Publishing a Result: Rules of Thumb

[center o bass]

Rules of thumb for when to publish a result.

Being novel to the game of academia I wonder if there are any rules of thumb to follow
when one should publish a result or not.

Suppose one finds a result which one suspects might be new. How should one then proceed?
Should one go about looking around carefully to see of someone else has done it? If so, how carefully should one look? What would qualify as a 'new' result anyway? Is it enough if the result comes with a new twist/perspective, or if one builds on an earlier result?

 

[center o bass]

As a theoretical physicists, in which journals should one ideally wish to publish one's paper? Does it depend on the content of the paper? (If so, which journal should one wish to publish a result in gravity, quantum gravity and unified theories?) Can it count against you if you publish in one of the less recognized journals?

 

[ZapperZ]

As theoretical physicist, one should already have read dozens, if not hundreds, of papers already. One would expect that such physicist already are aware, based on the papers they read, the journals that these papers are published in and the different tiers and prestige that each journal has.

Zz.

 

[f95toli]

This depends on many factors, and is something you learn gradually if you work in a field. Basically, the rule-of-thumb is to submit the paper to a journal YOU read, you know your collegues/competitors read and that publish results in your field.

This is something best discussed with your supervisor.

Edit: ZZ was faster

 

[f95toli]

Should one go about looking around carefully to see of someone else has done it?

If it is a purely theoretical result then yes, you should look very carefully. The last thing you want is for one of the referees to tell you that the result has already been published (this is one of the things a referee looks for). Experimental results also have to be at lest somewhat novel, but experiments are by their very nature rarely exact duplicates of work in other labs (we always get somewhat different results or use other methods).

What would qualify as a 'new' result anyway? Is it enough if the result comes with a new twist/perspective, or if one builds on an earlier result?

There are no clear rules and it depends on the journal. Journals with lower "prestige" often publish work that e.g. uses a new method to reach a known result, or perhaps is just another "version" of a known results.

If you want to publish in Nature/Science it has to be entirely novel. if it is a conference proceeding for a minor conference it is much less strict and can also be work in progress.