Parsing

["A.W.T." <Ti@fa-kuan.muc.de>]

Trying to translate a Ch'i-lü style poem to Lojban, I'm experiencing proble=
ms when parsing the following line:
{.i latyte'a gi'e bau la madjar. sipta'a}
whereas after changing the word order it parses with
(i {<latyte'a VAU> gi'e <sipta'a [(bau {la madjar}) VAU]> VAU})

Does this mean that the first utterance is ungrammatical (which I doubt) or=
 just that the algorithm expects a selbri to immediately 
follow the {gi'e}?

co'o mi'e .aulun.