Re: How to write down a language design?

Thank you, dear fellow conlangers, for your replies! :-))

Even though some of you thought that I was referring to
an external reference grammar, I actually meant the personal
notes which a conlanger makes to keep track of his/her
invention.

So far my notes consist of mere bits and pieces - mostly
examples and test sentences which I compiled into a big
text file. The major problem arose when I changed my
inflectional system - I had to redo all my sentences
and was about to get confused with some ideas, thinking
hard how I had designed them before I partially changed
my mind. This way of keeping notes is natural to the
design phase but it is such a mess that I have to do
something against it. ;-)

Ok, you have convinced me: I'll write a reference grammar
- for anybody who will desire to read it but firstly and
most neccessarily for myself. Obviously I should precisely
explain to even myself how my language works. Since I came
up with a very strange way to represent the meanings of
nouns, verbs and sentences, I should probably make sure
that I document my little thought experiement quite well.

So, thanks for your comments! I shall go and begin my
reference grammar now. :-))

Yours smiling and waving,
Harald :-)

>Quoting Andreas Johansson <andjo@...>:
>
>> Quoting Harald Stoiber <stoiberh@...>:
>>
>> > Hello ladies and gentlemen,
>> >
>> > Please view this as a kind of brainstorming query to get myself ideas
>> > how other people handle this matter. And here is the matter:
>> >
>> > How do you keep notes of your creations? Especially I am talking about
>> > the case of an unusual grammar and/or syntax. Which way do you prefer -
>> > tons of examples plus explanations? Or some more abstract description?
>>
>> I usually write really dense notes, but accompanied by (frequently
>> untranslated) examples.
>
>This, of course, refers to my personal notes. When I make a presentation
>supposed to be readable for others, the result is much like my webpage at
>andjo.free.fr/conlang .
>
>                                                Andreas

Reply