Michael Dirda has an essay over at book forum about the negative impact of bestseller lists. He suggests limiting the number of times an author can appear on the list: once. This, he argues, would allow other authors, whom can't compete with literary franchises like Tom Clancy and John Grisham, to still get exposure and introduce the general reading populace to a wider range of books.
It's a good suggestion, but one that isn't likely to happen. Our entire culture of reviews and marketing is based on numbers. Still, I'd like to see things go even further: remove the grading scales from reviews.
In book culture, this isn't too much of a problem. The NY Times has assigns no grades to their reviews. One of my favorite online magazines for poetry, Coldfront, however, does. We see this more in film and music. I can't think of a single venue that doesn't rate an album or movie with stars, numbers, or letters. Aside from advertising, what good does this do?
Stars, numbers, and letters don't communicate any of the things a well written review do, but how often does someone see a movie because of how many thumbs Ebert pointed at the sky or buy and album because it got a 9.0 on Pitchfork? These numbers reflect, more than anything, the editorial bias these venues operate under.
There's also the problem of expectations. If a trusted critic praises something, it sets your expectations somewhere. This is true regardless of the presence of numbers, but at least, were they to disappear, the expectations would be based on the content of a review and not the number assigned to it.
Of course, like Dirda's suggestion, this will never happen. The larger institutions who produce reviews are integral to marketing campaigns, but how many times have critics been wrong, how often have you found your taste at odds with what the prevailing attitudes of the mainstream are?