“He’s a lot slower than his brother was.”

“She still isn’t putting words together.”

“I think he understands, but he can’t say any words yet.”

These are familiar concerns to any health or education professional working with families and young children; most of the time the concerns are unfounded or will soon prove to be so. The advice that many children are slow to speak, or make slow progress to begin with but soon catch up, is right nine times out of 10 - or even more often than that.

The dilemma is posed by the few children whose slow talking indicates more severe problems requiring specialist intervention. For speech - expressive language - is a sensitive indicator that a number of systems, not just language, are intact and working well. Because speech is the supreme fine motor skill (more so than walking or hand control) it is particularly sensitive to neurological and sensory impairment, and may be the first clue that something is wrong.

Relatively few children are profoundly deaf, but a much larger proportion do not have adequate hearing for the full range of speech sounds (usually high frequency loss) with potentially disruptive consequences.

But the most common underlying problems manifested by delayed speech are severe and profound learning difficulties. Early identification here can be important because parents have to learn to support their child’s learning in ways that are not “normal”, especially the structured teaching of skills that most children just pick up incidentally.

The experience of parents of children with Down’s Syndrome exemplifies the importance of early identification, and the success of early and systematic training.

Of course there are other, rarer problems: specific language disorder of one kind or another where thinking or understanding use of language isn’t developing in children otherwise “normal”; and the whole spectrum of autistic and autistic-type problems which run much deeper than language alone. All of these require detailed specialist assessment. However, the process starts with accurate identification of children whose spoken language development is significantly below normal. The generally accepted level of “severe delay” is the bottom 2.5 per cent of any age group. But how does one identify this small minority?

Most children start imitating words before the age of 12 months (often exclamations like “Stop it!”). This is not “true” talking, but it shows they are alert to speech and it is a form of communication (babies enjoy “imitative” games). Around the end of their first year they show increasing understanding of what people are saying; and understanding is ahead of speech up to about 24 months. Around 11-13 months the first true words appear - “true” in the sense that the child means something by them - like the little girl who called trains “whoosh”.

Children take several months to produce their first 50 words - usually at about 18-20 months - and then about as many weeks to produce the next 50. By 24 months two-word (plus) sentences are the norm. But many children don’t fit this norm. They may be slow to produce their first words, but speak fluently in near-adult sentences a few months later.

How can one cope with all this variation in a screening procedure that is quick and simple to use but identifies children accurately?

James Boyle, Nicola Smith and myself make up the research team at Strathclyde University developing and standardising such tests. The three-year project has been funded by Scottish Office Home and Health Department. The tests are intended for use by a wide range of professionals - GPs, health visitors, nursery teachers, paediatricians, speech therapists, psychologists - and will be published in 1996 by Hodder Stoughton.

The First Words test is in two parts: a book of 28 picture-words most likely to appear in children’s first 50-100 words; and a parent checklist of 43 words which are (not just nouns) also high-frequency first words.

Children are asked to name the pictures (colour photographs, not drawings); and parents are asked to tick the words their children can say. Children who fall below the level of the bottom 5 per cent of their age group (the fifth percentile) on both the vocabulary book and the checklist should be referred on for further assessment. Those who are below this level on either the test book or the checklist should be reassessed after two months.

One important feature of these tests is that there are plenty of “easy” items, ones below the average level. Most tests of all kinds are made up of “average level” items. The trouble with this is that such tests don’t discriminate very well between children in the bottom 15 per cent of the population - and it is in this range that critical decisions are made.

Children who score below the fifth percentile on the expressive First Words test can be given the First Words Comprehension test, which concentrates on receptive language. The child has to identify one out of three pictures on each page. Any child who has poor expressive language but good comprehension has a much better prognosis.

Additionally, the First Sentences test uses pictures and a parent checklist to try to elicit the longest sentences the child can produce. Children who score below the fifth percentile on this test would be referred or followed up.

Parental anxiety runs high and so a give-away booklet for parents (“How to help your child”) is included in the scheme. Advice on activities and record-keeping offer parents the chance to contribute to their child’s progress.

These tests are currently being standardised on a sample of children aged 18-36 months. Our hope is that they will replace guesswork when parents ask, “What do you think?” Dr Bill Gillham is senior lecturer in the department of psychology at the University of Strathclyde.