Four Dedicated Individuals. One Important Mission.
The story of Scientific Learning begins with four research scientists: Michael Merzenich, William Jenkins, Paula Tallal, and Steven Miller.
When the work of these four scientists intersected, their collaboration proved that the underlying cognitive processes that influence speech and language problems could be identified—and permanently improved.
These findings led to the development of the Fast ForWord program, a groundbreaking computer-based reading intervention. The scientists then founded Scientific Learning to bring their program out of the lab and into the lives of struggling readers.
The Changing Brain
As a student at the University of Portland in the early 1960s, Michael Merzenich discovered an interest in science that led him to the field of neuroscience.
He earned his Ph.D. in neurophysiology at Johns Hopkins and went on to the University of California at San Francisco, where he pursued his interest in how the brain processes information.
Among his achievements: developing the cochlear implant, which electrically translates acoustic signals into the nerves used for hearing.
During the 1970s and 1980s, Merzenich and his colleagues at the University of California, San Francisco, ran a series of experiments designed to illuminate how the brain interpreted stimuli. They discovered that the brain actually changed physiologically when it learned or experienced something new.
More significantly, collaborative experiments by Merzenich and William Jenkins, Ph.D.—who joined the UCSF lab in 1980—showed the adult brain also demonstrated change and adaptation in response to behavioral stimuli.
“We established that the brain is modified on a substantial scale, both physically and functionally, each time we learn a new skill or develop a new ability,” said Merzenich. “Our brains were created to reinvent and reconfigure themselves throughout our lifetimes.” This ability is known as brain plasticity.
Another exciting discovery emerged when Jenkins spearheaded a study that showed progressive training could actually accelerate the rate at which the brain changed.
Auditory Processing and Language
Meanwhile, Paula Tallal was pursuing her Ph.D. in experimental psychology at Cambridge University. She theorized that the speech and language difficulties of children suffering from language-learning difficulties were related to auditory processing problems, meaning the children had difficulty distinguishing between speech sounds.
She found that that these children did well discriminating long-duration speech sounds, but had trouble differentiating rapid sounds such as the consonant sounds “ba” and “da.”
Intrigued, she used a computerized speech synthesizer to extend the duration of these quick sounds and tested the children with the modified sounds. Amazingly, the children were now able to distinguish between the sounds.
“This was a unique and completely novel finding,” says Tallal. “We were able to find the root of the difficulty in temporal-spectral processing, and you could manipulate the results by changing the duration of the sounds.”
A Meeting of the Minds
The Santa Fe Institute is a think tank devoted to fostering multidisciplinary collaboration between scientists who might not otherwise work together. In 1993 the institute held a conference at which Merzenich, Tallal, and Steven Miller—who was working with Tallal as a post-doctoral graduate student—were invited to present their research.
When she and Merzenich heard each other speak, says Tallal, it all fell into place. “It really clicked that we should work together,” she says.
Merzenich saw the possibilities, too. “Bill Jenkins and I had discussed using our training tools, as applied in monkeys, for impaired human populations, and we both realized that Paula’s kind of kid problem might be addressed with our kind of solution,” says Merzenich.
The four scientists obtained research funding to develop model training tools. Jenkins took the lead on creating the computer software that would be the foundation of the training components.
Jenkins and his team developed complex algorithms that could stretch the speed and enhance the components of speech, but the challenge was how to package the software so it would engage children.
The solution? Make a game of it.
The software component used to train the brain to increase its sampling-rate characteristics was disguised as something called Circus Sequence, while another component became Old McDonald’s Flying Farm.
Within six months, Jenkins, Merzenich, and their colleagues had a prototype product ready to go.
Putting it to the Kid Test
Tallal and Miller focused on setting up a “summer-school” study at Rutgers University designed to evaluate the efficacy of the software. They ran the first study for four weeks in July of 1994 with seven children.
Throughout the four weeks, expectations were high. So intent were they on keeping tabs on the children’s progress, Jenkins even created a software tracking tool designed to communicate daily data back to the UCSF team via the Internet. Finally, the results were in—and they were excellent.
“Six of the seven children made very substantial improvements,” says Tallal. “We were blown away.”
Added Merzenich: “The results from this initial study were excellent, wonderful."
Although the scientists were eager to bring the product to market on the strength of those results, they decided to run a second study with a larger sample group. In the summer of 1995, the product, which by then had been substantially refined by Jenkins, was set for a second test.
This study included 22 children broken into two groups matched for age, intelligence, and language impairment. One group used the software product with acoustically modified speech; the second group performed the same speech-therapy exercises using non-processed speech.
This time, the results were stunning. The children using the version with acoustically modified speech showed a huge improvement in the rate of auditory processing.
“The results were spectacular,” says Tallal. “We moved many of the kids well into the normal range, and that’s what I didn’t think could happen. I did not think you could change something as basic as a psychophysical threshold to that degree.”
Taking the Next Step
The study results were published in the journal Science in the summer of 1995, and were presented at a conference in November, which sparked an article in the New York Times. The public response was immediate and overwhelming.
“Something like 20,000 people tried to call Rutgers to get information, but we don’t know exactly how many because the phone banks blew up,” says Miller. “In the end, we took about 17,000 calls, and CNN covered it.”
Soon, Rutgers and UCSF, who jointly owned the technology and the ideas behind it, looked into licensing it for commercial use, but the quality of the licensing applicants was underwhelming.
“My view was, the companies and the kinds of software that were out there would not leverage the science correctly,” says Jenkins. “I saw a lot of edutainment software out there purporting to provide educational benefit, but the designs weren’t neuroscience-based. The science was so important and the potential was so big, it didn’t make sense to risk it not being done well.”
In early 1996, Merzenich, Tallal, Jenkins, and Miller formed Scientific Learning Corporation. The company began offering the Fast ForWord program to speech language professionals and school districts across North America, and expanded over time to more than 40 countries worldwide.
Today Scientific Learning continues to create educational software that accelerates learning by improving the processing efficiency of the brain.
The Fast ForWord® family of products provides struggling readers with computer-delivered exercises that build the cognitive skills required to read and learn effectively.
Scientific Learning Reading Assistant™ is the only reading solution to combine advanced speech recognition technology with scientifically-based courseware to help students strengthen fluency, vocabulary and comprehension to become proficient, life-long readers.
Find out more:
