Here's an updated spreadsheet that I've made based on the post Learning RTH (Simplified) after RTK by David556, with many things I was able to improve — the main complaints were that there were too many errors and that it was not in a spreadsheet format, and I think those are not a problem anymore (I hope). The main goal of this work is to find the commonalities between RSH (Remembering Simplified Hanzi) and RTK (Remembering the Kanji) to make the "transition" smoother, or to satisfy our curiosities, or something like that, and then improve it from there as much as possible. I also paired them with the, what... RTH (Traditional) keywords!

==================================================
2012-07-04: Update
I have created a new spreadsheet now with all the characters from RSH1&2, RTH1&2, and RTK1&3. Check the link for more information.
==================================================

Additional information:

• I checked an corrected all the RSH frames and keyword pairs based on the 1st. edition of Remembering Simplified Hanzi and the Errata;

• Three keywords were modified according to the errata:
- 匕 was 'spoon', became 'ancient spoon'
- 坠 was 'crash', became 'plunge'
- 统 was 'overall', became 'continuum'

• When informed in the book, I added the Part of Speech information — I believe they are helpful because by using them you are able to know that, e.g., 飞 (fly) is not a fly (the insect) but the verb;

And that's it for now. If you find any errors in the spreadsheet, please post here and I'll see what I can do.

Thanks for posting this! Actually, about 75% of David556's "has no RTK equivalent" hanzi do have RTK equivalents. To take just the first few:

18 叶 -> 葉
56 儿 -> 児
57 几 -> 幾
122 厅 -> 庁
164 压 -> 圧
171 灭 -> 滅
187 丽 -> 麗
189 响 -> 響
220 查 -> 査
224 桌 -> 卓
225 杂 -> 雑
228 艺 -> 芸
230 宽 -> 寛

Any chance of creating a copy of this spreadsheet on Google docs that everyone can help modify directly?

JimmySeal Wrote:Thanks for posting this! Actually, about 75% of David556's "has no RTK equivalent" hanzi do have RTK equivalents. To take just the first few:

Thanks, I'll add these when I have the opportunity.

JimmySeal Wrote:Any chance of creating a copy of this spreadsheet on Google docs that everyone can help modify directly?

Sure! I created an additional version that anyone can edit (so beware, heh):
Google docs file - testing version (anyone can edit)

I'll also keep the link to the version that only I can edit (that I'll call stable), where I'll insert only verified updates. I've added both links to the original post.

Major update! Changelog:

* Added all the RTK frames (up to 3007! I used the shared Anki deck for this) to make it easier for kanji/hanzi and keyword comparisons (if/then/else) and future updates;
* Reordered the columns, putting RTK first, followed by RSH, and later possibly RTH (traditional hanzi);
* Made a few keyword updates (around 10);

The column 'Kanji and hanzi are...' is generated automatically and doesn't consider font differences, as you can see in the screenshot (for example the kanji for RTK 73. straightaway [Mincho] / hanzi for RSH 72. straight [SimSun]).

Later I'll add the changes that JimmySeal pointed (I'm such a procrastinator...)

Obs.: This is the spreadsheet that I'm working on (the "stable" version), linked above – the testing one I'll not update by myself for now, I'll only do it after incorporating the changes you've added – there is someone modifying it right now!

Thanks for adding an editable version. I modified your editable version to add RTK kanji versions for all the ones I knew, and added RTK numbers and keywords for the first 20 or so. Please merge these in when you get the chance.

And FYI, you might realize this already, but you can use a nested IF expression to handle the 3 way logic of same-different-no equivalent, like:

=IF(C1251=G1251,ʺthe sameʺ,IF(G1251=ʺʺ,ʺno equivalent?ʺ,ʺdifferentʺ))

That way the "Keywords are..." and "Kanji and hanzi are..." columns can take care of themselves.

Ok, I've already merged quite a few (I'm doing this locally, so the newly updated version is not up on Google Docs yet); it's going to take a little while to finish them all. I'm going to bed right now, but don't worry, I'll continue it first thing when I get up! Then I'll see what nice things I can do with the nested IF and, after that, I'll publish the updated file as an editable one.

Cool beans. I'll hold off on making any more changes until you've finished merging what we have so far. Then you can replace the current editable version with the content of your locked version so things are in better sync.

Now that you've got separate "Keywords are..." and "Kanji and hanzi are..." columns, it might make sense to fill in kanji that match the corresponding hanzi even if they're not Heisig kanji. Would you say so, or do you want to keep this spreadsheet strictly Heisig?

Right JimmySeal, everything is up-to-date now! I've added your ~166 updates, and I found a few things I need to ask you:

RTK # Kanji RTK Keyword RTK Stroke Count RSH # Hanzi RSH Keyword
a. 51 卓 eminent 8 48 卓 eminent
b. 51 卓 eminent 8 224 桌 table
a. is the original entry, b. is your update; I think here we should keep the original entry, right? Or am I missing something?

a. 2173 叶 within my ability 5
b. 228 葉 leaf 12 18 叶 leaf
- a. has 叶 (kanji = hanzi), b. has leaf (RTK = RSH, but different characters). What should we do here?

Where are these RTK # higher than 3007 from? (I'd like to add them to the spreadsheet)
3017 拉 yank
3025 辣 bitter

Nested IFs
I've added these rather complex nested IFs to make it easier to catch commonalities just by sorting by these columns (that is, if there are no errors in the algorithm*):

For the "Keywords are..." column:
=IF(C2=G2;"the same";IF(C2="";(IF(G2="";"RTK & RTK missing";"no eq., RTK missing"));IF(G2="";"no eq., RSH missing";"different")))

The nested IF is used to check if column C (RTK Keyword) is equal to G (RSH Keyword). Explanation:
1. If C and G are the same, no problem: tag "the same";
2. If C is empty, check if G is empty:
2.1. If G is empty too, tag "RTK & RTK missing" and, if not, tag "RTK missing" (issue: this condition never occurs since C=G prevents it)
3.If C is not empty, check if G is empty:
3.1. If G is empty, tag "RSH missing" and, if not, tag "different".

* Small problem I couldn't figure out: the 2.1 branch, when both C and G are empty, is never reached, since C2=G2 prevents it. I wanted it to tag "RTK & RTK missing", but I couldn't find a way to do it. This was taking too much time for me to get it right, so for now I'm used the code above. This part is actually not essential at the moment (there are no both C2 and G2 empty in the current spreadsheet), but can be useful for when adding new characters, according to JimmySeal's suggestion below.

For the "Kanji and hanzi are..." column:
=IF(B2=F2;"the same";IF(B2="";(IF(F2="";"RTK & RTK missing";"no eq., RTK missing"));IF(F2="";"no eq., RSH missing";"different")))
Where column B contains Kanji and F, Hanzi.

JimmySeal Wrote:Now that you've got separate "Keywords are..." and "Kanji and hanzi are..." columns, it might make sense to fill in kanji that match the corresponding hanzi even if they're not Heisig kanji. Would you say so, or do you want to keep this spreadsheet strictly Heisig?

I think adding these would be great! Would you do it?

After this we would only need keywords for additional characters*, and maybe we could order them with the same technique Heisig uses (adding new characters only after [or when] all the needed primitives were already introduced). The exact numbering is not important - we could precede it by C (Custom), like this:
C1 [hanzi/kanji]
C2 [hanzi/kanji]
...
C200 [ hanzi/kanji]

Where C[num] is not fixed, so we can insert a new character/keyword as we identify them and generate the numbering accordingly (it could have a frequency id., so we don't get bogged down by too many additional characters when studying in order, just when using it for reference).

This could be a kind of "beyond RTK/RSH" — so we don't depend on Heisig for characters that he didn't cover in his books —, although kind of hard and laborious to do, better suited for people who already know a fair amount of the target language, I think (in order to choose the right keywords). Anyway, the idea is here.

* Update: Actually I've mixed some ideas here — I was thinking, " how am I going to study these?", as you can see by the title. Yes, thinking long-term.

gdaxeman Wrote:RTK # Kanji RTK Keyword RTK Stroke Count RSH # Hanzi RSH Keyword
a. 51 卓 eminent 8 48 卓 eminent
b. 51 卓 eminent 8 224 桌 table
a. is the original entry, b. is your update; I think here we should keep the original entry, right? Or am I missing something?

This looks like a case where two kanji have been merged into one and they have remained as two separate characters in simplified Chinese, as 卓 definitely means "eminent" and "table" in Japanese and 桌 doesn't exist in JIS 2. I think we should keep it this way and note "c.f RTK 224", "c.f. RTK" in the Additional Notes section.

Quote:a. 2173 叶 within my ability 5
b. 228 葉 leaf 12 18 叶 leaf
- a. has 叶 (kanji = hanzi), b. has leaf (RTK = RSH, but different characters). What should we do here?

Well, "within my ability" is a Japanese-invented meaning for 叶, and 叶 is the simplified form of 葉. So it probably makes most sense to leave it this way, and again include cross-references in Additional Notes.

Quote:Where are these RTK # higher than 3007 from? (I'd like to add them to the spreadsheet)
3017 拉 yank
3025 辣 bitter

These are Kanji from the RTK 1 Supplement. I'm basing the numbers off of the numbers in this site's Study section, but I don't know where they came from.

gdaxeman Wrote:For the "Keywords are..." column:
=IF(C2=G2;"the same";IF(C2="";(IF(G2="";"RTK & RTK missing";"no eq., RTK missing"));IF(G2="";"no eq., RSH missing";"different")))

You need to invert the nesting order:
=IF(C2="";IF(G2="";"RTK & RTH missing";"no eq., RTK missing");IF(G2="";"no eq., RTH missing";IF(C2=G2;"the same"; "different")))

JimmySeal Wrote:I'm basing the numbers off of the numbers in this site's Study section, but I don't know where they came from.

Ah, it makes sense to use the same numbers, since one really useful thing this spreadsheet can provide is the ability to rapidly access the stories that are available here for the kanji that can also be used for hanzi.

JimmySeal Wrote:You need to invert the nesting order:
=IF(C2="";IF(G2="";"RTK & RTH missing";"no eq., RTK missing");IF(G2="";"no eq., RTH missing";IF(C2=G2;"the same"; "different")))

Thanks, it works now! I just changed the text from "RTK & RTH missing" to "RTK & RSH missing" (I frequently mix the initialisms for RSH and RTH, it makes me think it's "Remembering the Hanzi"). The new code will be on the next version, which I'll work on after you finish the updates, to avoid change conflicts and duplicated efforts.

For now I'm going to turn my attention to other things: adding the number of strokes for RSH and see what I can do with RTH (Traditional Hanzi) — I'm adding a separate tab where they'll be compared with the Simplified Hanzi, just like the one we're doing for RTK. The shared Anki deck for RTH takes care of most of what I intend to do (it even has one reading for every character, as seen in Heisig's index). According to my rough observation using the same IF I used for RTK, there are 971 traditional characters that are equal to the simplified ones, and 529 that are different (this is not too accurate, since there are characters that look the same but the IF says they're different because they have different code points, and also there are characters that it says are equal, but they look slightly different when using the right fonts).

I've made some changes, but I don't plan to be making any more for the next day or so, so please feel free to do another merge.

For now the only change I've made to the Google Docs files was to add your formulas to the aforementioned columns; the addition of the Traditional Hanzi information which I'm focusing on will be a major one and it is going to take me a while to finish, maybe a few days. After this (well, not right after, but follow me) we'll be able to confidently answer the question — or personally check any answer others may give, even Heisig himself: "does RTK have more in common with RSH (Simplified) or with RTH (Traditional Hanzi)?" This question has been previously asked a few times on this forum and some others that I've accessed but, for what I've read, they were all guesses (some saying that it was S, others T). Also, we'll be able to do a Ctrl+F that searches through all the characters and keywords, not only one set or the other.

Notes for all of you guys and girls:
Note 1: I generate the formulas on the "Keywords are..." and "Kanji and hanzi are..." columns automatically, so if something was changed manually in either of them, it has gone. It's better to put these kind of alterations in the Additional Info column so we can see them, otherwise we lose them in the middle of the automatically-generated values.

Note 2: Don't trust too much the fonts Google Docs use – they don't offer specific choices for kanji, simplified hanzi and traditional hanzi (and it replaces the ones I'm using locally), so the characters might look a little odd / different from expected, as discussed on this thread.

Ok, I've finished doing the Simplified and Traditional Hanzi pairing / list of equivalences according to the Heisig and Richardson's books. I couldn't think of a sensible way to merge them with the kanji one without making it very complex to upkeep and without too much horizontal scrolling, so for now this info is in a separate tab named RSH+RTH in the original file, like the Primitives tab. I have also corrected some errors, and the Google Docs are now updated. The next step would be to make a comparison of all these hanzi with the kanji, then indicate if the kanji is equal, or more similar, to the simplified or to the traditional hanzi.

Note: In the "Kanji component that was simplified" column, when I say "font variation" I mean that the characters look significantly different if you select specific fonts for Traditional and Simplified Hanzi — to the point of inverting a stroke direction and making it look like a different primitive, for example. The tests were made visually with Microsoft YaHei and JhengHei, so they are not definitive, and other fonts (e.g., SimSun and MingLiU, or any other combination of your choice) may or may not be affected depending on the character. Notice, though, that on Google Docs they look the same.

gdaxeman,

Your hard work is greatly appreciated!

Thank you.

Time for a new, massive update — this time I've merged Kanji, Traditional Hanzi and Simplfied Hanzi all into the same spreadsheet! It's probably one of the most comprehensive Heisig's comparison lists you'll find for the time being. I've updated the screenshot on the first post, so look there if you want to know what I'm talking about. I'm also outlining a kind of 'framework' that will make it easier to add and share keywords for characters that Heisig didn't cover, but that's a future project.

I'm using many abbreviations in order to make the columns narrower; these are listed on the Abbreviations tab, but are easily inferred (e.g: TH is for Traditional Hanzi). Sometimes I also invert words an write not-so-proper English so that the most important information can be easily seen.

This list is an improvement with many errors corrected since the last one, and it has changed its goals: it's not "what you'll need to learn if you've previously completed RTK and now want to learn SH" anymore, but a "I want to know what are the differences between all the books and character sets!" (of course the former question can also be answered, and in an even better way).

There's now a link for the original Excel spreadsheet I'm working on, using the following fonts:
* Meiryo for Kanji
* Microsoft YaHei for Simplified Hanzi
* Microsoft JhengHei for Traditional Hanzi
* Arial for everything else

Original Spreadsheet - Excel 2007 file

I've added this because Google Docs is very simple and changes all the fonts I'm using, so from there it's not possible to see the differences between characters when using different fonts that I'm talking about.

And that's it for now! Now I'm gonna stop procrastinating and I'll just start learning hanzi! The Anki deck will be an interesting thing to do...

@gdaxeman
Wow, this is exactly what I needed, thank you very much! For the last five days I have been working on an Anki deck for both traditional and simplified versions of RtK kanji. I have been copying and pasting RtK kanji from this site into Wiktionary to find out if the hanzi have any differences from the Japanese versions, and then I would come up with a story and make a card. This is a very slow process, and I am also studying for my college exams right now (one down, three to go), so I have only managed to check 178 kanji and make 41 hanzi cards so far. Your spreadsheet should speed up the process dramatically.
非常感謝!

Awesome, this has worked out wonderfully in using stories to link the characters to their reading. Thank you so much! Will this be updated when the second volume of Remembering the Hanzi comes out? If so, can I help in any way?

[ファブリス]

Thank you gdaxeman. I'm building the data for RevTH with the help of your spreadsheet.

My script output 1499 unique rows for RTH Vol.1. On further inspection RTH # 677 and 678 ("ring", "give back"); rows 837-838 on Google Spreadsheet; have the same traditional character.

EDIT: I have the following in RTH data that transalpin sent me. I think "give back" is the incorrect one:

ring 環
give back 還

LittleFish Wrote:Awesome, this has worked out wonderfully in using stories to link the characters to their reading. Thank you so much! Will this be updated when the second volume of Remembering the Hanzi comes out? If so, can I help in any way?

The spreadsheet can be easily updated (well, sort of) when the second book comes out, I just need a list with the characters and keywords. Then, it's just a matter of comparing the characters and keywords like it was done here. Maybe I'll do this — it depends on how long it will take for the book to be released. You can help by providing the list, comparing the characters, finding errors, etc.

There's one thing I should mention: a while ago I converted this spreadsheet to an Anki deck with extra fields (including RSH1, RTK1 and RTK3 primitives — it's a really complete deck but maybe slightly confusing, with 20 fields), and now I'm updating only this deck and studying the characters using it — that's why I haven't updated the spreadsheet anymore, it was hard/confusing to update both files. In this deck, I've corrected a few errors (aargh, there's always some error, no matter how careful you are) and already added more than 100 extra characters (keywords + primitives) that are not on RTH1/RSH1. I'm really enjoying the process of listing the primitives and selecting keywords by myself, so I probably won't use the next book (especially if it takes longer for it to be released than for me to add the extra 1500+ characters), which is why I'll have to rely on others providing the lists.

About this very complete and updated Anki deck I'm talking about: I was thinking of posting it here, but I don't know if it's allowed to do, especially because of the primitives fields. If someone says they want it and it's ok to share, then I'll do it.

ファブリス Wrote:Thank you gdaxeman. I'm building the data for RevTH with the help of your spreadsheet.

I'm glad this spreadsheet is being useful!

ファブリス Wrote:My script output 1499 unique rows for RTH Vol.1. On further inspection RTH # 677 and 678 ("ring", "give back"); rows 837-838 on Google Spreadsheet; have the same traditional character.

EDIT: I have the following in RTH data that transalpin sent me. I think "give back" is the incorrect one:
環
ring 環
give back 還

Hum, I hadn't noticed this error; yes, the correct traditional character is 還 for give back ('road' instead of 'king/ball'). There are, I think, three other small errors that I should report for the sake of precision, I just need to find them (I didn't keep track of the updates I made to the deck I mentioned, so I'll have to use another method; the errors are related to pairing the SH with the TH, so not that crucial for now I think.)

Also, if you're using the H. Reading field, I should say that the readings in the spreadsheet are from a shared Anki deck and I noticed some "imprecisions" where multiple readings are listed, at least if you're considering only Standard Mandarin (the characters where only a single reading is listed are okay.) I think it's best to stick with the readings listed in the books or in a reliable dictionary. I'll try to explain why with an example:

756 得 702 得
Readings listed before: děi; dé; de
Readings used now: děi

Why did I change it? Because the keyword Heisig uses is must, and when the hanzi is used in this context it's read děi; dé is for when the character means "get, reach, achieve" and de is for when it's used as a grammatical particle, according to the Wenlin dictionary.

There are other examples like this, and there are also some readings that I don't know where they got from — so let's stick with what Heisig / Collins / Wenlin say.

So, here are the single readings (except 爪 that has 2) for the fields where there were multiple ones (most of them are from Heisig's book, a few from a dictionary):

RSH | SH | RTH | TH | Readings
0884 呢 0822 呢 ní [ne is more common but it's not the pronunciation for woolen cloth]
0756 得 0702 得 děi
0046 下 0045 下 xià
0115 夕 0108 夕 xī
xxxx 采 0614 采 cǎi
0663 爪 0612 爪 zhuǎ, zhǎo
0653 叔 0604 叔 shū
0973 昔 0897 昔 xī
1338 期 1288 期 qī
0032 凸 0031 凸 tū
0129 削 0118 削 xiāo
0270 玩 0252 玩 wán
0353 识 0452 識 shí
0363 划 0313 劃 huá
0376 尧 0166 堯 Yáo
0384 企 0358 企 qǐ
0440 夭 0404 夭 yāo
0574 息 0538 息 xī
0596 担 1478 擔 dān
0597 括 0562 括 kuò
0598 提 0563 提 tí
0655 寂 0606 寂 jì
0660 质 0862 質 zhí
0760 微 0708 微 wēi
0764 和 0712 和 hé
0797 伪 1386 偽 wèi
0850 巫 1460 巫 wū
0929 暂 0860 暫 zàn
0948 档 0652 檔 dàng
0976 散 0900 散 sǎn
1055 骨 0978 骨 gǔ
1072 穴 0995 穴 xué
1073 究 0996 究 jiū
1074 突 0997 突 tū
1125 危 1057 危 wéi
1138 血 1075 血 xiě
1171 讽 0484 諷 fěng
1218 绩 1159 績 jī
1236 击 0636 擊 jī
1353 亚 1313 亞 yà
1420 播 1380 播 bō
xxxx 发 1391 髮 fà

[ファブリス]

OP Wrote:Also, if you're using the H. Reading field

I'm only using the RTH/RSH characters, keywords, and frame number. The perl script uses the unicode value as a key to look up readings and strokecounts from the UNIHAN database.

OP Wrote:Why did I change it? Because the keyword Heisig uses is must, and when the hanzi is used in this context it's read děi; dé is for when the character means "get, reach, achieve" and de is for when it's used as a grammatical particle, according to the Wenlin dictionary.

It is not going to be as specialized as your custom deck of course, in terms of hand picking the most relevant readings. I could use specially edited data such as that you describe, but I will eventually have to support all Chinese characters in the unicode range and need a consistent way to supply readings for all of them.

For comparative purposes, I could use the frequency from the Unihan_Readings.txt file, kHanyuPinlu field.

This one ( 血 ) gives a greater frequency to xue4:

U+8840 kHanyuPinlu xue4(349) xie3(69)

This one ( 企 ) has the same reading as you listed:

U+4F01 kHanyuPinlu qi3(244)
U+4F01 kHanyuPinyin 10116.060:qǐ

This one ( 和 ) has loads of readings, but the first reading has a much higher frequency than the others (the same that you picked):

U+548C kHanyuPinlu he2(9513) huo5(38) he5(24) he4(9)
U+548C kHanyuPinyin 10602.010:hè,hé,huò,huó,hú

This one ( 爪 ) has the readings the other way around, but the frequencies are very close:

U+722A kHanyuPinlu zhao3(18) zhua3(12)
U+722A kHanyuPinyin 32030.010:zhǎo,zhuǎ

Those frequency readings could be helpful to make a distinction between readings that are are more or less equally used, and cases where secondary readings are much less common.

Oh, so there's no problem. The thing about the readings I was talking about is just due to the fact that sometimes Heisig (and I) choose keywords that do not represent the main meaning and, because of that, sometimes don't use the most common reading. I actually don't know the value of aligning them this way instead of using the most common reading; I'm still experimenting. One thing I noticed is that having multiple pronunciation sounds during reviews is a little distracting for me, so I reduced them to only one sound per card (two at most.) Well, besides that, the great thing is that most hanzi have just one reading, which when vary are mostly based on very understandable rules (unlike Japanese), so this is just an "exception" thing.

You're going to support all the 25000+ characters? That's very interesting — maybe I'll add some of my stories for the very obscure, non-Heisig ones then.

[ファブリス]

OP Wrote:One thing I noticed is that having multiple pronunciation sounds during reviews is a little distracting for me, so I reduced them to only one sound per card (two at most.)

Oh I totally agree. Maybe I'll use just the first two readings.

Quote:You're going to support all the 25000+ characters?

Not in the free "plan" but as part of a subscription or membership of some sort, which I briefly mentioned here.

gdaxeman Wrote:Here's an updated spreadsheet that I've made based on the post Learning RTH (Simplified) after RTK [...]

Just a small note: you linked
澱 2300 sediment (RTK)
with (RSH) (RTH) 221 渣 sediment 208 渣 sediment

But the keywords are the same incidentally; the hanzi/kanji have no relation whatsoever. Instead, the TH is also 澱 (or 淀) and the SH is 淀.

By the by, I am interested in that very large Anki deck you mentioned.