Yes, it is about viewing (= finding) and not storing.
For those who can’t imagine why some of us think tagging is important, I venture to add some length to this thread, including some invocations of ancient software:
I may store 20 articles about the history of chemistry in a file so named, and that is useful. What do I have about “history of chemistry”? Just look at the list of titles in that file. But suppose there is something odd, off-topic, in one of those articles: by the way, famous chemist X happened to suffer from some neurological abnormality. On down the line, I want to find that dimly-remembered fact to integrate it with, well, whatever. I cannot remember the name of the individual or of the specific condition. How am I going to find this? Tags offer a possibility (not a certainty!) IF I tagged the original item AND if I have a visible list of tags.
Presently, I “tag” things just by adding words to the top of the article in DTP. But I’ve no way to see what tags I have used, short of putting each such term on a dedicated list every time I use a term, and then alphabetizing them all periodically. So, did I add “epilepsy”? “synaesthesia”? “neurological”? I cannot tell, unless there is a tagging system and it shows me a list of all my tags.
Also, if I use my workaround to add terms, say, “epilepsy” because that relates to some interest of mine, then all I can do is search for that term and look through every single item that contains that word. With a tag system, you apply the word where you think it applies and you retrieve only items you applied the tag to. (You can rely on DTP’s search to turn up an entire list of items containing the word, so you only need to tag for special purposes.) Without such a system, I have to look at every article that mentions the word once, even if only in the title of an article in the bibliography.
Finally, hierarchical tags such as the ones used in Keyword Manager for iPhoto (and in the old lamented citation organizer, Papyrus) enable refinement of searching. To keep with the epilepsy example, that term might be placed in a hierarchy: Brain > Neurology> Disorders> Epilepsy. Disorders could include various other subheads, so when you want a broader look that includes but is not limited to epilepsy, there you are. In iPhoto, I have Animals > Wildlife, and Animals> Pets > Dogs >then subcategories for each dog I’ve had fun photographing. These hierarchical categories are very useful in narrowing or widening my view of what I have.
Papyrus, a product of Research Software Design in the late 90’s, utilized a system of linked keywords which allowed the same term to appear in more than one hierarchy! Epilepsy (I am regretting this choice of examples, but too late now) could appear as: neurological disorders> category A >epilepsy, as well as social welfare >accommodations for disabled persons >epilepsy, and you could search for either one, and not retrieve the other set. I still have the manual for Papyrus, and I just looked at it to refresh my memory.
Another late lamented (by me, anyway) program is Mailkeeper, a text snippet organizer from the Nisus Writer folks. It allowed you to establish keywords in three parallel columns, attach any keywords to any item, and then do additive/subtractive searches by clicking or option-clicking on various terms. Show me everything tagged with Carnivores and Felidae and Africa and Population, but omit anything tagged with Lion or Nigeria—I don’t want to see those items. Is that a dream or what? Instant lists of qualifying titles, instant changes in search criteria: add Asia, tell it to omit Africa, add 19th century, add Africa back again…
What we are trying to do is add some of the abilities of the human mind to applications. Fuzzy searches, hits involving criteria that are somehow related to the stated search criteria, adding and subtracting sub-criteria, our brains do all this on the fly.
