WHEN YOU go out for a bracing winter walk in the country you have to decide on the best kind of coat to wear. Do you choose a raincoat, which provides a waterproof barrier to the rain, or an overcoat, which keeps your clothes dry using a different principle - that of absorbing the water and then drying out later when the sun comes out?

Your choice depends upon a number of factors, such as how likely it is to rain, how long you will be out for, and what sort of fashion statement you want to make to the sheep. There is rarely one perfect answer. This raincoat vs overcoat idea is regularly discussed in construction literature as a metaphor for the external finishes of buildings, but the point is often lost on architecture students, who take it to mean that the raincoat principle is always better. So many new buildings are sheathed in metal, plastic and glass, in the mistaken belief that this will keep them dry inside. If only.

Anyone who has spent a typical day out in the temperate British climate wearing a raincoat - or its modern incarnation, the anorak - will know that the major source of wetting comes from within. A moderate bit of physical activity can leave your clothes soaking, because that impermeable barrier doesn't just keep the rain out; it also keeps the perspiration in. The same is true of buildings; all the water vapour produced by human activity must be allowed to escape to the outside, otherwise it will condense as liquid water on the inside of the "raincoat" and cause all sorts of dampness problems.

Most new buildings cope with this moisture load by the use of energy- intensive heating and ventilating systems, but changes in the use or ownership of a building can mean that these are not employed properly. The other disadvantage of a raincoat is that when you're caught in a shower, your top half may stay dry, but only at the expense of your bottom half: unless your anorak reaches right down to the ground all the rain runs straight off and soaks your trousers and shoes. Again, buildings can have a similar problem.

Architects often seem to think that a waterproof cladding is all they have to think about. They forget that the accumulated rainwater has to go somewhere when it gets to ground level. Often it is left to flood the drains and pavements, and soak the basement and footings. All that water cascading down from on high can create serious moisture problems at low level which, bizarrely, are often ascribed to rising damp.

And it's not just modern hi-tech office buildings that are victims of this misconception. Many ordinary homes, too, are mistreated by being given waterproof wrappings. Pebble-dashing, masonry paint, colourless silicone coatings and proprietary textured finishes are all sold to householders by persuading them it is a good idea to seal the outsides of their homes against the elements. This is rarely the case; buildings usually function best when they are left to breathe, and to cope with moisture by the natural process of evaporation.

The people who flog these paints and coatings are guaranteed to come out with some pseudo scientific claptrap including the terms "micro-porosity" and "vapour permeability". Salesmen who use these terms are highly unlikely to know what they're talking about - and they'll probably be wearing dodgy anoraks, too.