I've presented at and attended some scientific conferences, and the rules at those conferences are usually explicit in the calls for papers: submitted work must be presented at the conference for the first time, and it must not yet have been published. I think the rules on work that has been accepted for publication but not yet published vary, and they depend on how crucial it is to get information out to peers.
Granted, there are ways to bend these rules. Often, a scientific research project has been produced by a whole team of people and the data can be analyzed and presented in many different ways. If the team members come from different disciplines, then they may be able to present their research from different disciplinary angles at various conferences. Also, there may be different products that come out of a single study--one paper might focus on methodology, another on a new use of a technology, while another analyzes the consistency of that study's results with similar studies. But in any case, it's widely recognized as wrong to submit the exact same paper to two similar conferences, just as it would be wrong to submit it to two journals.
In philosophy, though, our work builds and our views shift in a more organic way, and there may be a good reason to present the same work to two quite different audiences. It seems fully acceptable to give a paper at a local symposium and also at a national conference. Is it acceptable to submit the same paper or abstract to more than one conference simultaneously, without intending to make any changes? To submit to more than one division of the APA? To submit to a group session and to the main program at the APA, perhaps betting that one would be rejected or could be withdrawn? To submit to an APA session some work that has been accepted by a journal but is still forthcoming? To submit to an APA work that has already been published? Where is the line?
More than once, I've attended papers delivered at the PSA which I've also seen delivered in a similar or exact same form elsewhere. In some cases, I've already read the journal article. In one case, I had read the journal article 4 years before seeing the presentation. Something like that might be acceptable if it extended or modified a prior published article, but in this case no changes had been made and the article's publication was not mentioned.
Conference program committees should review papers anonymously (that is, without knowing who the authors are, their affiliations, or their status). Is it fair for them to look up paper titles to see if a paper has been published or presented elsewhere? Or doesn't that undermine anonymity since many people now post paper drafts online?