THE hunt for Saddam Hussein has come to Herefordshire. News crews turned to one of the world's few expert 'ear witnesses' able to positively identify his voice. She is Kington-based Dr Kate Storey-Whyte, a specialist in forensic phonetics and linguistics - applying the sciences of speech and language to policing and international security. Outside of academic and professional circles Dr Kate doesn't say much about her work, preferring to let that work speak for her. But in a rare interview she tells the Hereford Times how it is much more than a matter of semantics.

"ONLY the actions of the faithful who struggled and fought can evict the invaders. Our belief is strong that God can grant us victory."

It's him. It has to be. That hectoring, heavily inflected, hauntingly disembodied voice. Saddam Hussein.

The news crews need to be as sure as the CIA. They send the tape to one of the few people really able to tell them what they want to hear.

But Dr Kate Storey-Whyte doesn't do 'easy' listening. She can't say. Saddam stays hidden.

Maybe the next tape to come her way will be the one.

A muffled, multiplied wash of sound. Elastic waves lapping in a latex sea; the bloated, leaden wheezing of a bellows; bubbles bursting in bottomless lava.

The sound of breathing. The sound of life.

As a child in Australia Kate would listen to herself breathe, honing an ear for language fine-tuned by an 'inspirational' English teacher. Kate took up teaching herself, entered academia as a specialist in phonetics - the science of speech. She did voice identification work for detectives and drifted into what she watched on her favourite 'court and cop' TV shows.

Through friends Kate found herself in Kington and an office as unobtrusive as its occupier - more favourite aunt than aural enforcer - at the end of a row of cottages. There - under the company name Audio Lex - she clips on headphones, carefully settles them over her ears and concentrates.

That concentration makes Kate one of only 40 expert 'ear witnesses' police and security agencies around the world employ as hearing aids.

From Isaac's suspicions of Jacob in the Biblical Book of Genesis to 21st century 'art of the state' surveillance, evidence of, or by, voice has a long and varied history, though it wasn't until the widespread use of audio recording that the field of forensic phonetics came into being.

So rare is such expertise that Kate doesn't just break criminals by voice but also builds defence cases too, with the odd 'private assignment' in between.

Whenever and wherever sound affects an investigation Kate will hear of it. But to her discretion is the watchword of a good listener.

Kate's is a voice for quiet rooms. She will speak of what she has heard in court and let the court decide.

The courts want to hear what Kate has to say. Her diary dictates demand.

Weeks can run on routine, then it is Basildon for Friday, Melbourne by Monday, Hong Kong on Wednesday to make Exeter for Thursday. Southampton, Liverpool and Canberra are scheduled along the way while taking requests from Turkey, South Africa and the USA.

To do what Kate says neither she, nor any other expert, can ever define as an exact science - give a tape to 10 different people you get 10 different transcriptions in return.

Imperfections

So when in court they come back at Kate to claim she heard what she expected to hear. Kate counters that phonetics - by its nature - is always open to interpretation. Where the mind is up against instinct, the true expert accounts for imperfections in acting on the evidence in front of them. Being wrong gets them to right.

Loop playbacks, analogue filtering, binaural listening, dialectical strategies, spectrographic imaging, fundamental frequency tracking ... forensically speaking phonetics is as complex as it sounds. Best to think of it at its simplest - a 'usual suspects' line-up testing the ears instead of the eyes.

Voices are as easily recognised as disguised by emotion or manipulation. To work on speaker identification from an original tape Kate needs a sample, at least 30 seconds of clear speech from a suspect or subject that she can break down to particular vowel and consonant pronunciation patterns.

Those patterns are points of phonetic reference when the sample voice is compared to that on the original tape.

From that computer-aided comparison Kate can provide an opinion on the presence or participation of a particular speaker. She can do the same with sounds through Microsonics - a science so specific that it can pinpoint a gun being put down on a table, indicate what sort of weapon it was and hint at who might have pulled the trigger, says Kate in a teasing reference to a recent case.

Right now, the toughest challenge facing forensic phonetics is cleaning up 'dirty' tapes. The poor quality of most recordings - especially covert surveillance given the places bugs lurk - means much time is taken up with screening for extraneous sounds or testing for tampering.

The next 'big thing' will be the difficulties inherent in digital recording, says Kate. Like the pictures, the sonics are so much easier to selectively edit.

Though the world comes to Kate at the click of a mouse, she enjoys the travels on which her work takes her, keeping apace of forensics with other experts who, like her, believe they have a social responsibility to apply their skills as they do.

Kate only has to think of the victims that she has met to say why. And sometimes the suspects. Investigation, she says, is about individuals not 'entertainment'. As more and more crimes are solved behind their scenes the science should respect 'real lives'.

There's more to Kate's life than reels. She switches off from other people's conversations by riding, gardening or watching TV.

But not appearing on it. Kate catches up on cases she has worked on through those she has worked with.

She lends an ear to the news to hear of those headed her way.

How voices betrayed a drugs gang

BIG players, big stakes, big talk - but voices broke the international drugs gang that thought speaking in tongues would baffle the detectives bugging their hotel room.

Those detectives did have an informant inside the gang planning a record shipment of heroin into Australia, but no one really knew whose side he was on. And anyway, he'd gone.

So the case rested on the surveillance tape sent to Dr Kate Storey-Whyte.

The first hearing was not good, an hour-and-a-half of poor quality recording in about eight different languages.

"My immediate reaction was that it could not be done," says Kate.

The sound was so bad some interpreters could not make out anything said. Others found the work too boring to continue.

The languages were from minority groups; it was tough to find an interpreter who didn't know the suspects.

It took five months of work and many interpreters to transcribe the tape, translate it and identify the speakers.

But one section still posed problems. Amid the English, Malay, Thai, Cantonese, Hakka, Hokkien and Mandarin was an apparently unidentifiable language.

"We had streams of interpreters coming in, listening and shaking their heads," says Kate.

The transcription team was near surrender when one last minute interpreter started translating this strange tongue freely. He said it was Mandarin - only spoken badly.

The two key defendants were Malay nationals who had arranged the arrival of the heroin, its collection, storage and distribution. And had done so in such a way that another Chinese gang member would get the blame if they were 'busted'.

With the translation and transcription done, Kate had ample comparison material in the target languages. Eventually the Chinese suspect and one of the Malays decided it was best to 'assist with inquiries'.