ROILAing in the Grave

Originally written 1/9/12

So I bought a copy of Learn ROILA. ROILA stands for Robot Interaction Language, a strange marriage between the goals of Lojban and the syntax of Toki Pona, which is intended as a means of communication between robots and humans. It is allegedly simple enough for humans to learn and clear enough for robots to understand. Since I do not own a robot, I cannot speak about the mechanical end. Thus my comments will be limited to the human side.

The vocabulary generation is atypical of a constructed language, the opposite of the oligosynthetic end of the conlang scale. The method of vocabulary generation ensures maximal distinction of words, but results in atypically long function words and semantic and phonological disconnections.

The vocalic phonology is a little odd. The choice of a five-vowel system is not extraordinary, but the specific pronunciations of the vowels (and this is not a language that favors allophony!) are not as spread out as a 'normal' vowel set. A human can manage it perfectly well, but the distribution is maximized for robotic visual perception. An examination of the consonantal phonology bears this out, since the inventory is heavily weighted towards the labial/labiodental column, which is the most visible point of articulation.

The use of 'tuji' as a plural marker, intensifier, and 'many' suprised me. One of the things I do not like about Toki Pona is the polysemy of the word 'mute' (many) for these meanings. In the case of Toki Pona, the paucity of the vocabulary excuses this polysemy, but one would expect greater clarity from a language designed for machines that need precise instructions.

What especially shocked me was the near-homophony of the markers for past and present tense. 'Pito make jifi lakowo' means 'I saw the cat', and 'Pito make jifo lakowo' means 'I will see the cat.' 'Pito make lakowo', without any tense marker, means 'I see the cat'. Given that one of the frequently mentioned faults of Esperanto is the similarity of the past, present, and future markers on the verbs. This is not only a problem for the eterna komencanto, but for fluent Esperantists; otherwise, why would fluent Esperantists abbreviate the past/present/future forms estis/estas/estos , with accent on the e, to stis/stas/stos, with accent on the i? I am sure this critique arose during the creation of ROILA, and that the creators have good reason for not removing it.