I heard something this morning that initially sounded plausible but the more it niggled away at me the more I decided that the speaker was just another I-think-I-know-better-than-you (but I'm wrong) pundit.


It was another of those TV discussion shows that only people (like me) who are either currently unemployed or actually mentally ill (you can make your own mind up about me) would ever watch. This particular show as they run into the advertising breaks always asks the audience a multiple choice question. This multiple choice question was along the lines of (and I'm making up the specifics, though not the intent)


How many people think the current exam system is a failure?

a) 25%

b) 37%

c) 63%



The host, when the show resumed, made a comment to the effect that he understood the difference between a number and percentage and the people who wrote the question clearly didn't. But I think he's wrong. Yes, it's true, a question beginning "how many" normally requires a numerical answer and a percentage answer normally requires a "how much" question. Normally, but not always.



Grammatically it's a well formed question but would anyone, anyone at all, expect an answer along the lines of 23,617,321?


Of course not. It would be utterly ridiculous.


On the other hand the alternate formulation of the question would be either,


"How much of the population thinks," or

"What percentage of the population thinks"


both of which sound rather more clunky and inarticulate than the actual question.


So, the question is, can a "how many" question be answered with a percentage and I think that, though it may depend on the question it certainly can be.


So, all that remains to ask is how many of my fellow (slightly more informed) language pundits agree with me.

I think "How many?" requires an exact (or reasonably exact) number. For percentage you need to know the set of people and number who believe the question as stated. Unless you are certain the question refers to everyone in the world, no matter what their circumstance, there is no way of knowing if the percentage given is accurate universally (although statisticians may argue the point).

I'm with Proofreader on this one.

quote:

I'm with Proofreader on this one.

What the hell happened?

quote:

Originally posted by Geoff:
I'm with Proofreader on this one.

So am I.

Technically, I agree with Geoff, arnie and Proof. However, nobody would have any difficulty in understanding the results, so what's the problem? (I have learned that perspective from none other than...Wordcraft!)