On Tue, Jan 06, 2009 at 02:14:52PM -0500, WJhonson@aol.com wrote:
When a secondary source brings forth a statement, it can be balanced by a primary source. What would be wrong would be to present a brand-new claim directly from a primary source, which no secondary source mention whatsoever.
What would you make of decades-old papers that are well known and accepted by everyone in the area, but not covered by review texts because nobody feels a need to do so? This is the situation with much mathematical research. It's simply impossible to include every fact about a topic in a text, so the author chooses a certain perspective and set of topics. Results that don't fit are left out.
More briefly: the assumption that all journal papers include contingent results or experimental data that might be invalidated later is not correct.
"The Neutrino has no mass. In other news, it's been recently found that the neutrino is made of Spam."
The only difficulty here is that the "made of Spam" claim should be attributed to the authors:
"The Neutrino has no mass. Jones and Jones (2009) have recently published a paper in which they claim the Neutrino is made of Spam."
That is assuming that the Jones/Jones result is of interest to people in the field and not just a crank paper of some sort. For example, if they published their paper in Science, that would be a sign it is of interest.
As another example, if some new researcher claimed to have verified [[cold fusion]] and published in a respected peer reviewed journal, we could certainly include that in the article even if no other source had commented on it - but with appropriate attribution.
This is assuming that all journal papers are actually primary sources in the NOR sense, of course. My own position is that NOR makes non-experimental papers "secondary sources", and in those cases there is no issue.
- Carl