I think that it is time to re-evaluate the English-like concept of the word order in a sentence. In many languages, the word order in a sentence is rather flexible. In English, the position of words in sentences is necessary to inform, whether it is a noun, adjective, verb, subject, object, or something else. But in many natural indo-european languages (e.g. Latin, Greek, Slavic, ...) words have their own endings (declension and conjugation), in which is stored the complete information about their grammatical category, so there is no need any word to respect a specific position in the sentence.
We can say that English needs the fixed position of words in sentences in order to add words missing information about whether these words are subjects, objects, verbs or something else. It is obvious that other languages operate with words containing more unambiguous grammatical information without the need to use fixed positions. Free word order can then be used to express the finer details of communication in these languages. See our solution in our NS conlang: http://www.neoslavonic.org/lessons/2
Much better concept than English linear rules (e.g. first word, second word, third word, ...) is a concept of a branched tree we use: