I think that any paper should not be accepted without some scientifical verification, because in the other case a lot of good works will sunk among the millions of really bad ones.

On the other hand the modern scientific publishing system is seems preety strange to me. One journals publish any payed paper and get relatively high impact factor as a result of open access, other - sell papers for strange amounts of money. A lot of top-ranked USA journals rejected papers from not top world counties or university without any considerations (if the paper has no USA authors or USA grants mentioned in it).

Referees (not all of them, surely,but a very large part) want only to see citations of their own papers in the paper reviewed and oftenly reviewing papers only by abstract and bibliography.

I think, that all scientific papers should be free of any charges and open accesible.

Journals do not spend a lot of money today on the electronic publicationsm and referees are working without salary too. Selling of paper journals to libraries will give them some quantity of money and the missing amount should be payed by the top world/country universities of goverments which all are interested in scientific development.

And all referees of the paper should be mentioned in it with or without their decisions on it. That will make them to be much more thorough in their work and not too interested in getting a lot of additional citations.

You got the right point here. I myself experienced the above scenario. My article is rejected in an open access low impact journal, but ironically has been accepted by another journal (high impact factor).

I am also agree with the point mentioned that let the scientific community read our article. Their decision should be accepted.

I think that any paper should not be accepted without some scientifical verification, because in the other case a lot of good works will sunk among the millions of really bad ones.

On the other hand the modern scientific publishing system is seems preety strange to me. One journals publish any payed paper and get relatively high impact factor as a result of open access, other - sell papers for strange amounts of money. A lot of top-ranked USA journals rejected papers from not top world counties or university without any considerations (if the paper has no USA authors or USA grants mentioned in it).

Referees (not all of them, surely,but a very large part) want only to see citations of their own papers in the paper reviewed and oftenly reviewing papers only by abstract and bibliography.

I think, that all scientific papers should be free of any charges and open accesible.

Journals do not spend a lot of money today on the electronic publicationsm and referees are working without salary too. Selling of paper journals to libraries will give them some quantity of money and the missing amount should be payed by the top world/country universities of goverments which all are interested in scientific development.

And all referees of the paper should be mentioned in it with or without their decisions on it. That will make them to be much more thorough in their work and not too interested in getting a lot of additional citations.

@Vladimir, its understood that paper has to have a reasonable scientific vigor before its accepted to be published but let the scientific community as a whole decide the effectiveness of proposed work instead of a couple of reviewers. Even the language barriers can be overcome if some service can translate the original submission into the universal English language.

Also, I dont agree with your comment that good work will sunk among the bad works as high number of articles published even in high impact factor journals are never cited and still researchers have plethora of junk to go thorugh before finding out related work. Rather, researchers will be more cautious by knowing that their published work will be perished if not cited later.

@Imran: I think community reviews have been tried before (PLoS?) and just didn't work. It is already hard enough to find three reviewers for the torrent of papers that are submitted every day, with a "voluntarily" review process 99% of the papers will see no readers at all. Reputable journals and conferences at least try to identify the "better" papers to save the readers some work, letting search from them, everyone on its own, simply doesn't scale.

Nowadays everyone is free to publish with world-wide availability (arXiv, personal websites, etc.), and the scientific community as a whole has judged to ignore such publications anyway. I don't think it is malice from the reviewers, it's just that the frequency of "rare gems" that didn't make it unjustified is quite low.

@Matthias, right but I am not saying community review and then accept, rather its reverse i.e accept and then cite by community. If the article is not acceptable then wont be cited and thus will be removed from the annuls.

However, your point is 110% correct.

I think what they did at PLoS was to publish the papers after a technical check and let people vote for good papers and write comments. The idea was that there should be a low entry barrier for the assessment, and come from a few "gatekeepers" to a more democratic approach. But the interest was quite low, you have to back up any criticism very well, and substantial reviewing takes time.

I'm not satisfied with how the current publication system works either, there should be a better way to leverage internet technology. However, this would require to get the major players on board, like technical societies, leading scientists or governments. Only they can provide the incentives to participate or can check if there are no malicious acts (there are enough people who would cheat to save their job or increase their prestige).

How about this:

- there is one central system

- For every paper you want to publish in the system, you have to do N reviews for other papers first (let's say 5)

- Your paper gets 3 or more reviews, but is published anyway (if the authors want)

- Reviews are public but anonymous, but readers can vote or provide new reviews if they don't like them

- Researchers can become supporters of papers they like, and the more supporters with high reputation you have, the better the ranking of your paper

- accepting reviewers will become invisible supporters, they could also cluster to TPCs where all support the decision of a few reviewers

- If you write good reviews and papers, your personal assessment will rise

This brings more people to reviewing, and actually rewards good reviews. However, this will only work if it is the major system, many good publications and researchers should be there, and attacks like voting up your own papers or having lots of fake "friends" are infeasible.