"Books are different," Tim Waterstone, the most refreshing wind to blow through British bookselling since Sir Basil Blackwell, wrote in the Guardian (23 January). "Fifty-five thousand new titles a year are published, and a further 400,000 are available in stock. It is the pluralism of book publishing that gives it its joy and strength, and this pluralism is protected by the fact that the publisher and bookseller are able to provide a market place for minority titles at stable prices and stable profit margins."

Tim Waterstone is addressing a subject that bores most readers. The media have been more agitated; what are the pros and cons of retail price maintenance on books (the Net Book Agreement - NBA)? "Surely the effect of abolition will be minimal," one writer argued: "only bestsellers will get discounted, and so what?" Yet the issue affects all those who buy, write or borrow books.

Abolition would hit particularly - it is said - the endangered species of small, independent booksellers. In a price discount war they would be the first victims of competition from large chains. Booksellers have to buy in volume to command the trade discount to cut stock prices. "Good for large booksellers, bad for booksellers; good for big writers, terrible for writers," D.J. Enright said.

Sir Simon Hornby, Chairman of W.H. Smith, champions NBA. Arguing with Pentos - the spearhead of the attack on "this obsolete form of protectionism" - Sir Simon spoke for small booksellers. Pentos claims that 'efficient', 'successful' booksellers won't suffer. Sir Simon believes that what is at risk is the book trade at large.

Success and efficiency can be variously measured: even in an age obsessed with margins and stock turn, they can relate to specialism, or to benefiting specific communities. Some booksellers may try to give excellent service, costly to the bottom line for single-copy and special orders. Some make substantial cultural contributions on tiny margins. Bookselling might still be a service as well as a business. The margins that spell the difference between survival and extinction are sustained by NBA.

Abolition of NBA would affect specialist publishing - poetry and literary lists for example - and writers. Tim Waterstone again: "The abolition of retail price maintenance for books has destroyed the whole texture of the US book retailing scene." He should know: he once worked in the American trade. "Discount and remainder stores are in every mall and in most main streets. Sales are brisk ...but choice is extremely narrow, and anything from the literary backlist is unobtainable."

The 'balanced' media presentation of the case addresses only its commercial implications, just as the debate about VAT on books considered primarily market consequences. An Independent leading article accepted abolition and the depletion of the independent bookshop sector as inevitable. It wanly conceded that 'inefficient' shops might die.

Yet what is implied is a fundamental change in fiscal and perhaps popular attitudes to books. It is salutary for bookbuyers to reflect on the criteria for success of many booksellers: it's less a matter of comprehensive stock than of stock-turn per shelf foot and ratios of turnover per employee. They differ only in degree from those publishers who remainder books after twelve months because 'demand has ceased', who publish only a front list and move from one Major Promotion to the next.

When books come to be regarded by producers and suppliers as commodities with an unwritten 'sell-by' date, it is not surprising that the market privileges books enjoy seem anachronistic. In the trade today, with only a few of the chains performing as booksellers in the traditional sense of the word, offering stock range and customer service, books seem to have forfeited their protected status.

The NBA should stay, even if as an anachronism. Not quite all bookselling and publishing have accepted the prevailing prejudices. Those reduced independent-minded sectors which survive need protection, as traditionally they have done. If fast-food publishers profit because books retain privilege, it is a small price to pay for a varied trade with some pockets of informed specialism and service. The vulnerable infrastructure of literary publishing and literary dissemination is never sound. It must not be deliberately weakened.

Books should remain privileged merchandise. Every book, from a romantic novel to a physics textbook, addresses the individual reader. The phrase 'educational use' hovers just beyond the frame of this argument and should be incorporated. Distinctions have been drawn between Dickens and the latest Mills & Boon romance: one is educational, the other 'entertainment'. It is possible to read David Copperfield for entertainment, a romantic fantasy in the context of a course in semiotics or post-structuralism. No hard and fast distinction can be made: education is an end of any act of reading, however primitive or degraded it may seem to severer critics. Some books are taken seriously in the press and the public arena; every book is taken seriously by the reader. It is also taken privately.

The case for the 'difference in kind' of books is based not on a book's content but on the nature of the act of reading. Is it not possible to agree that the book is a privileged product in terms of the liberty it confers on author and reader within a civilized society, and that it should remain privileged even if producers and disseminators grow cynical about privilege, even at a period when 'market forces' have become an ideological imperative as reductive to culture as the crudest applications of the dialectic were in Havana in the 1970s?