Theiling Online    Sitemap    Conlang Mailing List HQ   

Re: Word order (Was: Conlangs of mischief (Which in turn was: Re: I'm back!)

From:Ray Brown <ray.brown@...>
Date:Sunday, September 26, 2004, 10:19
On Friday, September 24, 2004, at 07:28 , David Peterson wrote:

> Ray wrote: > > <<It is a problem - it's an impossibility if we are producing sounds > serially or writing in any way that is recognized as writing. The sounds > (and characters) come one after another, i.e. there is an order.>> > > Oh, oops.  I of course meant "totally free word order".  If words > *couldn't* be ordered, then, logically, they couldn't very well be > produced, could they? 
No, they couldn't. That's why I was questioning "no word order" claim currently being touted for some conlangs. If the the word order is not determined by syntax, it will be determined by other considerations such as focus & topicalization - even, as we see the rhetorical styles developed by the ancient Greeks & Romans, by considerations of rhythm. [snip]
> But I can't imagine a language where you could take, for example: > > "The man on the roof gave a book with a blue cover to an unhappy girl in > the garden." > > And produce: > > "The a a an the on the in man roof to with blue gave garden book unhappy > cover girl." > > Even if every element was so explicitly marked that there was no way of > confusing > which elements formed constituents and which didn't.
Quite so - in the latter case, it would be possible for a computer program to detect all the explicit markings and generate a nice little parse tree that could then be read off in a more intelligible word order. But "The a a an the on the in man roof to with blue gave garden book unhappy cover girl" is, I think, not a possible word order for a human-to-human language, even with explicit marking of every element. ============================================================== On Friday, September 24, 2004, at 08:40 , Rodlox wrote: [snip]
>>> Take >>> a look at the Metes text, and how I translated it >>> Metes: >>> >> >> I have looked - the letters come serially. There is order. >> >> Even if Metes makes single utterances one word, the words are ordered, >> unless everyone is speaking at the same time. > > that's a whole other conlang, and one more suited to a herd species. :)
Maybe - but I assure you it is not uncommon among humans. However, it rarely IME leads to satisfactory communication. What I was getting at was: - it was claimed Metes has no word order; - it is also claimed - and indeed you have reiterated the claim - that Metes is polysynthetic & whole sentences form single words; - if there is _no order_ then all sentences are produced at the same time. If the sentences are not produced at the same time, but one after the other, they are ordered, i.e. as far as I can see, Metes must have word order.
>> The single-utterance words >> are obviously composed of morphemes - they are surely ordered, or did >> Rolox simply get all his morphemes and then apply a randomizing function >> before writing them down. > > I used Proto-Indo-European as a guide (read, I tried to apply sound > changes > to Proto-Indo-European words),
This has been done several hundred times in the last 5000 years or so ;)
> and the result was Metes. > > my *big* problem, imho, was that, when I crafted Metes, I accidentally > confused grammar with word order...since, up to that point, everyone I > talked to (admittedly, not in school & off the internet), had seemed to > use > them interchangeably.
They are not interchangeable. Word order is to do with that part of grammar known as syntax. 'Grammar' is one of those words that has acquired a range of related meaning. Trask lists three meanings: "1. The system by which words and morphemes of a language are organized into larger units, particularly into sentences, perceived as existing independently of any attempt at describing it. "2. A particular description of such a system, as embodied in a set of rules. "3. The branch of linguistics dealing with the construction of such descriptions and with the investigation of their properties, conventionally divided into _morphology_ and _syntax_." Morphology: "The branch of grammar dealing with the analysis of word structure, conventionally divided into _derivational morphology_ (the study of word formation) and _inflectional morphology_ (the study of the variation in form of single lexical items for grammatical purposes)." Syntax: "The branch of grammar dealing with the organisation of words into larger structures, particularly into sentences; equivalently, the study of sentence structure." It might be argued that in Metes morphology & syntax are the same thing - but that is quite different from no word order. =================================================
> On Friday, September 24, 2004, at 08:43 , Muke Tever wrote: > > On Fri, 24 Sep 2004 21:40:11 +0200, Rodlox <Rodlox@...> wrote: >> (initially, I was puzzled...most of Proto-Indo-European seemed to be >> prefixes (what with alll the - at- the- end- of- words- ). > > Those aren't words or prefixes, but stems that endings must attach to.
Lexical morphemes, in fact.
> To make a parallel example from a natural language, one could make a > Latin entry _gaud-_ "rejoice", covering all derivatives such as _gaudeo_ > "rejoice", _gaudium_ "joy" etc. but also with the hyphen to note that > _gaud_ itself is not a word.
And gaud- itself is derived from the _root_ morpheme gau-, which we see in: gaui:sus sum "I rejoiced", the perfect indicative tense of the _gaude:re_, to rejoice. Indeed, whether we want to analyze _gaudium_ as gaud-d-i-um, and _gaudeo_ as gau-d-e-o is questionable; it is difficult to give meaning to each morpheme in such an analysis. I think _gau-de-_ and _gau-di-_ might be simpler. Certainly, we should say that: gaudi - is the _stem_ of the adjective _gaudia:lis_ "joyful" (base: gaudia: l- ); and both the _stem_ & _base_ of the noun _gaudium_ "joy". gaude: is the _base_ of the infectum (the so-called 'present stem' tenses etc) of _gaude:re_ "to rejoice". gaui:s- is the _stem_ of the future active participle _gaui:su:tus "going to rejoice" (base: gaui:su:r- ); and both the _stem_ 7 _base_ of the perfect active participle _gaui:sus_ "having rejoiced". _gaude:_ occurs as an unbound form in the singular imperative "rejoice!". It is, therefore, not stem, but it is the base form of all the tenses and other forms of the infectum. Note (for those unfamiliar with the terminology): - root morpheme = "In morphology, the simplest possible form of a lexical morpheme, upon which all other bound and free forms involving that morpheme are based" [Trask] - stem = "In morphology, a bound form of a lexical item which typically consists of a _root_ to which one or more morphological _formatives_ have been added and which serves as the immediate _base_ for some further form or set of forms" [Trask]. - base = "In morphology, a morph, variously consisting of a root, a stem or a word, which serves, upon the addition of a single further morpheme as the immediate form of some particular formation" [Trask] Beware: there are other meaning of 'base', for example in Transformational Grammar and Government-Binding theories of derivational grammar ;) Ray =============================================== =============================================== "They are evidently confusing science with technology." UMBERTO ECO September, 2004