From: Guy Chapman aka JzG
Look up a fact? No problem. Join the dots from a series of facts you looked up? Original research, in my book.
I was recently working on an article (not in Wikipedia) on the phrase "Slippery slope," which these days usually means "a course leading inexorably to disaster." I wanted to support a statement that although there _are_ old uses of the phrase in its modern meaning, it became much more popular starting around 1980.
My public library (and _many_ others) provide online access to a searchable full page-image database of complete back issues of the New York Times. In a few minutes, I was able to compile this:
Hits on exact phrase "slippery slope", all article types, ProQuest Historical Newspapers The New York Times (1851 - 2003) (a database available to patrons of many public libraries)
1857-99: 5 (average about 1 per decade) 1900-49: 41 (average 8.2 per decade) 1950-59: 11 1960-69: 23 1970-79: 36 1980-89: 144 1990-99: 402
Well, can I use this in an article? Under Wikipedia's present rules, I don't think so.
And yet.
It is certainly original research. But it's verifiable, sensu Wikipedia, in that it the database is very widely available so it's easy to confirm my results. In fact it's easier than verifying an obscure print publication with no online copy. And, in this particular case, my assemblage of facts is intellectually honest: it is reasonable to want to know the history of the phrase's use, and this is a reasonable way to find out. This is not a selective assembly of facts made for the purpose of suggesting a biassed conclusion.
And I was _not_ able to find a published source that said in so many words that the phrase bloomed in popularity around 1980.
Granted this _is_ original research, what exactly is the harm in it?
Other than its being the first step on a slippery slope, of course.