In 2006, Dr Dana Suskind, a paediatric surgeon at the University of Chicago, started performing delicate two-hour operations to attach a cochlear implant in the inner ears of babies born with hearing problems.
The implants enabled deaf and severely hearing-impaired children to hear electronically by capturing sound, translating it into digital signals and then electrical energy, and transmitting that to the auditory nerve.
But within a few years, she noticed that outcomes among her young patients varied.
Some of the children she operated on started to speak after the implant surgery, but others fell behind. What ate at her even more was that the children who tended to lag behind were the ones from poor, disadvantaged homes.
She signed up for a course on children's language development, and found her answer in a study by child psychologists Betty Hart and Todd Risley.
Their landmark study in the 1990s found that children whose parents were professionals were exposed to far more spoken words - more than 1,500 per hour, on average - than children from homes on welfare. Over one year, that amounted to a difference of nearly eight million words - by age four, that amounted to a total gap of 32 million words.
Moreover, it was not just the number of words, but also the type of words that varied. There was a substantial gap in the complexity of words being used.
The researchers also found that higher-income families provided their children with far more words of praise than low-income families did.
Key language development study
Child psychologists Betty Hart and Todd Risley studied 42 families from three income levels for three years, recording the number of words spoken to children in these households.
They found that children from all three groups started to speak around the same time.
But those in professional families heard more words per hour, which resulted in larger cumulative vocabularies.
In these families, children heard an average of 2,153 words per hour; in working-class families, the figure was 1,251 words. For families receiving welfare, it fell to 616 words.
Over a year, the average came to 11 million words in professional families, six million in working-class families and three million in welfare families.
By age four, the gap between children from welfare and professional families could have grown to 32 million words.
There were also differences in the kinds of words and statements heard. By age three, children of professionals had heard 500,000 encouragements and 80,000 discouragements on average. In contrast, welfare children heard an average of 75,000 encouraging phrases and 200,000 discouraging ones.
The researchers concluded that professional parents were giving their children an advantage with every word they spoke, and that the advantage just kept building.
Children from families with professional backgrounds experienced six encouragements for every discouragement. Those from working-class families received two encouragements for every discouragement. The ratio was reversed for those from families on welfare, who heard two discouragements for every encouragement.
Twenty-nine of the 42 families in the original study were recruited for a follow-up study when the children were in third grade. Researchers found that measures of accomplishment at age three were highly indicative of performance at the
ages of nine and 10. The foundation built at age three had an impact later on in life.
Moved by the findings, Dr Suskind started Project Aspire, an educational programme for parents of children with hearing loss. It taught parents how to provide the right language environment at home.
But she was mindful of the fact that the research did not point to a gap just for hearing-impaired children, but for all children born into families unable to provide for their healthy development.
"We look at these two-year-olds and already you can predict their life course. How can we, in all conscience, allow that?"
She launched the Thirty Million Words Initiative in 2009, with the goal of teaching parents and early-childhood caregivers how to nurture brain development through frequent, high-quality communication.
Her message is simple, but powerful.
"Children aren't born smart. They are made smart by their parents talking with them."
She adds: "Thirty Million Word Gap - it's like a headline for a news story. It grabs the attention of people. But behind it is this rich science that shows how the way parents interact with their children affects the way the children's brains develop.
"And it is really about helping parents understand the important role they can play."
The programme is delivered over a three-month period by trained staff who make weekly visits to the child's home.
The children's language development is tracked and fed back to their parents. The home visitors give each family - all of whom volunteered to participate - a tiny recording device, known as a Lena (short for Language Environment Analysis), which can be inserted into a vest worn by the child.
The recorders distinguish between words overheard from television or other electronics and words heard in live human conversations. Computer software then analyses the words spoken.
Explains Dr Suskind: "It is like a pedometer, but it keeps track of the words that children hear."
The strategies are simplified for the parents. They are told to follow three Ts - Tune In (to what the children say), Talk More (to the children using a rich vocabulary to build up the number of words heard) and Take Turns (to have a conversation with the children and actively listen to them).
Parents are taught how to incorporate rich, meaningful conversations into routine activities.
For example, while dressing their children, parents are encouraged to discuss the activity - which arm goes first while putting on a shirt, for example, or why hot weather makes shorts a good choice.
Instead of telling a child, "Put your shoes on", the parent might say instead, "It is time to go out. What do we have to do?"
"We teach them to add more language and interactions to things they're already doing, rather than adding more things to their already busy lives," says Mrs Beth Suskind, who is the project's co-director and Dr Suskind's sister-in-law.
Other topics covered include how to go on a television "diet", how to incorporate numbers into a conversation and techniques for reading to a child.
Throughout the 12-week programme, the Lena data is used to gauge progress and set new goals.
Small trials conducted by Dr Suskind's team have shown dramatic increases in parent-child interaction.
In one randomised control study involving 40 families, the experimental group received eight weeks of lessons on topics and strategies for integrating more words into daily life. During the intervention, in addition to speaking more with their kids, parents in this group significantly increased back-and-forth conversation.
Another study published in a journal two years ago showed how the programme could help children from affluent homes.
Optimising ability to learn
Dr Dana Suskind is professor of surgery at the University of Chicago, director of the Paediatric Cochlear Implant Programme, and founder and director of the Thirty Million Words Initiative.
Based on scientific research that shows the critical importance of early language exposure for the developing child, Thirty Million Words helps parents enhance their home language environment in order to optimise their child's brain development and thus the child's ability to learn.
An evidence-based intervention programme, Thirty Million Words is supported by a broad coalition of public and private partnerships. The initiative is an extension of Dr Suskind's Project Aspire, which she created to ensure that her patients from disadvantaged backgrounds would reach their full potential in terms of listening skills and the spoken language.
Her ultimate goal, and that of her team, is to help all children achieve their full potential and to close the ever-widening achievement gap.
Her book entitled Thirty Million Words: Building A Child's Brain will be out later this year.
Dr Suskind was married to fellow paediatric surgeon Donald Liu, who passed away in 2012. She has three children - Genevieve, 15, Asher, 13, and Amelie, 10.
Dr Suskind and her team monitored 17 nannies, most of whom worked for University of Chicago faculty, who were caring for children aged 10 months to three years. After a single home visit focusing on language enrichment, the nannies used 395 more words per hour on average, a 32 per cent increase. They gave children a chance to respond 14 times more per hour, a 25 per cent increase.
Dr Suskind stresses that it is important to study if the programme really works - which is why a long-term study tracking between 200 and 250 children was launched early this year. They will be followed from the age of 15 months to five years, and their vocabulary development and school readiness will be monitored.
"You really need to be able to measure differences," Dr Suskind says. "Unless you can show that you are making it better, you are just running a programme to make yourself feel better."
She attempts to explain what drives her. Her mother is a social worker and her father, once a Peace Corps physician in Senegal, spent years as a paediatrics professor specialising in child malnutrition.
Her late husband was also an inspiration. Dr Donald Liu, who was surgeon-in-chief at Comer Children's Hospital, excelled at minimally invasive surgery procedures for children. He drowned three years ago trying to save two children who were caught in a strong current in Lake Michigan. The boys made it safely to shore but not Dr Liu.
"When he saw the two children struggle, there was no hesitation on his part. The children needed help, the children got help. He would never have considered standing on the shoreline while two children struggled, even if he had known it would cost him his life.
"In our country, we have too many children who are struggling against the odds of achievement. They are floundering. We cannot stand on the shoreline."