A perplexed copy editor at another paper wrote after a visiting professional laid down the law on likely. Use it only as an adjective, not an adverb. "He will likely be proved wrong" should be rewritten as "He is likely to be proved wrong." Next question.

Well, not so fast. Henry Fowler wrote that likely as an adverb must be coupled with very, most or more in educated speech and writing, though it can be found used otherwise in Scotland, Ireland and the United States. (Take that, outlanders.) Theodore Bernstein wielded the same cudgel. But Merriam-Webster’s concludes, in a lengthy entry, that "the use of likely as an adverb without a qualifier … is well established in standard general use in North America. It is an old use, dating back to the 14th century. The strictures on it seem to have developed because it dropped out of mainstream literary use in England during the 19th century."

Yet those strictures survive among the shibboleths to which newspapers’ in-house style guides are prone. And copy editors throughout the land are expected to enforce them. "He likely will" to "he is likely to," lawyer to attorney (or the reverse), "half a mile" to "a half-mile."

Some little things count for a lot, and some not a whit. Authorities disagree, and that ukase from three managing editors back may not be the best advice on what to do today.

True, the copy editor’s lot is to be responsible for the little things, but part of the responsibility is to determine which little things matter. You may recall my kvetching in a previous post about irritating pleonasms — safe haven, close scrutiny, armed gunman. And this morning’s Sun carries a bold headline, O’S REFUTE CLAIMS, though the text makes it clear that the Orioles rebut, dispute, disparage and contradict accusations that they used performance-enhancing drugs. To refute is to demonstrate conclusively that a statement is false, not merely to take exception to it.

Editing involves making a multitude of small decisions. The trick is to make the right ones.