Excuse what might turn out to be a fairly rambling message; there's
been a lot to reply to but in not having a chance to do so I've
probably lost some thoughts.

On Thu, Sep 10, 2009 at 6:39 AM, Christophe Grandsire-Koevoets
<[log in to unmask]> wrote:
> 2009/9/9 G. van der Vegt <[log in to unmask]>
>
>> 2009/9/9 Lee <[log in to unmask]>:
>> > I learned the cursive "X" (uppercase) as "96". "x" (lowercase) was more
>> of a fluid "\" connected with adjacent letters, finishing with "/" after the
>> word was written.
>>
>> I learned uppercase X as what was basically an upside down version of
>> yours, lower case was )/(, with the slash ideally blending into the
>> parantheses.
>>
>
> Same here. My handwriting still has them as I learned them (unlike other
> letters which have mutated much since then).
>
> In any case, I'm not surprised that there are many variants of cursive
> styles taught to children across the world. That's one of the effects of
> having a script for which general shapes are more important than how you
> actually drew those shapes.
>
> Even in a single cursive style sometimes letters end up being drawn
> differently depending on context. I know I have two ways to write my t's,
> depending on the letters around it (basically the one-stroke and two-stroke
> methods Paul Hartzer mentioned, with the added fun that my one-stroke t's,
> when they connect to the next consonant, do so via the horizontal end. I
> typically do that for the t+h sequence, since I loop my h's vertical bar).
> My o's look different depending on whether they are followed by an f or not.
> And I have three ways to write my 9's.
> --
> Christophe Grandsire-Koevoets.
>
> http://christophoronomicon.blogspot.com/
> http://www.christophoronomicon.nl/
>

So a good system of describing letters would, I think, really need to
consider handwriting just a representation of the more essential
printed letters. However, such an approach could be dangerous since
the printed letters could occasionally be influenced by the
handwritten ones, and I'm sure the changes which took place in the
middle ages necessarily involved not having a class of 'more
essential' letters.

A LOTEP based system or a topologically based one would not be
appropriate except when those features are the ones readers and
writers consider graphemic. Since topology is not graphemic in today's
writing, it can be varied pretty freely: I don't care if my <a>
connects all the way and I add closed loops onto <g>'s tail, and I
have two topologically different versions of <p>. However I have a
professor who I'm convinced sees the topology of 2 as the main
graphemic feature: there is always a little curl at the top and at the
joint in his handwriting, but he feels free to stretch the letter out
or leave off most of the bottom bar. His question marks appear to be
the same graph, and he will write a big red question mark over an
assignment as little more than a more or less straight line over the
whole sheet with a little curl at the top and bottom.

So these are good things for the list of what *can be* graphemic, but
they would only be necessarily specified in a narrow transcription.

What I'm more interested in, though, is describing a writing system's
graphemic inventory, which I suppose could get pretty narrow.

On Wed, Sep 9, 2009 at 12:22 PM, Paul Hartzer <[log in to unmask]> wrote:
[snip]
> Because of the complexity of the characters, it's important that people know how many strokes it takes to make a character, and Chinese has strict rules about strokes, stroke direction, and stroke order. 囗 (enclosure) consists of *3* strokes (not 4, which outsiders might expect), in this order: the left bar, the top and right bars in a fluid stroke, and the bottom bar. Any other way of drawing that character is considered incorrect.

This reminds me of how I was taught that writing <0> and <o> and <O>
HAD to start from the top of the letter. Many in our class were
starting at the bottom and we spent part of a day practicing and were
informed we would be marked down on exercises if we didn't keep the
habit of starting at the top. For a little while I started at the
bottom but wrote very carefully to cover it up, just for fun.

> None of this says it's impossible to create a dictionary of possible strokes, but if such a thing existed, there are significant issues to consider. Since order and direction are so important in Chinese, we'd have to decide how to treat those; off hand, I'd say they're similar to issues like tone, so they'd get their own marking.

Writing systems, just like spoken languages, have things that they
contrast for and things they ignore. In English and (afaik) the Roman
alphabet in general stroke order does not matter. But conscripts
should decide for themselves whether orientation, absolute or relative
height or placement, stroke order, topology (or subsets of it), etc.
etc. matter.


The bottom line for me is I need to study world writing systems more
before I could really attempt this project. All I've studied so far in
this direction is some Latin palaeography.