Richard Baggaley (Letters, December 22/29) suggests scientists should be encouraged to write books. A better goal would be to induce scholars in sociology and the humanities to write fewer.
Articles in good journals are easily accessible electronically; books may be dear to buy, difficult to borrow and deadly to read. But books have been fetishised by promotion panels, despite being subject to softer peer review, so academics feel compelled to crank them out.
Baggaley finds some journal articles narrow, cautious and inconsequential: too many books are all those things at ten times the length.