A box on the front page of the New York Times on 30 May:
To Our Readers
On June 1, the Monday-Saturday newsstand price of The Times will increase from $1.50 to $2.00. The Sunday newsstand price will increase from $5.00 to $6.00. Home delivery prices will go up by $.70 to $1.54 per week, depending upon the days of delivery.
(my emphasis). When I got to the boldfaced bit, I was at first startled to read that apparently a current home delivery price of $.84 per week (certainly not what I’m paying) will increase to $1.54 per week (still a bargain), but then “depending upon the days of delivery” made it clear that the reading I’d gotten for the boldfaced bit was not the one the Times intended — though both readings are possible for the boldfaced bit out of context.
What set me up for the misreading was the prior context, where the price changes were of the form increase from OLD to NEW, so I read the third price change as similarly focused on the new price, that is, as go up by CHANGE-AMOUNT to NEW, but in fact the intended reading of by … to was not this one, but one expressing a range of change amounts, that is as go up by CHANGE-AMOUNT-1 to CHANGE-AMOUNT-2.
The intended reading could be brought out by marking CHANGE-AMOUNT-1 with from:
go up by from $.70 to $1.54 per week.
I suppose there are people who find the sequence by from ugly and wordy, but googling on {“increase by from”} nets a fair number of relevant (and unambiguous) examples, for instance:
The global population is projected to increase by from two to three billion to around nine billion. (link)
(I’ve switched from go up from to increase to reduce the number of irrelevant hits.) In case it’s the other reading you want, a comma will make things clear:
go up by $.70, to $1.54 per week.
But otherwise, you have to rely on background information and real-world plausibility to choose the appropriate reading — which is just what I did.