complexity research was clear: The more difficult work, learning in the weaker language, proved to facilitate relearning in the stronger language.
Kiran will continue her work with Spanish–English bilingual patients to find out whether Ricardo’s experience is reproducible across a broader population. In addition to a systematic clinical study in which she and her colleagues will provide therapy in each patient’s weaker language, she plans to use fMRI (functional Magnetic Resonance Imaging) neuroimaging of her patients as they recover to better understand which parts of the brain they are using as they relearn language.
Axial fMRI images show patterns of neural activation in two patients with bilingual aphasia, both of whom were more proficient in English than in Spanish prior to a left-hemisphere stroke. Areas activated during word processing in Spanish and English are shown in blue and red, respectively; horizontal lines at right indicate the sections of brain illustrated.
MRI scans courtesy of Swathi Kiran
To extend the reach of her research, Kiran is also working with a colleague at the University of Texas at Austin to develop a computer model that simulates a bilingual person learning two languages at differing levels of proficiency. The model can then be given a simulated aphasia-causing stroke, allowing the researchers to test various therapeutic methods to see which provide the greatest improvement in both languages.
“To make a conclusive study of these theories, I’d need to study 1,500 aphasic patients,” says Kiran. “But with a model, I can do this theoretically. As I work with real patients, I can compare their outcomes with the model’s outcome, to see how effective a predictor the model is,” and refine or recalibrate it as needed.
That interplay between theory and clinical practice provides an apt parallel to Kiran’s research. She uses theoretical knowledge of how language is organized in the brain to create more effective strategies for providing therapy to her aphasic patients. Her clinical research, in turn, has the potential to offer new insight into neuro-plasticity—the brain’s process of restructuring itself in learning or in response to injury—as well as the very nature of how the human brain processes language.