Some mistakes are unambiguous and simply need to be corrected. Some things are less clear. If I’m checking an academic piece, I’m not going to check the names of all the authors you quote, let alone the date and title of their work. Although I suppose I could quote for doing that if you’d really like me to. So where do I draw the line?
1. Simple mistake
The capture rate near Rupert House, one if HBC’s trading posts, had been approximately 2,000 animals annually
I hope I would correct this 100 times out of 100.
2 Unexplained initials
The UNISDR (2005) definition discusses the capacity to adapt
Standard practise is to write the name in full the first time you use it, followed by the initials in brackets; after that you can just use the initials. You might think everyone knows what the initials stand for, but once you make the assumption, you introduce inconsistency. Remember your reader doesn’t want to have to work things out, so stick to the rule rather than wondering what you can get away with.
3 Inconsistency
… hymn based on the 3rd letter of John … highly relevant to our 21st century churches
If you’re using superscript, you should probably use it all the time. You could argue that it is more appropriate in some cases than others; or that ‘third’ is better than ‘3rd’. You wouldn’t be wrong, as long as you were consistent about that.
4 Referring to something that isn’t there
Ordnance Survey (1990) Coventry City Centre Sheet 55. 1:500000, Warwickshire Series
As the OS has a very popular 1:50,000 series, I thought a 1:500,000 map of a city centre would be surprising. And then I thought, it would also be completely useless. So I’m fairly confident there is no such map.
5 Bad science
When it [the cloud] is gone, Venus is already covering half the sun
A transit of Venus is when Venus appears as a tiny dot in front of the face of the sun. The observers saw it get halfway across, which isn’t the same. This was either a mistranslation, or a rephrasing by someone who hasn’t thought what a transit looks like.
This isn’t quite a sliding scale of seriousness. Only one of these examples is a straightforward error. Two or three of them ought to be picked up every time by a proofreader. The last two rely on the proofreader knowing enough about the subject to know that what you have said can’t be right.
Accuracy Works 