Dear Vishwanath, thank you for the good question. As (co-)author and co-editor of some conference proceedings and an online journal, I also asked myself the same question, especially when I had to reformat the references of submitted contributions. However, by doing this, I developed some understanding for the differences.

As Fulvio and Svetlozar have pointed out, the phenomenon is related to those standardisation questions that are often difficult to complete. And, in my opinion, in this question there is no deep industrial or community need as in the case of metric system, for instance.

As a multidisciplinary scientist I had the opportunity to deal with papers of completely different fields, and these researcher communities have their own traditions. Moreover, in some cases there are certain specific needs. If you consider the high number of varieties of proceedings, books, etc., it is easy to recognise the standardisation problem. (For me, being aware of all these problems, it was difficult many times to apply the half-heartedly defined "rules" that a specific guide4authors contained.) It is not easy to use the same citation rules for a journal in veterinary science, education, geochemistry and cartography, for instance. Their material cited (and its composition) are really colourfully different (experiment reports, legal documents, research reports, chemical descriptions, maps, etc.).

Maybe for a "simple" journal contribution it seems to be very easy to standardise the format, but if you consider the journals published by AGU, in which there are no page numbers in the sense as in other journals, it is easy to recognise the difficulty. And AGU (in my opinion) was right to introduce the paper numbering/identifications system (because of exorbitant large page numbers of e.g., Journal of Geophysical Research like 41,112-41,123 that does not seem to be a page range at all). Even if it is not easy to implement in your (or my) favourite reference manager software, I can understand very well their reasons.

If we are already at it, I think the solution (well, yes, a 90%-solution) is to use a reference manager system that is able to reformat (most of) your references in the requested format. This is the user's side.

What concerns the publishers' side, I think they could help their authors as well by providing (a) well-defined (REALLY well-defined!) formatting instructions and (b) proper (I mean PROPER) templates for their formatting (on one hand in LaTEX and on the other hand reference manager system). A good news for us that at least the big publishers use the same format for relatively long time, so the third-party solutions that are being developed may have a good "best before" date. Interesting, but understandable that the journals applying the open-access model (in other words the author pays everything) typically do a better job in this field.

Despite all this I cannot exclude that a community-driven effort may develop (like in the case of JPEG and PNG image formats or the TCP/IP) a generally applicable, well-designed bunch of formatting rules. But before initiating such a "movement", some enthusiastic experts should prepare an RFC-like white paper that can be commented. And to introduce that afterwards, you will still need the "critical mass" of supporters at the key positions.

Well, a good question, which would need a long formal answer, if starting speaking on the different bibliographic styles, reference formats and the historical, geographical and other types of origin of their rules. However, in order to "feel" the nature of the problem just think on why for example there are so many currencies throughout the world (wouldn't be easier to have one only?) or why there are so many typing font styles, e.g. Arial, Times New Roman, Calibri, etc.?). You may think on many other similar problems....

Dear Vishwanath, thank you for the good question. As (co-)author and co-editor of some conference proceedings and an online journal, I also asked myself the same question, especially when I had to reformat the references of submitted contributions. However, by doing this, I developed some understanding for the differences.

As Fulvio and Svetlozar have pointed out, the phenomenon is related to those standardisation questions that are often difficult to complete. And, in my opinion, in this question there is no deep industrial or community need as in the case of metric system, for instance.

As a multidisciplinary scientist I had the opportunity to deal with papers of completely different fields, and these researcher communities have their own traditions. Moreover, in some cases there are certain specific needs. If you consider the high number of varieties of proceedings, books, etc., it is easy to recognise the standardisation problem. (For me, being aware of all these problems, it was difficult many times to apply the half-heartedly defined "rules" that a specific guide4authors contained.) It is not easy to use the same citation rules for a journal in veterinary science, education, geochemistry and cartography, for instance. Their material cited (and its composition) are really colourfully different (experiment reports, legal documents, research reports, chemical descriptions, maps, etc.).

Maybe for a "simple" journal contribution it seems to be very easy to standardise the format, but if you consider the journals published by AGU, in which there are no page numbers in the sense as in other journals, it is easy to recognise the difficulty. And AGU (in my opinion) was right to introduce the paper numbering/identifications system (because of exorbitant large page numbers of e.g., Journal of Geophysical Research like 41,112-41,123 that does not seem to be a page range at all). Even if it is not easy to implement in your (or my) favourite reference manager software, I can understand very well their reasons.

If we are already at it, I think the solution (well, yes, a 90%-solution) is to use a reference manager system that is able to reformat (most of) your references in the requested format. This is the user's side.

What concerns the publishers' side, I think they could help their authors as well by providing (a) well-defined (REALLY well-defined!) formatting instructions and (b) proper (I mean PROPER) templates for their formatting (on one hand in LaTEX and on the other hand reference manager system). A good news for us that at least the big publishers use the same format for relatively long time, so the third-party solutions that are being developed may have a good "best before" date. Interesting, but understandable that the journals applying the open-access model (in other words the author pays everything) typically do a better job in this field.

Despite all this I cannot exclude that a community-driven effort may develop (like in the case of JPEG and PNG image formats or the TCP/IP) a generally applicable, well-designed bunch of formatting rules. But before initiating such a "movement", some enthusiastic experts should prepare an RFC-like white paper that can be commented. And to introduce that afterwards, you will still need the "critical mass" of supporters at the key positions.

Thank you for your opinions/comments.....

There is a lot of tradition in the reference styles, and different "schools" on how to format. I don't think this will change, because as long as you can find the cited works the references are functional. (Of course you should not mix styles in one issue/proceedings/...)

The problems start to occur when the references are wrong or incomplete, which may also be caused by the style. For example, if authors first names are abbreviated then it is more likely that author names collide and you cannot tell if it is the same person or not. This makes it hard for example for services such as Google Scholar to track citations.

To solve this I think it is a good idea to split the reference in a "human-readable" part that can follow the classic or even more abbreviated styles, and a unique identified like the DOI (http://www.doi.org/). I saw this while browsing web publications of Science (or Nature):

First Author et al., Title, DOI

Each reference was only two lines, but everything was more accurate than it is currently. Proper references take a lot of space, so this is a plus. An additional benefit is that reference formatters like BibTeX could auto-generate the human-readable part, because seemingly some authors are not capable of getting them right on their own.