Why Are Thai Letters So Hard To Read?

[Richard W]

.... why are all thai fonts so small???

<{POST_SNAPBACK}>

I think you mean "why are they so difficult to read when they are small" and I think this is not so much that they're so small, it's that you (and me) need to be able to identify every loop and curve in each character in order to recognise them. Thai people will recognise them just from the general shape - just think of your own language and how small the letters can be while you still can recognise them. There are only a few bumps, loops and curves on the Thai letters that differentiate them, but you - and I - as beginners just haven't learnt which bits are critical.

<{POST_SNAPBACK}>

I disagree on two counts. Firstly, the combination of a Thai letter and the vowels and tonemarks above and below it is is much taller than an English letter, especially in handwriting (think of กี่). Normally, when scripts are mixed in a posting, the characters are all taken from fonts with the same pitch. However, the pitch is the distance between lines, so Thai characters of the same pitch as Roman letters will be much smaller. Tahoma, which seems to be the best font for mixed text, reduces this problem by squeezing the tonemarks and superscript vowels. Recent versions of Word (or is that only the Thai edition) addresses this problem by having different fonts and pitches for different scripts within the same 'style'. Perhaps one day Web browsers will similarly solve the problem.

Secondly, I think the differentiating squiggles in Thai letters are relatively small in Thai letters. For example, there is only a sight difference between ก, ภ and ถ. I came to the conclusion long ago that although the letter size of the Roman script in bilingual road signs was smaller than the size of the Thai letters, it was genuinely more readable than the Thai, and not just because I am far more familiar with the Roman alphabet.

Where Thais do have an advantage is that they are probably reading a word at a time rather than a letter at a time. At that level of recognition, the smallness of the squiggles probably doesn't matter.

On a practical point, and I am deliberately leaving the letters small here, it is very difficult to distinguish ด and ต. The loop is lower in the latter, but that is well within the normal range for ด.

[The_Moog]

ifwedidnotleavegapsyouwouldgetusedtoiteventually

[RDN]

.... why are all thai fonts so small???

<{POST_SNAPBACK}>

I think you mean "why are they so difficult to read when they are small" and I think this is not so much that they're so small, it's that you (and me) need to be able to identify every loop and curve in each character in order to recognise them. Thai people will recognise them just from the general shape - just think of your own language and how small the letters can be while you still can recognise them. There are only a few bumps, loops and curves on the Thai letters that differentiate them, but you - and I - as beginners just haven't learnt which bits are critical.

<{POST_SNAPBACK}>

I disagree on two counts. Firstly, the combination of a Thai letter and the vowels and tonemarks above and below it is is much taller than an English letter, especially in handwriting (think of กี่). Normally, when scripts are mixed in a posting, the characters are all taken from fonts with the same pitch. However, the pitch is the distance between lines, so Thai characters of the same pitch as Roman letters will be much smaller. Tahoma, which seems to be the best font for mixed text, reduces this problem by squeezing the tonemarks and superscript vowels. Recent versions of Word (or is that only the Thai edition) addresses this problem by having different fonts and pitches for different scripts within the same 'style'. Perhaps one day Web browsers will similarly solve the problem.

Secondly, I think the differentiating squiggles in Thai letters are relatively small in Thai letters. For example, there is only a sight difference between ก, ภ and ถ. I came to the conclusion long ago that although the letter size of the Roman script in bilingual road signs was smaller than the size of the Thai letters, it was genuinely more readable than the Thai, and not just because I am far more familiar with the Roman alphabet.

Where Thais do have an advantage is that they are probably reading a word at a time rather than a letter at a time. At that level of recognition, the smallness of the squiggles probably doesn't matter.

On a practical point, and I am deliberately leaving the letters small here, it is very difficult to distinguish ด and ต. The loop is lower in the latter, but that is well within the normal range for ด.

<{POST_SNAPBACK}>

In my original post in the "De" thread (here http://www.thaivisa.com/forum/index.php?sh...dpost&p=303116) I mentioned that raw reua ร looks like an S in some fonts and it was this simplification of Thai characters that I was trying to explain.

But back to the first point, I agree that in order to keep the pitch the same, computers when displaying mixed Thai and English characters, must make the Thai characters smaller, to give room for the symbols above and below the characters.

The point about differentiating similar Thai characters that I was trying to make - not too clearly I admit - was that some characters which have loops, and are similar to other characters that have loops, don't really need all the loops to keep them unique.

I'll try to explain by showing the fonts in BP Becker's dictionary: (If you see "user posted image", just refresh the page - the picture will appear eventually)

For example, naw noo and maw ma both have loops top left:

but they don't need to, to keep them unique. Just look at the second font shown for each character - no hint of a loop, just a straight line either right or left. So the point I was trying to make was that beginners - like me - will look at the character in it's entirety to determine what it is, but with experience you learn not to worry about the loop at top left, it's the position of other loop that determine which character it is.

Similarly saw reu-see:

It looks pretty complicated, but the second font is just a U with a line through the right hand side - and that's the important thing to look for when identifying it. You just don't have to worry about the top left loop or the bottom left kink.

And it is just that line through the right side that differentiates it from baw bai-mai:

So really, all I tried to say was, when you know what to look for in Thai characters that makes them unique, and therefore identifiable, it's a lot easier to read Thai even when the body of the characters is smaller than the Roman characters.

I also agree with your point about word recognition - sometimes I notice a group of characters and can identify the word from that group. But - as yet - not too often

[The_Moog]

Right - and its that second 'simplified' font that appears quite often in everyday life. Newspaper headlines and posters in particular.

Vowels in that font are little more than slashes.

[Boon Mee]

In many texts it's hard to distinguish between "mai toe" ้ and "mai han aakaat" ั unless you're totally familiar with the grammer rules which I'm not.

One thing I like about Higbees "Thai Reference Grammer" is the way mai toe is always positioned significantly higher in all situations than "mai han aakaat".

Makes it a little easier.

[Thaiboxer]

The modern Thai font used in many mall and restaurant advertisements is VERY difficult to read for beginners. The standard font isn't too bad if it's not really small. For me, I look for indicators that help me distinguish one character from another.

For example, "lor ling" looks exactly like "sor seua" except that "sor seua" has that little hook that sticks up from the top right side.

"Dto dtao" and "dor dek" are identical except for the little notch in the top of "dto dtao".

Others are obvious since no other character looks like it. Examples are "ngaw ngu" and "wor waen".

For the vowels it gets tricky. Certain consonants in a word or phrase can change not only the vowel sound but also the tone as well. Sometimes, the consonant "hor heeb" will give the word a rising tone even though the "H" sound is silent. A Thai word ending with "lor ling" has an "N" sound when it is pronounced. The vowels, tones and little grammar tricks are the hardest things to learn I think. Thus far I have learned most of the consonants and about half the vowels and I can already make out many words. I have a long way to go though.

As one poster said, Thais tend to read words and phrases as whole units and not letter by letter. A vowel in a word may come before a consonant yet be pronounced as if it came after (like in English).

[Andrew Mac]

I have the feeling that some of you may be interested in a paper called "How do Thais Tell Letters Apart?". It was written by someone called Doug Cooper at the Center for Research in Computational Linguistics (which I found while surfing around one day!).

A direct link to the document is http://seasrc.th.net/paper/tellthai.zip -- it's a 483K zipped postscript file, so you'll need to be able to convert that format into something more usable (such as PDF), or send it directly to a postscript printer. Converting it to PDF results in a 1.1 mb file, which I can't send to too many people ... but if a few of you are interested, and can't convert it, I might be able to send it to you. Alternatively, someone might have somewhere they can place it for easy download.

Doug's interest looks to be mainly Optical Character Recognition, but he has written quite a few other papers that might interest people. If you'd like to have a peek, you can go to http://seasrc.th.net/paper/paper.htm

I once tried to get in touch with Doug, to ask about a resource that that had been posted on the site, but with an incorrect link. Unfortunately, the email address didn't work any more. I'm not sure what happened to the whole project ...

I hope this is of some interest -- and maybe even some help -- to people on this forum!

[RDN]

.... Higbees "Thai Reference Grammar" is the way mai toe is always positioned significantly higher in all situations than "mai han aakaat".....

<{POST_SNAPBACK}>

Yes, the font he uses is excellent:

- all the tone marks are higher than the other symbols (gaa-ran, mai-dtai-kuu, mai han-aa-gaat) and the vowels.

I have the feeling that some of you may be interested in a paper called "How do Thais Tell Letters Apart?".....

<{POST_SNAPBACK}>

Absolutely! I downloaded and extracted the .ps file, but can't do much with it. I'll search the web for a converter tool and post what I can find. Maybe one of our PC experts knows of a converter.

I did download the abstract, and it looks just what I need. Here's a bit of it:

So Andrew, if I don't get anywhere with the conversion, I'll PM you my email address and hope you can send the PDF version. (Remember, you can send it just once to everyone who asks for it, and you can put all the email addresses in the "BCC" - blind carbon copy - field, and no-one will see the others' addresses).

[Andrew Mac]

One thing I like about Higbees "Thai Reference Grammer" is the way mai toe is always positioned significantly higher in all situations than "mai han aakaat".

Makes it a little easier.

<{POST_SNAPBACK}>

It's funny how we're all different, isn't it? That's one thing I really quite dislike about Higbie's otherwise excellent book! (Others are his system of phonetic transcription, which I personally find quite illogical and inconsistent, as well as the lack of an index in Thai script.)

In my case I honestly think that Thai script is much easier to read when there *isn't* that vertical gap between the consonant and the tone marker! I can't actually explain why, however ... that's just the way it is for me!

When it comes to distinguishing between may thoo and may han aakaat, I usually find that the difference in the horizontal position of the two characters relative to the consonant is enough to indicate which it is. Where the font designer, for whatever reason, has not offset may han aakaat a little to the right (which I think is standard practice), it's been my experience that s/he will usually make the visual distinction much clearer.

[Andrew Mac]

So Andrew, if I don't get anywhere with the conversion, I'll PM you my email address and hope you can send the PDF version. (Remember, you can send it just once to everyone who asks for it, and you can put all the email addresses in the "BCC" - blind carbon copy - field, and no-one will see the others' addresses).

<{POST_SNAPBACK}>

Yeah, thinking about it a bit more I realise that it shouldn't be too much of a problem after all. So please feel free to let me know if you (or anyone else) want me to send you a copy, OK?

[Richard W]

I have the feeling that some of you may be interested in a paper called "How do Thais Tell Letters Apart?". It was written by someone called Doug Cooper at the Center for Research in Computational Linguistics (which I found while surfing around one day!).

Yes! Very! Thanks!

A direct link to the document is http://seasrc.th.net/paper/tellthai.zip -- it's a 483K zipped postscript file, so you'll need to be able to convert that format into something more usable (such as PDF), or send it directly to a postscript printer. Converting it to PDF results in a 1.1 mb file, which I can't send to too many people ... but if a few of you are interested, and can't convert it, I might be able to send it to you. Alternatively, someone might have somewhere they can place it for easy download.

I've converted it to PDF using Adobe Acrobat 4.0 - the English text is a bit grotty (I think because font substitution on the Roman font), but it's about the same size (466K) when zipped up after conversion. I intend to put it up in my webspace tonight. I'll see if I can improve it, e.g. by replacing Helvetica by Arial. Perhaps the free tools (e.g. GhostScript) will do something better.

Doug's interest looks to be mainly Optical Character Recognition, but he has written quite a few other papers that might interest people.

<{POST_SNAPBACK}>

He's also interested in Thai etymology. I get the feeling he isn't getting much help.

[Richard W]

In many texts it's hard to distinguish between "mai toe"  ้  and  "mai han aakaat"  ั  unless you're totally familiar with the grammer rules which I'm not.

If the text is more than a line or so, you can identify mai tho because it's the only one of the two that goes above other vowels. Mai han aakaat is the only one that can seem to straddle two consonants on its own, but you can't rely on it doing that.

My wife admits that she writes them exactly the same!

One thing I like about Higbees "Thai Reference Grammer" is the way mai toe is always positioned significantly higher in all situations than "mai han aakaat".

Makes it a little easier.

<{POST_SNAPBACK}>

This is the typewriter style that modern font design strives to avoid! MacThai (and the 8-bit fonts, such as DB Thai) had different character codes for the different positions!

[RDN]

...I've converted it to PDF using Adobe Acrobat 4.0 ...  I intend to put it up in my webspace tonight. ...

<{POST_SNAPBACK}>

Hmm, maybe I should play around with my version of Acrobat a bit more, but I'll certainly be having a look at your web site later (http://www.thai-language.com/).

P.S. Congratulations on getting it back up and running, and finding the sound files.

[bina]

i'm glad u all had a good time discussing font sizes;

i just have to start wearing my bifocals when reading during the day (a pain in the a--- since i use prescription sunglasses and work outdoors so switching back and forth i;m just lazy lazy lazy so try to read w/o.....

but learned alot from this thread anyway....

plus i think that this newspaper uses small print to save copy since the paper is more or less free sort of....

[RDN]

This is the typewriter style that modern font design strives to avoid! ....

<{POST_SNAPBACK}>

You'll have to have a word with the UBC sub-titler:

[Richard W]

...I've converted it to PDF using Adobe Acrobat 4.0 ...  I intend to put it up in my webspace tonight. ...

<{POST_SNAPBACK}>

Hmm, maybe I should play around with my version of Acrobat a bit more, but I'll certainly be having a look at your web site later (http://www.thai-language.com/).

It's now up and accessible in PDF from my webspace. It's probably best accessed through my cover note. I've zipped it up (466kB) for the benefit of those with slow connections. The file looks fine when viewed with Adobe Acrobat 6.0 - I don't know why the letters were horribly misaligned when I was viewing it with Adobe Acrobat 5.0.

P.S. Congratulations on getting it back up and running, and finding the sound files. 

<{POST_SNAPBACK}>

Thai-language.com isn't mine - it's Glenn Slayden's! I'm just an occasional contributor.

[RDN]

...please feel free to let me know if you (or anyone else) want me to send you a copy, OK?

<{POST_SNAPBACK}>

Thank you Andrew! I just got your email and this document describes exactly what I was trying to explain - how to pick out the important bits in Thai characters to distinguish them:

...It's now up and accessible in PDF from my webspace.  It's probably best accessed through my cover note....

<{POST_SNAPBACK}>

Thanks Richard - I recommend everyone to get this and read it. It is soooo good!

....Thai-language.com isn't mine - it's Glenn Slayden's!  I'm just an occasional contributor.

<{POST_SNAPBACK}>

I often wondered why the author of the site was listed as Glenn Slayden, but it was temporarily hosted on your NTL World homepage. Anyhow, I'm glad it's back!

[The_Moog]

Sorry - having problems accessing that zipped file.

Is the text available elsewhere ? Can the doc be e mailed if I send someone an e mail address?

Thanks

[RDN]

Sorry - having problems accessing that zipped file....

<{POST_SNAPBACK}>

Could you have a look at it Richard please, because I get a "missing 10 bytes in zip file" error from WinZip.

Moog: I've got a zipped version (of Andrew's PDF), if you PM me an email address

[Richard W]

Sorry - having problems accessing that zipped file....

<{POST_SNAPBACK}>

Could you have a look at it Richard please, because I get a "missing 10 bytes in zip file" error from WinZip.

Moog: I've got a zipped version (of Andrew's PDF), if you PM me an email address

<{POST_SNAPBACK}>

I'll fix it when I get home this afternoon and let you know what happened. That will be about 17.00 GMT, so midnight in Thailand, I'm afraid. I've just checked it from a Windows 2000 machine at work, and I too get the 10 bytes missing report. They must have gone missing when I uploaded the file.

[Richard W]

....Thai-language.com isn't mine - it's Glenn Slayden's!  I'm just an occasional contributor.

<{POST_SNAPBACK}>

I often wondered why the author of the site was listed as Glenn Slayden, but it was temporarily hosted on your NTL World homepage. Anyhow, I'm glad it's back!

<{POST_SNAPBACK}>

I was trying to make it available until such time as Glenn could retrieve it from the FBI (or whatever). Most of what I had I had tediously gleaned from Google archives. I never got round to compressing all the data so it could fit in my webspace. I hope some of those who are ripping the site at the moment would have such noble intentions should it vanish again. However, Glenn now has backup plans to keep it going in case he should disappear from the scene.

[RDN]

.... I too get the 10 bytes missing report.  They must have gone missing when I uploaded the file. 

<{POST_SNAPBACK}>

Ah well, these things happen. The main thing is, it'll be there soon for everyone to look at. And it certainly is useful!

...Glenn now has backup plans to keep it going in case he should disappear from the scene.

<{POST_SNAPBACK}>

Good man!

[Richard W]

.... I too get the 10 bytes missing report.  They must have gone missing when I uploaded the file. 

<{POST_SNAPBACK}>

Ah well, these things happen. The main thing is, it'll be there soon for everyone to look at. And it certainly is useful!

<{POST_SNAPBACK}>

It's there now. This time I've done a test download before announcing it's presence. Last time I forgot to check that it wasn't FTPed in ASCII mode, whence the missing 10 bytes (Windows client to Unix host).

[bina]

thanx rdn,

the thai reads well, and part of the english fonts in the commentary, but the main body looks shadowed.... what to do???

oops just went back and checked, my acrobat is 5

maybe i have to update first?? to six or seven?

[RDN]

thanx rdn,

the thai reads  well, and part of the english fonts in the commentary, but the main body looks shadowed.... what to do???

oops just went back and checked,  my acrobat is 5

maybe i have to update first?? to six or seven?

<{POST_SNAPBACK}>

(I emailed bina my zipped PDF from Andrew Mac)

Yes, my Acrobat reader is v7. But if you look here:

http://www.thaivisa.com/forum/index.php?sh...ndpost&p=304567 there's a free PDF reader that I now use, that's quicker than Adobe (and smaller).

Edit:

I've just tried both Andrew's PDF and Richards new zipped PDF, and they look terrible using Acrobat 5.

But both look fine using v7 and the Foxit PDF reader I got from the link above.

So bina, you do need to upgrade

[easyb]

Maybe my eyes are just getting worse by the day and it's time for new glasses but I find the Thai in many books very hard to read. I've been meaning to post and see if there is a dictionary that has bigger Thai script type than Becker. I can’t read that.

I am very worried I’m going to be soon walking around Thailand with a one of these

[RDN]

Maybe my eyes are just getting worse by the day and it's time for new glasses but I find the Thai in many books very hard to read. I've been meaning to post and see if there is a dictionary that has bigger Thai script type than Becker.  I can’t read that.

I am very worried I’m going to be soon walking around Thailand with a one of these

<{POST_SNAPBACK}>

20 dollars! Wow, there are lots of big magnifying glasses on sale in Thailand - a lot in stalls on Sukhumvit road. But you'd have to jerry-rig the rest against your chest!

[manutoo]

...It's now up and accessible in PDF from my webspace.  It's probably best accessed through my cover note....

<{POST_SNAPBACK}>

Thanks Richard - I recommend everyone to get this and read it. It is soooo good!

yup, thanks to Richard.

It's quite an interesting book. Since I learnt to read Thai classic style, I always wondered about how was created the modern one and how Thai people learnt to read it..! Moreover, knowing the real differences to distinguish characters will be probably greatly improve my common script reading skills (I can read it, but something like 5 times more slowly than thai people )

And so far I was totally unable to read most thai hand writting, but now, my hope is back about this !

[RDN]

... Since I learnt to read Thai classic style, I always wondered about how was created the modern one and how Thai people learnt to read it....

<{POST_SNAPBACK}>

Hopefully it will enable you to understand why this stuff is not "USA", but the Thai equivalent of "Breeze":

[Richard W]

The paper didn't really address handwriting, or, to any great extent, script-like typefaces. I remember three features that had me baffled:

• ข can look like a gamma sitting on a line.
  
• ห can look like a curly capital U - the bottom left loops round rather than doubling back, and the knot at the top right hand corner can have a large vertical extent.
  
• า sometimes join on to a preceding consonant. The script combination ภา at the start of the title of the magazine ภาพยนตร์บันเทิง threw me - I had to ask my wife what the magazine was called! I couldn't find a better on-line example - the บันเทิง part is inside the circle to the right, and in a 'modern' font.