[The word EXTRAPOLATE's] useful life began, and largely remains, in the realm of statistics (the last great wellspring of metaphorical fads before the sensational debut of personal computing). EXTRAPOLATE describes a mathematical process by which one makes predictions about unobserved phenomena by carefully noting and quantifying patterns among observed events, and then assuming that those patterns continue beyond the range of observation.
The word has made a grand tour of the social sciences (in which people never like to be caught guessing, but are perfectly willing to indulge in the practice if everyone agrees to call it EXTRAPOLATION).
A paper on social disintegration that once circulated in the foundation world offers this example of the sad fate of EXTRAPOLATE: “From the compounded anomie of Vietnam, Watergate, and Iran, it is possible to extrapolate to a gradual erosion of the social compact on which community, commerce, and democratic governance are founded.” It’s a defensible idea, no doubt.
What it is not is extrapolation. It’s a reasonable surmise based on no quantitative measures or demonstrated patterns. It is, in fact, nothing more than a (doubtless accurate) assertion that things are going to hell, and that the road to hell has lately acquired some handsome new milestones.