(Originally appeared in Lognet 91/4)

Assigning Case Tags

by James Jennings

The following article was adapted from a letter to JCB.—Tisra

I am fascinated by case tags. I have spent a lot of time trying to understand how to assign them. I am convinced that we need some general principles, or perhaps a collection of standard patterns so that your average case-tag-assigner can be consistent with what has been assigned before.

I have several ideas for such patterns. Let me propose two.

My first idea is perhaps the most controversial, so I will take some time to defend it. I also think it is rather important, so I call it:

The Golden Rule: A single case tag has no meaning by itself. The relation between the tags assigned to a predicate is what is important.

Corrolary: One place predicates can always be assigned the “B”-tag pattern.

The reason for this rule is that [Loglan] predicates can fulfill any [English] grammatical role: noun, verb, adjective, whatever. If one tag seems more appropriate than another, it is only because you are being misled by the English translation. Consider Da sonli = ' X is sleeping' (K, X is acting)[while in] 'X is a sleeper' (B, X has the property of, the ability to sleep). I claim that the 2nd is a better translation because it better represents Loglan’s timeless tense.

I took the case tag list from NB3 (1st with the original list and later with [Dr. McIvor's] update) and sorted it by tag pattern. This made it easy to find the one place predicates, and a lot of mistakes for that matter. How many one-place predicates are not “B”? Not many.

dalra C is a dollar (a mistake: see peso, ruble, mark, etc.)

brute K breathes. (how about: B is a breather.)

clivi K is alive. (B is a living thing.)

kraku K cries out. (B is a crier, is capable of crying)

tsani V is a sneeze. [This entry in the NB3 list is errroneus. It should be 'tsani sneezes'; see L1:530.—JCB] (B is a sneeze. Use ‘po’ for the event of sneezing.)

And the international declensions follow the pattern:

junga B is of the Chinese language. (OK)

jungi K is a Chinese person. (X has to actively be Chinese?)

jungo V is of the Chinese culture. (B has the property of being Chinese.)

ponjo P is of the Japanese culture. (Is it P or V? Make up your mind!)

As I hope my parenthetical comments show, all of the above could be cast as a “B” without losing anything. I believe that they should be so cast.

Possible Exception to the Corrolary: One place complexes might want to reflect the structure of their root words. I’m not sure I like this exception. It depends on how often complexes do borrow the place structures of their roots. It should be considered.

My second idea:

Matma Class Words Should Use S, D, V or Equivalent:

NB3 says: 'matma S is the mother of P by K'. I can almost buy the S-P (mother is the Source, child is the Product) but the K looks tacked on because nothing else fits. The problem is that we have a one-way [or non-reciprocal] relationship between equivalent objects, three human beings. If it were a [reciprocal] relationship, we could do something like [the pattern in the entry for] narti: 'B is separate from B'.

My inspiration comes from [another entry]: 'rutma B is a path from S to D via V'. The S, D, and V are equivalent; they are all places, and yet their order matters. It is a one-way [or non-reciprocal] description. I propose:

matma S is the mother of D by/via father V.

brudi S is the brother of D with parents V.

In a sense, motherhood “points” from mother to child. This is a more metaphorical approach to assigning tags than most, but I like it.

Eo cutse letu [nu] jupni mi.