because she hit her head and fell from the boat, she drowned

From Lojban
Jump to: navigation, search

This is an excellent demonstration case for ".ijoibo":

darxi fi le stedu .ijoibo farlu fi le bloti .iseri'ake jacmrobi'o

If .ije appeared instead of .ijoi. it would mean "Because she hit her head, she drowned; and because she fell from the boat, she drowned".


More commonly it would be

darxi fi le stedu gi'e farlu fi le bloti .iseri'ake jacmrobi'o

Not the point. The semantics of je/gi'e vs. joi are what count here; using either .ije or gi'e fails to capture the essential fact that both are needed to correctly express the cause.

I was about to flame this, and then... realised it's absolutely correct, and a distinction logicians don't make often enough. Kudos! -- nitcion (and myxyl, if that's you, sorry for giving you such a hard time. :-)

  • John Cowan transcribed it, but the credit for this example, and its resolution, goes to ... wait for it ... TLI! I found it while scanning their Loglan Updater for interesting ideas.

Please explain. Doesn't gi'e require both parts to be true? How is that a closer binding than joi? If you really want to be pedantic, use ganai gi with da'inai. What am I not seeing here?

OK, let's take this slowly:

  • lenu darxi fi le stedu kei .e lenu farlu fi le bloti cu rinka lenu jacmrobi'o

This claims that A causes C and that B causes C, not that the conjunction of A and B cause C (i.e. not one alone could have done it.) It can be rephrased as:

  • lenu darxi fi le stedu kei cu rinka lenu jacmrobi'o .ije lenu farlu fi le bloti cu rinka lenu jacmrobi'o

Ditto:

  • lenu darxi fi le stedu gi'e farlu fi le bloti cu rinka lenu jacmrobi'o

which can also be rephrased as

  • lenu darxi fi le stedu kei cu rinka lenu jacmrobi'o .ije lenu farlu fi le bloti cu rinka lenu jacmrobi'o

That is false. See the Book chapter 14, section 19 (p. 365).

The question now is, does this generalise past a sentence boundary? Does

  • darxi fi le stedu gi'e farlu fi le bloti .iseri'ake jacmrobi'o

expand to

  • {darxi fi le stedu .i bo seri'ake jacmrobi'o} {.ije farlu fi le bloti .ibo seri'ake jacmrobi'o}

... And, now that I think of it, surely it doesn't.

  • Why "surely"? I think it surely does. See the discussion of how logical connectives break sentences in two, whereas non-logical connectives don't, in the refgram. --John Cowan
  • Well, I did, but I thought that any such breaking up was restricted in scope to the current sentence, and you couldn't do the rephrasing above. Guess not, huh?

On the other hand, without the joi, I still think this makes a weaker claim: {A and B} caused C, which is not the same as A caused C and B caused C, but is also not the same as A and B, occuring in conjunction, caused C. If you want to stress that it was the combination, rather than the cooccurrence of the two, that did it, you say joi.

I don't think the refgramm resolves this. And you know, this really is the kind of thing that should be defined formally somewhere. Anyone know any better? -- nitcion, who now wonders whether he shouldn't take this to the list...

What is the difference between combination and coocurrence? --xod

... You're right, there is none. Either the combination of A and B cause C, or they can each cause C separately; it was bogus of me to claim there is some middle ground. John's point still stands, though: gi'e is not the same as joigi. -- nitcion.


Off topic discussion moved to da'i