Acceptable Level of Error

In the 1980s Northern Irish politicians of a Unionist flavour complained that the British government operated a policy of an acceptable level of terror, so long as terrorist violence was restricted to Northern Ireland and did not cross the Irish Sea into the mainland (their term for Britain). There was no evidence that any such policy existed, rather the politicians were making capital by creating a conspiracy theory out of practical reality. That reality was that you can never stop all of the terrorists, all of the time. It was a realisation that would eventually lead to the present peace in Northern Ireland. Something similar is going on in the world of publishing, when the defenders of traditional publishing accuse indie publishers of pursuing a policy of an acceptable level of error. Again, there is no evidence that such a policy exists and no matter how error-strewn a text is, it does not provide that evidence. For there to be an acceptable level of error, the publisher has to see an error and decide not to fix it. In the realm of digital publishing that takes a few seconds to fix and very few, if any, authors, editors or publishers would not fix the error prior to publication. A different set of considerations come into play after digital publication or a paper print run. Errors discovered at that stage need to be balanced with the ease with which a corrected copy can be delivered to the reader or store. Prior to publication, it is not a case of an acceptable level of error, but the reality that you cannot catch all of the errors, all of the time, in a novel-length text.

There is, however, an acceptable level of error for the reader. Very few texts are error free at first publication, whether in traditional or indie publishing. Yet readers have been enjoying reading such books for centuries. On the reader side of the fence the error is spotted but cannot be fixed. The reader engages in a question of balances, just like a publisher discovering post-publication errors. In the reader's case, the question is whether the level of errors is sufficient to detract from the reading experience. That will vary from reader to reader, dependent on, for example, their own knowledge of the grammar, location or subject-matter that the author is getting wrong. Even with that knowledge, different readers will have difference levels of tolerance for error. Therefore, the point at which the level of error ruins the reading experience is a subjective matter. For me, that point also varies dependent on whether I begin to note down the errors, because that does disturb the flow of my reading. Although, I could argue that once I decide to note the errors, I have already reached the point where I am expecting more to follow, so the threshold has been passed.

What is clear, but often denied, is that the level of error can be reached in texts edited by the largest publishers and is not a problem exclusive to indie publishers. What is also clear, and cannot realistically be denied, is that the statistical chances of a reader encountering an unacceptable level of error increases with indie published work. Digital self-publishing means that it is easy for those with weak grammar or research skills to publish work so full of errors that it would spoil most readers' experience. That does not mean that the self-publisher adopted a policy of an acceptable level of error, but raises the question of their abilities to notice the multiple errors in their own text. This is the reality that the ease of self-publishing has brought and the only solution is for readers to use the returns, complaints, and review systems of the retailers. That is an advance on the world of print publishing, where you normally cannot get your money back because there is an error in the book. Although, most digital retailer returns policy would require that you return a book very soon, so will not help those of us who maintain large To Be Read digital collections. In general, neither traditional nor indie publishers have a policy of an acceptable level of error. What they do have is an acceptance that once they are ready to publish, it is with the knowledge that there will still be a few errors that readers will complain about. That, however, belongs in the realm of accepting that errors, like terrorists, cannot all be found before they do their damage. So rather than an acceptable level of error, there is acceptance that there will always be a level of error.