I now feel that a number of unresolved issues related to the Wiki ultimately derive from the fact that tags and encyclopedia articles should not both be created in accordance with the same criterion. Specifically, it seems to me that a topic that is suitable for a tag is sometimes too specific to be a suitable topic for an article.
I wonder if this problem could be solved, or at least reduced, by allowing article section headings to also serve as tags. I think this would probably be most helpful for articles that cover particular disciplines, such as psychology or computer science. Here it seems that it makes most sense to have a single article covering each discipline, yet multiple tags discussing different aspects of the discipline, such as research on that discipline, careers in that discipline, or applications of that discipline. Currently we take a hybrid approach, sometimes having entries for the discipline as a whole and sometimes for specific aspects of it.
Another advantage of allowing article sections to be used as tags is that some tags are currently associated with a very large number of posts. This suggests that a more fine-grained taxonomy of tags would organize the contents of Forum better, and allow users to find the material they want more easily.
A complication is that not all section headings will be suitable for tags. This issue could be solved in various ways. For example, the search field that opens when the user clicks on ‘Add tag’ could by default only show the tags corresponding to article titles, just as it does currently. However, the user could be given the choice of expanding the tag to display the corresponding headings, and allow them to select among any of these. Perhaps headings already selected as tags by previous users could be shown by default in future searches.
I’m not particularly confident that this is a good idea. But it does seem like something at least worth discussing further.
These are reasonable concerns, but adding hundreds of additional tags and applying them across relevant posts seems like it will take a lot of time.
As a way to save time and reduce the need for new tags, how many of your use cases do you think would be covered if multi-tag filtering was supported? That is, someone could search for posts with both the “psychology” and “career choice” tags and see posts about careers in psychology. This lets people create their own “fine-grained taxonomy” without so many tags needing to have a bunch of sub-tags.
I think something along these lines feels promising, but I feel a bit unsure precisely what you have in mind. In particular, how will users find all posts tagged with an article section heading tag? Would there still be a page for (say) social psychology like there is for psychology, and then it’s just clear somehow that this page is a subsidiary tag of a larger tag?
Inspired by that question, I think maybe a more promising variant (or maybe it’s what you already had in mind) is for some article section headings to be hyperlinked to a page whose title is the other page’s section heading and whose contents is that section from the other page, below which is shown all the tags with that section heading tag. Then if a user edits the section or the “section’s own page”, the edit automatically occurs in the other place as well.
And from “the section’s own page” there’s something at the top that makes it clear that this entry is a subsidiary entry of a larger entry and people can click through to get back to the larger one. Maybe the “something at the top” would look vaguely like the headers of posts that are in sequences? Maybe then you could even, like with sequences, click an arrow to the right or left to go to the page corresponding to the previous or following section of the overarching entry?
Stepping back, this seems like just one example of a way we could move towards more explicitly having a nested hierarchy of entries where the different layers are in some ways linked together. I imagine there are other ways to do that too, though I haven’t brainstormed any yet.
I now feel that a number of unresolved issues related to the Wiki ultimately derive from the fact that tags and encyclopedia articles should not both be created in accordance with the same criterion. Specifically, it seems to me that a topic that is suitable for a tag is sometimes too specific to be a suitable topic for an article.
I wonder if this problem could be solved, or at least reduced, by allowing article section headings to also serve as tags. I think this would probably be most helpful for articles that cover particular disciplines, such as psychology or computer science. Here it seems that it makes most sense to have a single article covering each discipline, yet multiple tags discussing different aspects of the discipline, such as research on that discipline, careers in that discipline, or applications of that discipline. Currently we take a hybrid approach, sometimes having entries for the discipline as a whole and sometimes for specific aspects of it.
Another advantage of allowing article sections to be used as tags is that some tags are currently associated with a very large number of posts. This suggests that a more fine-grained taxonomy of tags would organize the contents of Forum better, and allow users to find the material they want more easily.
A complication is that not all section headings will be suitable for tags. This issue could be solved in various ways. For example, the search field that opens when the user clicks on ‘Add tag’ could by default only show the tags corresponding to article titles, just as it does currently. However, the user could be given the choice of expanding the tag to display the corresponding headings, and allow them to select among any of these. Perhaps headings already selected as tags by previous users could be shown by default in future searches.
I’m not particularly confident that this is a good idea. But it does seem like something at least worth discussing further.
These are reasonable concerns, but adding hundreds of additional tags and applying them across relevant posts seems like it will take a lot of time.
As a way to save time and reduce the need for new tags, how many of your use cases do you think would be covered if multi-tag filtering was supported? That is, someone could search for posts with both the “psychology” and “career choice” tags and see posts about careers in psychology. This lets people create their own “fine-grained taxonomy” without so many tags needing to have a bunch of sub-tags.
I think something along these lines feels promising, but I feel a bit unsure precisely what you have in mind. In particular, how will users find all posts tagged with an article section heading tag? Would there still be a page for (say) social psychology like there is for psychology, and then it’s just clear somehow that this page is a subsidiary tag of a larger tag?
Inspired by that question, I think maybe a more promising variant (or maybe it’s what you already had in mind) is for some article section headings to be hyperlinked to a page whose title is the other page’s section heading and whose contents is that section from the other page, below which is shown all the tags with that section heading tag. Then if a user edits the section or the “section’s own page”, the edit automatically occurs in the other place as well.
And from “the section’s own page” there’s something at the top that makes it clear that this entry is a subsidiary entry of a larger entry and people can click through to get back to the larger one. Maybe the “something at the top” would look vaguely like the headers of posts that are in sequences? Maybe then you could even, like with sequences, click an arrow to the right or left to go to the page corresponding to the previous or following section of the overarching entry?
Stepping back, this seems like just one example of a way we could move towards more explicitly having a nested hierarchy of entries where the different layers are in some ways linked together. I imagine there are other ways to do that too, though I haven’t brainstormed any yet.