From nobody@digitalkingdom.org Fri Nov 10 22:11:17 2006 Received: with ECARTIS (v1.0.0; list lojban-list); Fri, 10 Nov 2006 22:11:18 -0800 (PST) Received: from nobody by chain.digitalkingdom.org with local (Exim 4.63) (envelope-from ) id 1Gim4w-0004FQ-1y for lojban-list-real@lojban.org; Fri, 10 Nov 2006 22:10:58 -0800 Received: from ug-out-1314.google.com ([66.249.92.169]) by chain.digitalkingdom.org with esmtp (Exim 4.63) (envelope-from ) id 1Gim4p-0004FG-VX for lojban-list@lojban.org; Fri, 10 Nov 2006 22:10:57 -0800 Received: by ug-out-1314.google.com with SMTP id c2so660065ugf for ; Fri, 10 Nov 2006 22:10:50 -0800 (PST) DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:references; b=FvS5mFohGQauHzPGKtG+EV90Psk5iaUv1f2CY9wJVYqOmgCZur4saAE2Tj1GdLynjYIg3dUJQRtSn6xnSrVCbNXwSaat9C7ftrne2wzmf0yb4fkxojuUFNo6jSCsLhGJ+Tpzakf+T0XhE4cC/cuGQAPEqFB6z2u7idd62hjOArY= Received: by 10.67.19.17 with SMTP id w17mr4595802ugi.1163225450176; Fri, 10 Nov 2006 22:10:50 -0800 (PST) Received: by 10.66.216.14 with HTTP; Fri, 10 Nov 2006 22:10:50 -0800 (PST) Message-ID: Date: Sat, 11 Nov 2006 01:10:50 -0500 From: "Hugh O'Byrne" To: lojban-list@lojban.org Subject: [lojban] Re: D'ni orthography for Lojban In-Reply-To: <20061110223627.99365.qmail@web30408.mail.mud.yahoo.com> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_42339_26346301.1163225450124" References: <20061110223627.99365.qmail@web30408.mail.mud.yahoo.com> X-Spam-Score: -2.4 (--) X-archive-position: 13059 X-ecartis-version: Ecartis v1.0.0 Sender: lojban-list-bounce@lojban.org Errors-to: lojban-list-bounce@lojban.org X-original-sender: hobyrne@gmail.com Precedence: bulk Reply-to: lojban-list@lojban.org X-list: lojban-list ------=_Part_42339_26346301.1163225450124 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Content-Disposition: inline On 11/10/06, Jon MacLeod wrote: > > --- Hugh O'Byrne wrote: > > > On 10.11.2006, 4:50, Jon MacLeod wrote: > > > > I would like to know what Lojbanistan thinks of the D'ni orthography > I > > > am > > > > submitting... > > I think it's pretty. It looks like it's designed to be easy to write in > > cursive, adjacent letters join up nicely. That's a good feature. > > > > It seems to me that anyone using this for an extended time is likely to > make > > the 'i' shape more rounded, just because it's quicker and easier than > > drawing corners (making it less 'L' shaped and more 'C' shaped). (The > > intent may be for the scribe to lift the pen at the end of the top > stroke > > and put it back down near the beginning, but in a hurry, or after they > get > > familiar with making the shapes, I'll bet a lot of people will just keep > > their pen on the paper because it's easier. For example, which do you > find > > easier to write, cursive or print?) The backward strokes on 'f', 'p', > and > > perhaps 'z', look like they could be tiresome, I imagine after a while a > > scribe might get sloppy with these, making them into loops perhaps. It > is > > good there are only two different types of feature you have to go back > and > > do after scribbling the 'main' line of the word - the vertical bars, and > the > > short horizontal ticks (like dotting 'i's and crossing 't's); any more > than > > two features, I feel, may be unappealing. > > The D'ni characters are meant to written with the continous line done > first, > then to go back through the word and put in the accents (ticks) and flags > (vertical lines). Writing the wors as such may be odd at first, but it > does > enable someone to scribble something in D'ni script rather quickly. I kind of figured that was the way it was intended. Just so you know my motivation: I think if a symbol set is going to be widely used, many people are, after time, going to get sloppy about making the precise prescribed shapes. The degree to which that sloppiness affects legibility (specifically, the ability to quickly and with certainty classify different symbols) can be reduced with planning. So, I was trying to anticipate how the characters may change over time. Take, for example, lowercase 'l'. In print, it's a vertical line (perhaps with serifs). In cursive, it's commonly drawn as a tall loop. As such, the cursive 'l' really has a much more fundamental feature in common with the print 'o' than it has with the print 'l'. No-one I know about has confused a cursive 'l' for an 'o', and the two are still quite different in that the 'entry' and 'exit' points are typically at the bottom of the 'l' loop and at the top of the 'o' loop, but this demonstrates the principle of how sloppiness can affect the characteristics of symbols. Anticipating how it is likely to manifest itself could guide in the design of a symbol set whose features are more resistant to sloppines. > Now for my pet crusade. > > The D'ni language has 24 different characters, 11 with accented versions > (the > xy. and by., for example), and has 35 individual sounds. I honestly don't > know > why the creators of the Myst series of games decided to attach a certain > symbol > to a certain phoneme, I just know which one's are which. Ah. I did not realise that the typeface came from a game. I wonder if copyright is an issue. I think if a new alphabet is to be chosen for Lojban, it must have design considerations which are compatible and harmonious with Lojban. The first consideration I can think of is that it should be robust against time; i.e. robust against sloppiness. Features which may get smeared or de-emphasised in haste should, as much as possible, retain as much recognisability as possible, and be as easily distinguised from other smeared features as possible. The second design consideration I would put forward is orthogonality (e.g. that the presence/absence of the accent in a symbol correspond to a *particular* feature of the corresponding phoneme, e.g. whether the phoneme is voiced or not). There are interesting consequences and interplays to take into consideration. I find it quite fascinating. It elicits some deep thinking on how exactly perception works, when one explores how one symbol can be made 'as different as possible' from another symbol, within rules. Omniglot (http://www.omniglot.com/) is a good place to explore shapes of letters. If you're interested in alphabets and orthographies, a visit there is highly recommended. And, of course, the IPA ( http://www.arts.gla.ac.uk/ipa/fullchart.html) is a valuable resource for classifying the other end of the orthography: the phones and phonemes (different words! different concepts!) that are represented by the shapes on the page. I don't have anything against altering Lojbanic usage of the D'ni character > set > to be more correct :) aha, you'll get me in trouble. I know better than to try and claim I am 'correct' on this issue. I express opinions. Others will express other opinions. History will decide which opinions are more valid. Or it won't. - if you are wont to, by all means alter it as you see fit > to improve for Lojbanic usage, I would love to see the result, and I'll be > more > than happy to fix the font mappings to your results. I have mused on matters of orthography for years, but I'm still not sure my expertise is up to the high standard I have set for a 'perfect alphabet'. I guess I should get started trying. > "As a percentage of total universal knowledge, what I know is > statistically insignificant." - me I like that sentiment. It makes me more eager to share what little expertise I have, and collaborate with those whose goal is also to contribute to the collective knowledge. mi'e .xius. -- Good night, and have a rational tomorrow! ------=_Part_42339_26346301.1163225450124 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline
On 11/10/06, Jon MacLeod <eye_onus@yahoo.com> wrote:
--- Hugh O'Byrne <hobyrne@gmail.com> wrote:
> > On 10.11.2006, 4:50, Jon MacLeod wrote:
> > > I would like to know what Lojbanistan thinks of the D'ni orthography I
> > am
> > > submitting...
> I think it's pretty.  It looks like it's designed to be easy to write in
> cursive, adjacent letters join up nicely.  That's a good feature.
>
> It seems to me that anyone using this for an extended time is likely to make
> the 'i' shape more rounded, just because it's quicker and easier than
> drawing corners (making it less 'L' shaped and more 'C' shaped).  (The
> intent may be for the scribe to lift the pen at the end of the top stroke
> and put it back down near the beginning, but in a hurry, or after they get
> familiar with making the shapes, I'll bet a lot of people will just keep
> their pen on the paper because it's easier.  For example, which do you find
> easier to write, cursive or print?)  The backward strokes on 'f', 'p', and
> perhaps 'z', look like they could be tiresome, I imagine after a while a
> scribe might get sloppy with these, making them into loops perhaps.  It is
> good there are only two different types of feature you have to go back and
> do after scribbling the 'main' line of the word - the vertical bars, and the
> short horizontal ticks (like dotting 'i's and crossing 't's); any more than
> two features, I feel, may be unappealing.

The D'ni characters are meant to written with the continous line done first,
then to go back through the word and put in the accents (ticks) and flags
(vertical lines). Writing the wors as such may be odd at first, but it does
enable someone to scribble something in D'ni script rather quickly.

I kind of figured that was the way it was intended.

Just so you know my motivation:  I think if a symbol set is going to be widely used, many people are, after time, going to get sloppy about making the precise prescribed shapes.  The degree to which that sloppiness affects legibility (specifically, the ability to quickly and with certainty classify different symbols) can be reduced with planning.

So, I was trying to anticipate how the characters may change over time.  Take, for example, lowercase 'l'.  In print, it's a vertical line (perhaps with serifs).  In cursive, it's commonly drawn as a tall loop.  As such, the cursive 'l' really has a much more fundamental feature in common with the print 'o' than it has with the print 'l'.  No-one I know about has confused a cursive 'l' for an 'o', and the two are still quite different in that the 'entry' and 'exit' points are typically at the bottom of the 'l' loop and at the top of the 'o' loop, but this demonstrates the principle of how sloppiness can affect the characteristics of symbols.  Anticipating how it is likely to manifest itself could guide in the design of a symbol set whose features are more resistant to sloppines.

> Now for my pet crusade.

The D'ni language has 24 different characters, 11 with accented versions (the
xy. and by., for example), and has 35 individual sounds. I honestly don't know
why the creators of the Myst series of games decided to attach a certain symbol
to a certain phoneme, I just know which one's are which.

Ah.  I did not realise that the typeface came from a game.  I wonder if copyright is an issue.

I think if a new alphabet is to be chosen for Lojban, it must have design considerations which are compatible and harmonious with Lojban.  The first consideration I can think of is that it should be robust against time; i.e. robust against sloppiness.  Features which may get smeared or de-emphasised in haste should, as much as possible, retain as much recognisability as possible, and be as easily distinguised from other smeared features as possible.

The second design consideration I would put forward is orthogonality (e.g. that the presence/absence of the accent in a symbol correspond to a *particular* feature of the corresponding phoneme, e.g. whether the phoneme is voiced or not).

There are interesting consequences and interplays to take into consideration.  I find it quite fascinating.  It elicits some deep thinking on how exactly perception works, when one explores how one symbol can be made 'as different as possible' from another symbol, within rules.

Omniglot (http://www.omniglot.com/) is a good place to explore shapes of letters.  If you're interested in alphabets and orthographies, a visit there is highly recommended.  And, of course, the IPA ( http://www.arts.gla.ac.uk/ipa/fullchart.html) is a valuable resource for classifying the other end of the orthography: the phones and phonemes (different words! different concepts!) that are represented by the shapes on the page.

I don't have anything against altering Lojbanic usage of the D'ni character set
to be more correct

:) aha, you'll get me in trouble.  I know better than to try and claim I am 'correct' on this issue.  I express opinions.  Others will express other opinions.  History will decide which opinions are more valid.  Or it won't.

- if you are wont to, by all means alter it as you see fit
to improve for Lojbanic usage, I would love to see the result, and I'll be more
than happy to fix the font mappings to your results.

I have mused on matters of orthography for years, but I'm still not sure my expertise is up to the high standard I have set for a 'perfect alphabet'.  I guess I should get started trying.

> "As a percentage of total universal knowledge, what I know is statistically insignificant." - me

I like that sentiment.  It makes me more eager to share what little expertise I have, and collaborate with those whose goal is also to contribute to the collective knowledge.

mi'e .xius.
--
Good night, and have a rational tomorrow!

------=_Part_42339_26346301.1163225450124-- To unsubscribe from this list, send mail to lojban-list-request@lojban.org with the subject unsubscribe, or go to http://www.lojban.org/lsg2/, or if you're really stuck, send mail to secretary@lojban.org for help.