What is The Truth? I've had a hard time with this one lately-it's tough to come by, and you'll almost never find it in the news-except maybe for reporting of accidents or yesterday's weather. I'd like to think that scientists come close to The Truth by asking highly specific questions, creating well designed experiments to answer them, and then presenting their results as "Proof".

The peer review process for scientific publication has long been criticized as a subjective and inadequate measure of scientific work. There has been a recent rise in retracted scientific works due to inaccuracies, data fabrication, and failures to reproduce published data independantly. (to see who's been naughty-check the case summaries here)

In addition, there are many scientists who believe their works are unfairly rejected due to conflicts of interest among reviewers, bias against less established scientists, bias against "less sexy stories", or results that conflict with what is considered "conventional wisdom".

Historically:
A paper is submitted to a journal by the scientists who performed the experiments and analyzed the data.
The journal editors do a primary internal review to determine if the submitted work is appropriate for the journal.
The editors then select outside reviewers to read the manuscript and comment on the scientific merit of the work and the validity of the conclusions drawn based on the data presented and recommend to Accept, Revise, or Reject. These reviewers remain anonymous.
These comments are forwarded to the authors who respond with a rebuttal or additional data.
The process may repeat itself with a revised manuscript, or the journal editors may make a final decision to accept or reject.

The journal Nature is opening the process to the scientific public with online publication of manuscripts under review, for open comment and scientific discussion. A bold step forward indeed.