THINGS are getting worse. It is a cosy, familiar idea, and when applied to young children's speech and language it offers ample opportunity for bemoaning the pernicious effect of television, dummies and single parenthood. There is only thing wrong with it. It's not true.
There is absolutely no evidence that the speech and language of children entering nursery and reception classes has declined in the past five years.
In fact, The TES has uncovered a small amount of evidence suggesting that pre-school speech and language is actually improving. So why do heads, teachers, politicians and newspapers all believe the opposite?
The most recent outbreak of "things are getting worse" was prompted by Alan Wells, director of the Basic Skills Agency. In a widely-reported speech earlier this month he said that whenever he visits a school, heads tell him that their three and four-year-olds are less able to speak clearly, express themselves coherently and understand instructions than children of the same age five years ago.
Mr Wells told The TES his agency's survey of 700 Welsh primary heads, due out in the spring, will show that nearly two-thirds of heads believe children speak less audibly now than in the past and 61 per cent say children know fewer rhymes and songs than in the mid 1990s.
Those figures are almost identical to those from a survey by the National Literacy Trust and the National Association of Head Teachers, reporting that 74 per cent of heads believe three-year-olds' speaking and listening have deteriorated. This has prompted a new NLT early language campaign - likely to be titled "Talk to your baby" - to be launched in a year's time.
But do heads' beliefs fit the facts? What is surprising, in the midst of widespread belief that children's language has deteriorated, is that no one until now has looked at the evidence which does exist.
In the past five years - exactly the period when heads identify a decline in children's skills - all children have been given so-called "baseline" tests at the start of their reception year.
The TES asked Britain's biggest baseline test provider, the Centre for Curriculum, Evaluation and Management at Durham University, which produces the Pips test used in thousands of primaries, to compare children's scores in the "language" elements of its test over the past five years.
There were two relevant elements: a vocabulary section, in which children are asked to name pictures, and a phonological section, in which children are asked to identify rhyming words. Between them, they should give some indication of what has been happening to both language and listening skills over time.
What the figures for 722 schools and 20,000 children show is that, since 1997, children's vocabulary knowledge has marginally improved, and their phonological skills have improved considerably.
Of course, that overall averaged figure disguises different changes in different local authorities and schools, says professor Peter Tymms, Pips director. But there is no apparent pattern to the differences: deprived intakes are as likely to show an improvement as a decline. In any case, most schools are clustered around the average no changeslight improvement state. A mere handful show the big decline that heads say they have seen.
To ensure that this is not just a Pips effect, The TES contacted Birmingham local education authority, the second-biggest baseline author with its Signposts tests. It has results for speaking and listening for 14,000 reception children in Birmingham. Over the past five years, they have improved too.
Finally, we contacted Reading University, where Dr Susan Edwards is senior lecturer in clinical linguistics. She is the author of the Reynell test, a complex screening used by speech therapists and psychologists for young children with serious delays in speech development.
Four years ago, Dr Edwards restandardised the Reynell. This involves creating a "normality" benchmark against which children with abnormal speech can be tested. She asked teachers to pick out 1,074 children aged two to seven with "normal" speech and language, and then extensively tested what they could do. Their results set the standard for normal development.
Since then, she says, speech therapists have frequently, though only informally, told her that children with speech delay find the new Reynell test too difficult. There are three possible explanations for this.
First, "abnormal" children have got worse. This is highly unlikely, given that this group of children included those with the most severe disabilities including no speech at all.
Second, "normal" children's speech has got better. This would concur with the Pips and Birmingham results.
But there is a third explanation: that teachers' idea of "normal" speech skills has changed.
That last explanation, say experts, could be a credible one. In the past 10 years, schools' awareness of speech and language - now identified as separate skills in the national curriculum and used as predictors of future literacy ability, has grown immensely.
Inclusion of children with speech and language difficulties in mainstream schools means that teachers have become both more conscious and more skilled at identifying language problems. Speech therapists, though in short supply, are now often found in mainstream schools when they once would have been confined to special units.
"Teachers have become sensitised to speech and language difficulties" says Geoff Lindsay, professor of special needs at Warwick University. "There has been a redefinition of children, rather like what has happened with autism.
There's a positive aspect to this, but it does also change our view of children."
Teachers' view may also be changed by the increasing focus on children's performance and the need to get good national test results. In the past, says child psychology lecturer Dr Marion Farmer of Northumbria University, parents and teachers would be happy to wait and see whether a four-year-old was a late developer. Now, they panic about how far that child has to travel to reach Level 2.
She says: "People are thinking 'My goodness this child can't do X, Y and Z and these are really important predictors of what they are going to achieve.' It is easy to see how such fears might have raised anxiety about standards of speech and language."
Even Alan Wells, who did not denounce TV or poor parenting in his speech, admits that the debate over language may have been coloured by the age-old tendency to think that things ain't what they used to be that "Everybody thinks summers in their childhood were more summery than they are now."
His soon-to-be published survey shows that Welsh heads believe that only 50 per cent of children start school with the speech and language they need to embark on the curriculum. It does not show that five years ago more of them did.
Young children's speech has not slipped back. This does not mean we should not set targets or allocate time and money to making it better. But we cannot justify doing so on the basis that things have got worse.
A LOT OF TALK?
* 74 per cent of headteachers think children's language skills when they enter nursery at three have deteriorated over the past five years.
* More than half said more parents should be encouraged to talk and listen to children and to teach them rhymes.
* 31 per cent said parents needed to be taught more about how children learn language.
Source: National Literacy Trust and National Association of Head Teachers'
2002 survey of three-year-olds' language.