Junior is fretting. He seems tired and wants to sleep. If only mum could understand what he wants. One Australian woman figures she does.

Ms Priscilla Dunstan, 33, claims to have decoded meanings behind the sounds that babies make.

She has developed the Dunstan Baby Language system, a set of sounds babies produce by reflex when asking for specific needs to be met.

They are used by babies aged zero to three months to 'communicate' the five basic needs of hunger, fatigue, the need to burp, pain from wind in the gut and general discomfort.

In an e-mail to Mind Your Body, Ms Dunstan said: 'When I had my son Tomas, I started listening to the cries and I thought I could hear differences. There had to be a reason why there were differences, so I kept a log. I also used audio equipment and recorded the cries.'

The baby cried a lot because he suffered from colic and reflux. Thus it became invaluable to know when he needed to be burped and fed, as this helped relieve his condition.

'In all the books that I'd read, I was told that the baby burps after a feed and that's it.

'But I discovered that when they said 'eh', they weren't hungry; they just needed to be burped and when you burped them they were happy again,' Ms Dunstan said.

Her background as a trained violinist and opera singer allowed Ms Dunstan to differentiate the different sounds a baby made.

The first sound she recorded off baby Tomas, who is now nine, was 'neh', and it seemed to come when he was hungry. He was about 12 weeks old then.

It turned out she was right, she said. So whenever he said 'neh', she would feed him. The result: He had feeds only when he was hungry and he did not fuss a lot thereafter.

It took eight years of research to uncover this system of sounds - a 'language' that is shared by all babies, said MsDunstan.

To test her hypothesis, she told Mind Your Body that she visited different countries.

'I found that all babies, regardless of race or culture, made the same sounds. I also had many friends who were travelling to remote areas to film observational video tapes. Again, they showed that all babies were making the same vocalisations,' she said.

Ms Dunstan asserts that her research has been verified.

Independent longitudinal studies were conducted last year in Sydney and Chicago, with over 300 mothers involved.

Conducted by an independent consumer research firm, it found that nine out of 10 mothers of young infants found the Dunstan System is beneficial. They would even recommend it to another mother.

Seven out of 10 mothers reported that their baby settled faster. Half of the babies experienced better feeding and half of the mothers experienced unbroken sleep.

Earlier this year, a third independent study conducted in Britain involving another 168 mothers of newborns had results consistent with the two previous studies.

Ms Dunstan said that thousands of mothers in Britain, Australia and the United States have experienced the system first-hand and are happy with it.

She is hoping to bring DVDs to Singapore through retailers such as Mothers Work and Mothercare.

Dr Maureen O'Brien, the director of education with Dunstan Baby Language, which is based in Australia, said: 'By learning to understand their baby's cries, parents are empowered to meet their baby's needs and thus avoid the feelings of frustration and stress that accompany prolonged and unexplained crying.'

The Dunstan system has its detractors among mainstream researchers and academics in language acquisition.

Dr Madalena Cruz-Ferreira, a senior lecturer with the Department of English Language and Literature at the National University of Singapore, said sounds uttered reflexively by babies do not constitute a language, any more than adult moans or cries of pain do.

'Unless, of course, we want to extend the definition of the word 'language' beyond any usable limits,' she said.

'We know what other people mean because of conventions in our languages which we learn through many years of nurturing, but the 'Dunstan Baby Language' is not conventional to start with.'

Dr Cruz-Ferreira, who is an expert in linguistics and child multilingualism, explained that communication begins when babies realise that making certain sounds or gestures results in specific parental responses like cuddling, feeding, changing nappies or any other behaviour.

Sometimes, babies as young as three weeks will fake a cry to get such a reaction from their mummies or daddies.

She said that the Dunstan claim to universality is also problematic.

'The 'words' of the purported Dunstan universal baby vocabulary are strangely English-spelt. Non-English speakers will certainly struggle to work out what 'eairh', for example, might translate to,' she said.

She said that peer-reviewed research since the late 1960s confirms the observation that the number of distinctive cries varies in each baby.

'Furthermore, the type, and even the duration of children's cries are unreliable. It is therefore extremely difficult to assign 'meanings' to infant cries.

'Research also shows that babies, like the rest of us, are individuals, and that healthy and impaired babies cry differently,' Dr Cruz-Ferreira said.

Ms Dunstan feels that while some linguists feel that the use of the word 'language' could be debated, the actual efficacy of the Dunstan System has not been challenged.

'Irrespective of our use of the word 'Language' as a product descriptor, our research demonstrates that infants the world over use the five sounds identified in the Dunstan System, allowing parents to identify their baby's everyday needs in their cries,' she said.