Without qualification, refers to the article Loglan written by James Cooke Brown, and published in Scientific American in June 1960.
People who wish to read it, may find it online, can borrow it from a reasonably large library, or ask TLI to send them a copy. That's how I got it. That was in 1997?, but if they have any copies left, they probably still mail them out for free to anyone who asks for them.
The reason that the CV templates of the Loglan words turned out the way we all know and love, is not given. It only says: "The reader is challenged to find a combination of possible word-forms that does not resolve."
The "little words" (cmavo) are unified as a class only semantically, not morphologically (phonotactically). The word classes preceding predicates (simple predicates (our gismu) and complex predicates (our lujvo)) were as follows:
The Loglan text that occurs in the tables in the article is written in all capitals, but the examples that occur inside the running text, is written in the SAE? orthography that is common both to natural languages that use latinate scripts, and Loglan of today. Still, there is some talk about audiovisual isomorphism, and
Loglan's "spoken punctuation" operators.
Most indicators (our attitudinals) were irrealis. In fact, there is a theoretical possibility that ALL of them were irrealis. If we entertain the possibility that even "ui" (our "ui") turns a predicate into a non-claim, that does not entail that it turns it into a claim of the opposite.
The only example of borrowing of vowel-final names, Mississippi, which becomes "lu misisipis", displays the addition of an s in the end, which is now the de facto standard of creating such names in Lojban (though not governed by any rules).
The original document is available at http://tiki.lojban.org/tiki/Scientific+American+article