Curiously, no one has provided a link to the column where Cecil answers nemo: Is “different than” bad grammar?
That out of the way, I must say that I’m disappointed in that column. If you ask Cecil a quetion involving physics, chemistry or biology, you’ll get an answer corresponding to the best knowledge we have of that particular science. What we got in this column was the equivalent of a medieval scholar telling Gallileo that he’s wrong because his physics does not correspond to Aristotle. And then he adds some ham-handed logic in a foolish attempt to support his argument.
When it comes to language, we have a science known as linguistics. Like other sciences, linguistics is based on an objective reality rather than the prejudices of the “authorities”. In other sciences, objective reality is what determines what is right and wrong. And so it is with linguistics. Linguists observe that people use language and other people understand what they say. Anything that corresponds to this is right, other things are wrong.
For example, one of the observations of this science is that some people use the construction different than, while others will use different from or different to and that in each case, other people understand them. Not only that, they understand that they mean exactly the same thing. Therefore, all these constructions are right.
People do not say *different with or *different glop or **different <anything else>*. If they did other people would not understand them. That makes those constructions wrong.
Now one thing that makes linguistics different from/than/to other sciences is that its objective reality is not fixed for all time. An electron will always act the same, whether last year or 100 years in the future. But language changes. At the current time, *different with is wrong, but for all we know, in a couple hundred years, people may start to use that construction and so it would become right.