"Thirty times weaker": Using a multiplier to describe the lack of something [duplicate]

I was watching CNN's coverage of the earthquake that struck northern California this morning, and I heard the following exchange between the CNN anchor and a seismologist, Walter Hays:

ANCHOR: This is 6.1 on the Richter scale, that means ... one-tenth the power of the one that struck in '89?

HAYS: No, that's the confusion ... It's nearly thirty times mo—weaker. It's one-thirtieth of the Loma Prieta earthquake.

(Emphasis mine.) As it happens, this touches on a longstanding peeve (and I am perfectly willing to describe it as such) that I have about the use of positive multipliers to describe the lack of something: twice as small, three times colder. In everyday use, of course, this pattern is readily understood as indicating the reciprocal of the opposite quality: something that's thirty times weaker is actually one-thirtieth as strong. It was interesting to hear Mr. Hays catch himself making the mistake, and then correct himself using more formally precise language—something I think a scientist would be more likely to do than a layperson.

So my question is: Is this really a mistake, or is it actually semantically proper/meaningful to use language indicating a positive magnitude to characterize a concept that is defined specifically as a lack of something else? We don't have a problem using ordinary intensifiers and superlatives in such contexts: a thing can be very small or much weaker. It seems intuitive to me that a multiplier should be treated differently from an intensifier, but I can't actually define why that should be the case. What about the reverse: if twice as small is semantically proper, what about half as small (the obvious ambiguity notwithstanding)? Or are these all more properly thought of as physics or mathematics questions, rather than language questions?

Solution 1:

"Thirty times weaker" is bad English, and atrocious mathematics. You should say "one-thirtieth as strong" - if this is what you actually mean.