Why can't terms be in two glossaries?
Hypertext-infused philosophy personal database software
Brought to you by:
jasonwinning
Menu
▾
▴
I have a term "control theory of luck" (nested under the term "control," which is in the "Causation" glossary) and I decided I also wanted a "Luck" glossary, which would be a nice place for the "control theory of luck" term. But it turns out I can either have two separate terms, or I can move the term from the former glossary to the latter. I can't nest the term under two glossaries. I can nest other stuff under two things: a position can be nested under two debates, for instance. Why this limitation?
Normally, in a case like that, I would have a position record for "control theory of luck" nested under Causation / Control debate records, and also nest it under a Luck debate record. I wouldn't normally have a term record for something like Control Theory of Luck. That's just my preference of course, and not necessarily the only good way to do things; I'm curious why you want to represent that structure using Term records.
I have a term for the theory because it seems like the sort of thing to have a term for, since I already have a "control" term, and if I have a "control" term then it seems like I should have a term for the control theory of luck. (I also have a Luck glossary for luck terms and this seems like a luck term.) I figure I might as well represent the structure using Term records (if I have those records) for the same reason to represent the structure anywhere else. (If you want to see me try to do it in real time and realize I can't: https://youtu.be/V-qyeO4zaL0?t=1617)
I'm not sure I agree with that inference :) I mean, there is a reason why there are record types for positions, debates, arguments, etc. rather than just having one type of record for everything. That's why I'm wondering why you don't just have a position record for "Control Theory of Luck". Is there some disadvantage to using a position record for that?
To me, it makes more sense for "Control" to be a Term record because it
etc. That is not the case for "Control Theory of Luck". The latter is the kind of thing you can imagine having arguments for and against; we definitely don't want to start having Terms be the immediate target of an Argument record. Again, what I would do is have Control and Luck be Terms that are united with Problem/Debate records with the same names, and have the Control Theory of Luck be a position nested under those Problem/Debate records.
Back in Ticket #19, when you requested the ability to have nested terms, you said "the goal is to tie the two more narrow terms to the broader term" which made perfect sense to me. You used the example of having "Epistemic Reasons" and "Practical Reasons" nested under "Reasons". But "Control Theory of Luck" isn't really a narrowing of "Control" in that way (like, e.g., "Guidance Control" or "Executive Control" would be); instead, it is a position in which control plays a central role. So again, that's why I'm wondering what is the resistance to just having it be a Position record.
If a Term definition will fit into multiple glossaries, then it should normally go in the parent of those Glossaries (which might be the General glossary); Glossaries aren't meant to be as granular as Debates and Positions. Glossaries can have multiple parents, however.
If the goal is not to have terms be targets of argument records, why can terms be united with positions? I've been using terms (united with positions) for positions with names that are terms, and positions without united terms for positions with names that are not terms. So e.g. the control theory of luck would be a term and a position because I can imagine coming across the term "control theory of luck," and so I thought I should put it in there as a term and not merely a position. But some position like "there is no such thing as moral luck" would not have an associated term because it's not like I would run across people using that term except accidentally.
The control theory of luck is not a more narrow form of control but it is more narrow form of luck, and luck is definitely a term. And I figure the control theory of luck should be related to control somehow. Nesting seemed like the best option since there's not really any other one.
The idea behind terms being united with positions is that you might find yourself mostly duplicating the description/key works for a term and a position. For example, I might have entered a term record for "Existentialism" and then found myself needing a position record for it as well. At the time (when I started encountering that situation a lot, and having to copy and paste description text from one record to another), I asked myself whether I should just allow terms to have an argument for/against them, fit under debates, etc. I decided it would be better to have them be separate records (so each record can enter into different relationships with other records) while eliminating the redundant descriptions/key works; hence "uniting" records was born. It has been about 6 years since then, and I never had a problem in the situation you are describing because I could always just have the position record be under multiple debate records.
You still haven't really answered my question as to why you don't find that sufficient. I get the impression it might be because you want to have a single hierarchical structure in the Tree tab containing every record, instead of separate hierarchies for Terms, Debates, Notes, and Labels. Maybe what you really need is the ability for the tree to switch to a different view that mixes the records into a single hierarchy, treating (for visualization/interactive purposes in the tree) united records as if they were one record (with only one tree node). In that case, I suggest you make an enhancement request. If you drag an argument onto a term without a united position, for example, it could create a position record behind the scenes and unite it to the term so that it will look like you've attached the argument directly to the term (but it would only do this when the tree tab is in "mixed mode").
I see what you mean about "Control Theory of Luck" corresponding to a particular conceptualization about the nature of luck. But wouldn't the corresponding term be something like "Control-based Luck" or something like that (Luck, itself, is not a theory)? Then that term would be united with Control Theory of Luck. That is similar to what I've done in the past. But even then, while the term is a narrowing of the concept of luck, like you said, it still doesn't seem obvious that it should be a child under the Control term record. But of course you could indicate the relationship by having the definition of control display in the definition of "control-based luck". But again, I'm guessing the reason you're not happy with that is because you want to see some relationship to control expressed in the hierarchical tree structure where you are seeing the term record in the Tree tab. For that, again, I suggest you make the enhancement request I described above.
I think the main reason is that it seems to me that the relevant things are terms (even if they're also positions) and so the terms should be related in the same way the positions are/would be related. If something makes sense for positions then why not the terms the positions are united with? I don't know what would be sufficient or not, really, since I'm not really sure what the optimal way of using Hypernomicon is. There are so many possibilities that it's hard for me to figure out what works best. But one thing I'm sure about is the hierarchical relationship between the relevant things. So I figure if the relevant things are in Hypernomicon and if there's the possibility of representing that hierarchical relationship, it's better to do that then not do that.
That's true for the specific example but other examples would instantiate the relevant properties more clearly. For instance I could have the term "socialism" and the term "anarchism" and the term "anarchist socialism," and the third would ideally go under both of the other terms.
I think rather than the enhancement you talk about I might just stop using terms that have associated positions. I was making those terms and then uniting them with the positions but I'm not sure that I get anything out of having the terms. The positions can already have a search key which will make the positions show up whenever the relevant word or phrase shows up, and I'm not sure what else being a term does. Uniting terms with positions seems sort of like a workaround to deal with the fact that you had a some terms that you later realized you wanted to treat partially like positions. Since I haven't run into this yet (since I started my Hypernomican database much later than you, already cognizant of what I wanted to do with positions and terms) I haven't yet created any terms that NEED to be united with positions or vice versa. I've just been uniting them since there's a "unite" button and I figured if a feature is there I ought to use it to see what it can get me. But I think this one is tricking me into making more terms than I should be making.
If the same structure is getting reproduced between sets of records for the same reason, then that seems like a bad form of redundancy because the burden is on the user to keep them in sync.
True, that is a good example where you might want both positions and terms for all those, and for the corresponding parent/child relations to be duplicated. But to me this case is somewhat exceptional in that regard (though not super uncommon), because it is a very simple case where "anarchist socialism" literally is just anarchism plus socialism; the fact that one is an adjective and the other is not is a little misleading. It could just as easily be "socialist anarchism" (with only a slight change of emphasis). In a case like that, duplication of record structure might be inevitable. But there are many other cases where a child position is not a simple sum of two parent positions that also have term records.
I think that you can probably go a long ways in Hypernomicon without using Term records.
But another place where they are useful is when the Term record is united with a Position record that has a different name. For example, I have a position record for "Instinct Theories" (with children that are particular types of instinctual theories of motivation) but the united term is "Instinct". They are united because whatever notes I want to write about the term Instinct should also go in Instinct Theories and vice versa.
As another example, I have a position for "Coherence Theory of Justification", which is united to the term "Coherence"; my "Coherence" term has the senses "Thermodynamic", "Epistemological", and "Semantic", and that position record is united to the Epistemological one. And I find it useful that those 3 "coherence" concepts share the same search key. So to me, that is a clear case where it is important to have Term and Position records that are separate.
Last edit: Jason Winning 2023-07-21
I still don't really need this since it's not a big deal, but I thought of one other reason I wanted this. I have a bunch of terms that can be categorized two ways: historically and topically. For example, pratītyasamutpāda (dependent origination) I want to have in both the "Buddhism" glossary (to indicate that it is a term central to Buddhism) and in the "Metaphysics" glossary (because it is a term relating to metaphysics).
Just looked up "pratītyasamutpāda" in Wikipedia because I've never heard of it. There it says it "is a key doctrine" stating that "all dharmas (phenomena) arise in dependence upon other dharmas." If that's correct, first, that is pretty interesting, and second, I'm wondering why not just have that be a Position record.
If it were me, I would have it be a position with problem/debate record "Metaphysics" being an ancestor, and also have it as an ancestor of the position "Buddhism" (to me, Buddhism is, first and foremost, from the standpoint of philosophy, a set of doctrines, i.e., positions). I would have it be a position for similar reasons stated above: it is the kind of thing there could be arguments for and against, etc.
If you want to categorize things historically (rather than being one of a set of doctrines), you could have Problem/Debate or Position records for things like "Ancient Greek Philosophy" (same applies to "Buddhism" if you are similarly thinking of that more as a historical period than set of doctrines/questions/debates) and have Problem/Debate or Position records about ancient Greek philosophers under that, or even Problem/Debate or Position records representing historical subdivisions like Presocratics (as well as under debate and position records organized more topically).
Maybe the real issue here is that "Problem/Debate" and "Position", as names of record types, don't do a good job of capturing how general they can actually be. Those names can mislead you into not using them to their full potential. Position records can potentially represent a huge collection of positions that are only related by being in the same historical period, etc. Maybe "Position" and "Problem/Debate" are misleading names for those record types. I'd be open to some other way of naming/describing them if so. Maybe "Problem/Debate" should have been called "Topic" and "Position" should have been called "Point or Region in Dialectical Space" to capture the real usefulness of them.
Think of Term records as inherently of more limited scope; they are there to solve a more specific problem (needing to track unrelated definitions sharing the same term, or track definitions and parent/child semantic relations that aren't tied to a certain topic or location/region of dialectical space) even though, outside of Hypernomicon, the word "term" has wider scope than "problem", "debate", or "position".
Maybe a better name for "Position" record would be "Position or Realm of Positions" and "Problem/Debate" records should be "Problem/Debate or Realm of Problems/Debates".