>Has anyone here done or seen work on a language relying on a grammar >that does not require the noun/verb distinction? I was faced with this issue when I tried to reinvent grammar from scratch. Before I would allow word types in the grammar I had to have it justified in some logical scheme. Are verbs and nouns required? Is the distinction required? I discovered in the analysis, that most definitely, verbs and nouns must exist and are distinct. HOWEVER, I found that there were two different issues. This is obfuscated by the fact that the words "verb" and "noun" in English have multiple meanings. In English "verb" means a word which semantically indicates an action (something that would be recognized in any culture and by speakers of any language). It also means a word which has a specific function in a sentence. It turns out, of course, that both of these things are required universally. That is, a language needs a way to express actions and it also needs a way to express the verbal function in a sentence. (Consider, for example, in Loglan where a "function" and an "argument" are required.) However, I realized that these two requirements were independent of each other. The reason that so many languages utilize the verb-noun model was because of a natural convenience. It is convenient for the action word to play the role of a "function" and a noun to play the role of an "argument". But even natural languages realize that this relation is not absolute as they employ verbal nouns, thus using action words as nouns. But here is the real mind-boggler! The verbal noun is NOT just the action in the role of a noun. It also includes the verbal function as part of the noun. That is, it is a noun that is actually referencing a sentence* (a sentence with just one word, the verb). This was a real difficulty and it took me some time to figure out that the action noun without the verbal function was merely the agent of the action. This solution was verified in Ankanian** when consistent conversions between nouns and verbs in either direction were re-entrant. That is, if you convert a noun to a verb and then you convert the verb into a noun, you end up with the same form and the same meaning. This would not work if the conversion of the verb to the noun resulted in a verbal noun/infinitive (e.g. Esperanto). >As I understand the matter, many >Native American languages modify their “action-words” (I decline to >call them verbs) in terms of the degree of validity with which the >speaker speaks. I also thought of employing this, but decided it would add unnecessary complexity, since this validity can be shown by other means as it is done in the majority of languages, and since one of the goals of Ankanian was to be simple. Even though ankanian was essentially an independent construction, general practises of natural languages were taken into consideration, the theory being that, if certain grammatical categories were present in a majority of languages, it might indicate a natural inclination of human thinking. This is why Ankanian has a basic plural form, even though, as some languages demonstrate, it is not technically required.*** =========================================================================== *Seeing this was difficult, since some English sentences seem to distinguish between the "fact" of the sentence and the verbal noun. But it seemed that these were actually artificial differences caused by English structure. **http://home.ix.netcom.com/~heensle/lang/avesta/parts.html ***Some might argue that I disobeyed the language parameters, since Ankanian has some grammatical elements that are unique. But these actually ended up being required to avoid other problems. For example, the "preexpectant" verbal form came about because I could not reduce the English adverb "already" to a derivative of a basic verb/noun type. This happened with a few adverbs, but not enough to require the creation of a whole new class of words.