Last week I was sitting in a phone interview by someone who runs a study about social software and especially tagging. One purpose of the interview was to find out how I am tagging information within IBM’s social software offerings but also in external social software I am using like flickr or del.icio.us ( in this interview I learned how to pronounce this correctly and where actually this strange URL comes from: delicious, not d-e-l-dot-i-c-i-o-dot-u-s ; hasn’t been obvious to me until now since English is not my mother language).
During that interview I suddenly heard me saying that tags and folders are pretty much the same.
Hummmh, is that really true ? Let’s try to sort this out.
First of all we all might agree that both are means to somehow arrange pieces of information ( usually called documents, can be articles, photos, videos, emails, bookmarks or whatever ).
A folder is actually an old term and has an equivalent in the physical world. Remember those paper or plastic folder where you can put in some paper sheets ? Now, those folder would have one special characteristic – at least those folder in physical world: you can put one sheet of paper only into one folder. This actually requires you to think carefully about how to name your folders. Assume you have folders for your paper mail you still receive once in a while and you name your folder after companies you receive mails from, but you have also a folder for “assurance” policies. What if you receive an assurance policie from company xzy ? Do you put that into your “xyz” folder or do you put it into the “assurance policy” folder ? Or do you may be create a copy of that document to satisfy both folders ?
Tags actually also have a physical world equivalent: a little label you can attach to a document for instance using a clip or some strap. Thus in case of the assurance policy: why not adding two tags, one labeled “xyz”, the second one labeled “assurance policy”. No problem with that, right ? You could even use different colors for those different tags.
Now assume you tag all your letters you receive like this and throw them on a big stack since because you use tags you do not need folders anymore. You would very soon end up with a huge pile of letters, right ? And do you think it would be easy for you to find anything in this pile of papers ? Even if you use colored tags – once you reach a critical mass like let’s say 1000 sheets of paper it would become hard to find a particular one, right ?
Thus it looks like tags are not a really practical approach in our physical world. We prefer to use folder, most probably even with some alphabetic or otherwise organized index within that folder, since even finding something within a folder once you hold the right folder in your hands still is a little challenge.
Once we deal with “soft-papers”, or let’s say software documents or files, it becomes easier to search. The computer and some piece of software can do that for us very quickly without doing mistakes and overlooking something. Thus tags might become the preferred means to arrange documents due to their advantage that I can use multiple tags for one document.
Now in software world – is it really true that the difference between tags and folders actually is that documents have a one-to-one relationship to folders and a one-to-many relationship to tags ?
It is probably not that easy, since a folder in software world might not be always modeled 100% equivalent to a folder in physical world.
Let us look at folder in computer file systems. Under Windows it might be true that a file can exist in one folder only. Of course we can create a shortcut to a file and move that to another folder, but still the file would be in one folder and that pointer to the file in a second folder. Under Linux this might be pretty much the same but look different. Through the concept of logical links ( I am not really a file systems expert, thus those experts out there: feel free to comment. Is a shortcut under Windows equivalent to a logic link under linux ? ) it might at least look like that a file exists in multiple folder.
Let us look at Lotus Notes: usually I move a document from one folder to another one, thus it always exists in one folder unless I do not create a copy. This is the standard behavior, but I also can “Add” a document to another folder – then one and the same document will show up in multiple folders and whatever I do to one ( including deleting it ) in one folder will become visible in the other folders as well – since this is really one and the same document.
In addition to folder Lotus Notes has the concept of views – one document might appear in multiple views. If I delete it from one view it is gone in all other views as well.
Did this blog article help to figure out the difference between tags vs. folder and what they have in common ? Probably not 100%, but to some extent I hope. Let’s take those two messages out of it:
- Both tags and folder are a means to arrange information.
- Typically a document exists in one folder but can have multiple tags.
The confusion about folder and tags is real, that much is sure when I look for instance at the latest move by google to come up with a new frontend for their “Google Docs & Spreadsheets” offering. They bascially have replaced tags with folders because obviously user have been asking for this according to this google blog article: “An entirely new way to stay organized“.
Almost from the day we launched people have been clamoring for folders. They’re here! Even cooler, our new folders continue to work like the tags they’ve replaced – your old tags are automatically converted to folders and documents can live in more than one folder at a time.
Documents can live in more than one folder ? This is against what I have said above about folders, at least about folders in physical world. Either they have done a bad job to model folders in their world or they just decided to name tags “folders” because user more like this traditional concept of good old-fashioned folders.
Some readers of my company blog actually pointed out two more significant differences between folders and tags in their comments:
- With folders you can realize a hierarchy. You can build a “tree” type order by putting folders into folders. Tags usually can not be tagged.
- Folders can be emptied and you could end up with empty folders. In case of tags – if you do not have any document with that tag then this tag basically is gone.