Character "etymology" dictionary for Pleco, plus other stuff

[werewitt]

Last year, after completely losing hope that the Outlier Dictionary will ever catch up with characters I already know, I looked up other "etymology" dictionaries. Why lost hope? Well, they have only just hit the 2000 definitions mark, after 3-5 years of selling it under grand advertisements, and many definitions are simply stubbed with the split key + phonetic (the last part is from when I looked at it half a year ago, perhaps they filled them out since).

 

I found one particularly intriguing "dictionary", or rather, a list in a PDF, by Lawrence Howell and Hikaru Morimoto ("late" Hikaru Moritomo, according to Mair). I bet the Outlier guys will say it is totally "unscientific" - they are heavy on "science", for some definition of "science", although I think there cannot be any science in arts, only tradition and recognition and maaaybe vague corroboration by primary sources. Aka 甲骨文 scratches on bones . Also an opinion by Mair http://languagelog.ldc.upenn.edu/nll/?p=3699

 

Yet I find that list useful for a fresh point of view on a character, and maybe even a mnemonic for tough cases. I took two PDFs from https://www.slideshare.net/KanjiNetworks and converted them to Pleco user dictionaries, including some rare characters that aren't well displayed in the PDFs themselves. No pinyin, traditional only, ~6200 definitions. "Mandarin redaction" lists "meanings" of characters in modern Mandarin and attempts to link them, often in a roundabout way, to the graphs. The original ("non-mandarin") uses more classical meanings, I recall.

 

I've put them on github, in case anyone wants to use them - https://github.com/agelastic/ChineseStuff/tree/master/Howell etymological dictionary 

 

There is also some more dicts/lists/cards in that repo https://github.com/agelastic/ChineseStuff

[werewitt]

Forgot to add - I sorta like Howell's idea, interpreted in the sense that, with the zillion homonyms in (even classical) Chinese, it totally makes sense that even the so-called "phonetic" 部首 was not chosen at random but from many existing similar-sounding signs that were somewhat related to the meaning of a new character. After all, there are several times more common "phonetics" used in traditional characters than there are different syllables in Chinese. 

[陳德聰]

Welcome back.

 

Right now, even though I think Outlier’s character list on Pleco is not big enough for my liking, it’s still a more accurate representation of actual character formation than anything currently available. I personally am not going to pay for it until it has more of the characters I would be likely to take interest in.

 

However, Mair has already said basically all that needs to be said in rebuttal of that interpretation; if it helps as a mnemonic then all the power to you. I feel like projecting new ways of classifying sounds onto the language can be helpful if they are accurate, regardless of whether they actually contributed to word formation.

[OneEye]

Well, much of Howell's stuff is fairly controversial and doesn't seem to be based on much evidence. I'm not actually sure we say that our stuff is "scientific" but rather that it's based on the most up-to-date research in the fields of paleography, historical phonology, etc. But if we were to say it's scientific (and maybe we have used that wording in the past), that's what we mean.

 

We also mean that advances in one aspect of the field should be consistent with, and will sometimes even advance, work in another. So for example, when we began to suspect that the 冖 in 學 may have originally been 六 rather than 宀 (as is usually said; their forms in ancient scripts were very similar), we had a hunch that 六 would likely also be a sound component. And sure enough, 六 is a very suitable sound component for 學. Although liù and xué seem very far apart in Mandarin, they're a bit closer in Cantonese (luk6 and hok6), and very close in Old Chinese (*k.ruk and *m-kˤruk in the Baxter-Sagart reconstruction). The corroboration from both a paleographic and phonological standpoint is a pretty strong indication that it was indeed 六 and not 宀.

 

It's quite possible to take a rigorous scientific approach to a humanities discipline; it's just not often done very well. Reliance on tradition is one of the main impediments.

 

But I'm obviously the biased party in this conversation, so I'll just leave it at that, and drop a link to another article by Mair.  

[werewitt]

17 hours ago, OneEye said:

But I'm obviously the biased party in this conversation, so I'll just leave it at that, and drop a link to another article by Mair.  

Yup:

Quote

The creators of ODCC came to see me about five (maybe more) years ago... it contains full entries for roughly the 1,500 most common characters,

 

Turns out I was right about "over five years" it took you to get to 1500 chars (I thought it was 2000 already. I was a few years early in claiming the number, it seems  )

[OneEye]

Oh, hi Vitaly. I forgot this was your handle. 

 

47 minutes ago, werewitt said:

Turns out I was right about "over five years" it took you to get to 1500 chars (I thought it was 2000 already. I was a few years early in claiming the number, it seems  )

 

The correct figure is a bit less than two years, and just under 2000 characters.

 

Prof. Mair must have misremembered. I actually visited him in January 2015,  before our Kickstarter in May 2015.

 

There has been a delay, and we've been pretty open about the reasons for it. You can check our Kickstarter page for more information if you'd like.

 

We released the initial version of the product February 2017, and it had only ~300 entries and a bunch of stubs. The stubs, btw, were requested by our Kickstarter users. We’re now at nearly 2000 completed entries, and we’re getting faster and more efficient at adding new entries. Doing quality work is more important than speed.

[furiop]

On 11/27/2018 at 5:09 AM, OneEye said:

We’re now at nearly 2000 completed entries, and we’re getting faster and more efficient at adding new entries. Doing quality work is more important than speed.

 

As a student I'm very satisfied about Outlier Dictionary, cause I don't need a mnemonic incorrect etymology, I'm very interested in state-of-art etymological study.

 

In my opinion to achieve the "complete" goal is like to win a Bingo: also in Bingo, it's relatively easy fill the first numbers, but it is complicated to fill all the numbers: the scientific work become more and more complicated, because there are different opinions about the single character etymology.  Also in my mother tongue the etymological dictionaries are sometimes in doubt: in chinese, I think, the doubts growt.

 

If I'm not wrong, in https://www.facebook.com/outlierlinguistics/ (October, 12) there is his post: "We've released an update to the Outlier Dictionary of Chinese Characters! It now has complete entries for the 1500 most frequently used characters + about 300 semantic components (with some overlap), for a total of over 1750 entries". In https://www.outlier-linguistics.com/purchase is written "Another 2000 characters (4000 total!)", for Commponent breakdowns + In-depth explanations for semantic components.

 

Outlier dictionary is a progressive work: long life and prosperity to authors, for a long but constant progression in etymology work.

 

Furio

 

 

[werewitt]

On 11/27/2018 at 3:09 PM, OneEye said:

Oh, hi Vitaly. I forgot this was your handle. 

No worries, John. It's one of, yes. In fact, my second favourite.

 

[Silvio Verbeek]

Hi Vitaly,

thank you for uploading Howell's dictionary files onto https://github.com/agelastic/ChineseStuff/tree/master/Howell%20etymological%20dictionary

I have always wanted to use Howell's mnemonics in Pleco, but didn't know how.

I am not too computer savvy, could you please explain HOW I upload them to the Pleco app on my smartpone (android)? And which one(s)? There are 6 :)
Thank you again.

Best from Europe,
Silvio
 

[LawrenceHowell]

@werewitt/Vitaly: Glad to hear you've found the lists of use. Just to clarify, the original was written to account for the meanings conveyed by the characters in Japanese.

 

@OneEye/John: Right. In the field of Chinese lexicography, there's only one true outlier, eh? As it happens, my stuff stopped being controversial way back in 2007, when Axel Schuessler came out with The ABC Etymological Dictionary of Old Chinese. There, as you know, he affirmed the presence of phonesthemic tendencies in Old Chinese, which serve as the foundation of the work (my late research colleague and) I did. To give him his mainstream scholarly due, Schuessler carefully avoided kneejerk-inducing terminology such as "phonosemantics" or "sound symbolism." In addition, he attempted to downplay genetic relatedness in what he was describing, but for all that he let the phonesthemic (orphonaesthetic) cat out of the bag, and the kitty won't ever be going back. As for the apparent lack of evidence to which you allude, one might as well ask Schuessler how he arrived at his own phonesthemic conclusions, a point he did not elucidate. My guess would be Schuessler judged, as I did for my own work, that the preponderence of evidence sufficed for the cause at hand.

 

@陳德聰: Rather than a rebuttal, Mair's post is better regarded as an attempt to stem the phonesthemic tide while incidentally offering up yet another paean to John DeFrancis. It appears the post is serving the intended braking effect, but (partially thanks to Schuessler) Mair is on the wrong side of what will become scholarly consensus in the coming decades. For a treatment of the most dubious points in Mair's post, see Ideographic Myth: A Response to Victor Mair. FWIW

[roddy]

Good to have you posting, Lawrence. 

[LawrenceHowell]

@Roddy: Thank you for the welcome, kind sir.

[LawrenceHowell]

I've posted elsewhere in this forum with the purpose of soliciting opinions about Han/Chinese character lexicography. I'd particularly like to encourage those who contributed here prior to my appearance to expand upon their viewpoints in the new post. Cheers.