How to learn long2gu3 (dragon bones), i.e., oracle bones

I began about 10 years ago by studying Wieger, Dr. L., S.J. (1927) Chinese Characters: Their Origin, Etymology, History, Classification and Signification. A Thorough Study from Chinese Documents, and Peng’s Fun with Chinese Characters series, but became dissatisfied with them, as they are not at all accurate; the former predates mature jiaguology, while the latter is nothing but a collection of folk etymology (although I did enjoy both). Another book which is very problematic in terms of its etymology is Harbaugh, Rick (1998). Chinese Characters: a Genealogy and Dictionary, which is useful only for finding characters by graphic components, and not at all reliable on etymology, as, like Wieger, it relies too heavily on Shuo1wen2 , and even its Shuowen (SW) scholarship is poor. So I strongly recommend against using any of these for learning etymology. TaiOanKok, you’re probably already aware of this, but I’m adding detail for other readers too. You can skim on down further for meatier stuff.

Later I found Wu2, Teresa L. (1990). The Origin and Dissemination of Chinese Characters, which is pretty well written, but limited in content and now out of print. It lays out a representative series of the OB, bronze, small seal, clerical and kaishu (modern) forms of a number of graphs, then goes into a reasonable analysis of each, based on the views of one or more scholars. It is much more reliable than most mass market books, but still not up to real academic standards, as it generally fails to present alternative theories, and the selection of only one OB and bronze form is often an unfortunate choice, as for many there were multiple forms.

A very similar work which is still available is by 謝光輝 Xie4 Guānghuī ed., (1997), The Composition of Common Chinese Characters: An Illustrated Account, Peking (sic) University Press. ISBN 7-301-03329-x. It’s a large softcover book with 652 pp., each showing the evolution of one character, with single, representative OB, bronze, and seal forms, accompanied by a brief paragraph of explanation in English and Mandarin (simplified characters), with illustrations and cartoons. Despite the cartoons, this is rather more accurate than the books by Peng, so it can therefore actually be recommended as an introduction to the topic for the casual reader. However, it still does not have the scholarly rigor to make it a suitable text for the beginning university student or serious amateur sinophile, and the single examples don’t give an accurate picture of the diversity of forms at any point in time. The layout is also tremendously wasteful of space, and it is only a partial translation of the larger, Chinese original. With only 651 characters (although more than most sources), its usefulness as a reference book is still very very limited, especially because many extremely basic characters are missing, such as 的, 是, 三, 五, 七, 切, 八, 巴, 把, 爸, 白, 百, 拜, 包, 九, 千 and so on, while a few uncommon (albeit etymologically interesting) ones such as


pān, 陟 zhi4 and 刖 yue4 are included. Most frustrating is the fact that, as with most currently available mass-market books, Xie does not adequately address competing etymological theories when appropriate. For each main character entry in the main lessons and the indices, traditional characters are given parenthetically. The book is organized logically into topical categories (man, utensils, architecture, etc.) and indexed by topic, pinyin head letter, and stroke number. Available at Schoenhof’s online; or in Australia, Worthwhile for the beginner.

In addition to this, I’d of course recommend THE fundamental introduction to jiaguology, which is Keightley, David N. (1978). Sources of Shang History: The Oracle-Bone Inscriptions of Bronze Age China. University of California Press, Berkeley. Dr. Keightley’s works are superb, scholarly yet readable sources of information in English on the oracle bones, their discovery, context and interpretation. This work is authoritative, technical, and comprehensive, covering all aspects of oracle bones, from their discovery, to early research, what reference sources to consider buying if studying the OB directly, and so on. The main text unfortunately clings to Wade Giles romanization without tone marks, and usually without Chinese characters; however, some appendices and the bibliography do better, adding traditional characters. It’s available in large format hardcover, ISBN 0-520-02969 (out of print, but widely available used on the internet); A cheaper but not cheap 1985 ppbk 2nd edition is still in print, ISBN 0520054555. There’s also a HC 2nd ed., out there used. I do strongly recommend the 2nd ed. over the 1st, as the info on references is more up to date.

Also worthwhile is Keightley, David N. (2000). The Ancestral Landscape: Time, Space, and Community in Late Shang China (ca. 1200 - 1045 B.C.). China Research Monograph 53, Institute of East Asian Studies, University of California - Berkeley. ISBN 1-55729-070-9, inexpensive ppbk. In this he goes through the contents of the OB inscriptions, gleaning what he can about all aspects of Shang life, which is quite interesting. Thankfully, he switches to pinyin, but still no tone marks. Unfortunately, he converts all the inscriptions into modern 楷 kai3 standard script versions, without providing the OB forms, so you learn about the Shāng world, but not how to read OB graphs themselves.

So the above two books will teach you a fair amount about what the OB sources and contents are, but will NOT systematically teach you how to read the actual graphs; nor are they suitable for looking up graphs by their modern form, to see their OB form or their entire evolutionary history, the way Wu and Xie above at least attempt to do.

After a couple years searching in frustration for good sources along those lines on Chinese etymology in English, I wrote to the中央研究院 Academia Sinica’s 歷史語言所 Institute of History and Philology asking the scholars there if they knew of any good resources, and if not, would they be interested in writing such a book – essentially, something to replace Wieger, and something like Xie but more comprehensive and a bit more rigorous in terms of scholarship, but still accessible (readable and affordable) to the foreign student of Chinese. One of the scholars, 陳昭容 Chen2 Zhāorong2, replied that she was unaware of any good materials like that, but she’d be willing to help me write one (ulp!). So I’m in the midst of a 10-year project to learn to read OB and bronze scripts, as well as to learn to read academic Chinese of the most impenetrable type, with the eventual goal of writing something like a dictionary, with around 1000-2000 characters, as well as a lesson-like book along the lines of Wieger, and perhaps an introductory textbook and workbook. I’m not sure how far I’ll get on these, but I do plan to have them reviewed by or written with the participation of professional scholars, so as to make them reliable and accurate.

So, the next step for anyone interested in learning more (before such works as those I plan are available) is to get one’s Chinese reading up to snuff so as to be able to turn to the most accessible sources in Chinese, starting with the simplest, reasonably reliable and inexpensive ones. I’d recommend you begin with these two:

趙誠 Zhao4 Cheng2 (Chao Ch’eng; 1988) 甲骨文簡明詞典 - 卜辭分類讀本 jia3gu3wen2 jian3ming2 ci2dian3 – bu3ci2 fēnlei4 du2bĕn. 中華書局 Zhōnghua2 Shūju2, ISBN 7-101-00254-4/H:22. Very inexpensive. I bought mine at; it’s a PRC bookstore, so you’ll have to pay by wire to be safe; I wouldn’t give them your credit card number. Also, they’re very slow so be patient (e.g., 3-5 days to respond).

劉興隆Li� Xīngl�ng (1997). 新編甲骨文字典 (new oracle bone dictionary), 文史者出版社, 台北. W�nshĭzhĕ (Wen-shih-che) Publishing, Taipei. ISBN 957-549-062-2. I got mine at 三民書局 on F�xīng N. Rd., might have been around 850NT.

These are both relatively small dictionaries. You check for characters by their modern forms (if applicable) or OB forms by stroke number in an index (sorry, no pīnyīn), then go to the relevant entry, which shows (in Li�) the OB, bronze and small seal forms, the kai3 form, the definitions in OB and bronze times, the Shuōw�n definition, and finally, examples of OB inscriptions (with both OB and kai3 graphs) in which the character is used. Meanings of other characters in the inscription, if relatively difficult, are footnoted. Li�, although not a professional scholar in this field, has done a fine job with this dictionary, which is really aimed at folks like us, serious amateurs. Zh�o is a very famous OB scholar, and this book of his is similarly targeting serious amateurs and roughly similar in content, although a tad more difficult to use, since his entry format varies � sometimes you have to read the whole entry just to find out what kai graph corresponds to the OB entry, whereas Liu puts this and other info up front in a more predictable, systematic format.

After reading some of both, you�ll probably be interested in having a crack at some real OB pieces to see how much you can read so far. If you have access to an Asian Studies library, check their catalog for any collection of published rubbings of OB, and photocopy a few dozen pages, then start reading them. There are many, but I�ll name a handful of famous ones:

H�j� is the nickname for 甲骨文合集 Jia3gu3w�n H�j�, 郭沫若主編 Gūo M�ru� ed., 胡厚宣總編輯 H� H�uxuān ed.-in-chief, 1978-1982, 13 vols. 中國社會科學院–歷史研究所, 北京 Bĕijīng: 中華書局 Zhōngh�a shūj�. A revised unauthorized reprint is also available by Yen Yi-p�ing, in Taipei. Huge and expensive, you won�t want to buy this, but it is THE comprehensive source for published OB rubbings. This monumental collection of 41,956 oracle bone rubbings in thirteen oversized folio volumes contains the vast majority of all known and published pieces of value, and virtually all those significant pieces unearthed at Anyang between 1899 and 1973. By �of value� and �significant�, it is meant that small fragments with indiscernible inscriptions and poor quality pieces which merely repeat common and uninformative graphs are not included; these account for about three quarters of the over 200,000 pieces originally considered.
Although many are also available in other, earlier publications, at least 4,200 are new, and many, perhaps 2000, have been newly rejoined. It is also very useful for scholars to be able to cite particular OB pieces according to one set of standard reference numbers based on the H�j� collection. The first twelve volumes contain the rubbings, divided by period, according to the most commonly accepted system of periodization, arrived at by 董作賓 Dong3 Zu�bīn (Tung Tso-pin) in 1935. Six of the H�j� volumes are devoted to the largest corpus, which are identified as belonging to period I; period II rubbings fill one volume, period III 1.8 volumes, period IV 1.2, and period V one volume. The thirteenth volume contains hand-drawings. Additional volumes are planned for indices, photographs, and pieces not yet included, such as those more recently excavated Zhoūyu�n pieces from 1977 .
Within period, pieces are classified by general topic, but this classification scheme has been designed and carried out in a very unsatisfactory manner; not only are some categories rather dubious, but the assignment of pieces to them is often even more perplexing. And pieces often have inscriptions pertaining to various topics, and even within a single inscription, there are often multiple ways to classify it, but the editors have only chosen one of these for each piece, so this is not a very useful way to find pieces, nor is it useful for statistical analysis by topic. Instead, one must consult one of two other works which focus on classification by content:《類纂》 L�izuan3 (short for《殷墟甲骨刻辭類纂》yīnxū jiagu k�c� l�izuan3 by 姚孝遂 Y�o Xi�osu� and 肖丁Xi�o Dīng), and 《綜類》 Sorui (short for《殷墟卜辭 綜類》Inkyo bokuji sorui by 島邦男 Shima, Kunio).
The quality of H�j�s editing and reproduction is generally very good, with only a limited numer of duplicates and other errors. Although a must-have resource, it is also prohibitively expensive, and students should check the nearest university�s Asian Studies or main library for a set.
One of the problems for students when looking at the rubbings directly is that the inscriptions are all helter-skelter on the bones, turning this way and that, running into other inscriptions, omitting words, and ending unexpectedly. Manual transcriptions are thus often easier to read than the rubbings, and are made even easier by accompanying 楷定 kai3d�ng translations (i.e., conversion to modern graphs, where possible). The first to be published with OB transcriptions and kai3d�ng translations for the H�j� compendium was《摩釋》M�sh� (short for《殷墟甲骨刻辭摩釋總集》yīnxū jiagu k�c� m�sh� zong3j�, by 姚孝遂 Y�o Xi�osu� and 肖丁 Xi�o Dīng). 1988; two volumes. 北京 Bĕijīng: 中華書局 Zhōngh�a Shūj�. This relatively affordable set contains manual transcriptions of the oracle bone rubbings in the《合集》H�j� compendium (q.v.), followed by 楷定 kaid�ng translations (aka 隸定 l�d�ng, i.e., conversion to modern graphs, where possible). It follows the order and index numbers of H�j� and is thus a convenient reference, but has quite a few errors and omissions, so it�s not authoritative. Being able to compare the OB and kai graphs together is quite helpful, but you have to supplement this with the above dictionaries (Zhao and Liu) to get anywhere. It follows the order and index numbers of H�j� and is thus a very convenient reference, but has quite a few errors and omissions, and a better version using the same order is now available by the H�j� team itself (called《釋文》Sh�w�n, short for《甲骨文合集釋文》Jiaguw�n H�j� Sh�w�n, 胡厚宣 H� H�uxuān ed.). Not only is their scholarship superior, but they�ve also had years of experience working with the rubbings themselves, rather than reproductions thereof, which also helps reduce errors. On the downside, the 釋文 Sh�w�n contains only the kaishū transcriptions, rather than pairing them with handcopied versions of the OB graphs; I find the pairing convenient, myself, and I think most students would agree. Thus, I recommend 摩釋 M�sh� over 釋文 Sh�w�n, at first.

Serious students who wish to own some reproductions of oracle bone rubbings or those without access to a library set of H�j� will want to consider more affordable compilations such as 丙編Bĭng Biān, which is the standard abbreviation for 殷墟文字: 丙編, aka中國考古報告集之二, 小屯, 第二本, Archaeologica Sinica Number Two: Hsiao-T�un (sic) (The Yin-Shang Site at Āny�ng, Honan), Volume II: Inscriptions; Fascicle 3: Inksqueezes of the Restored Specimens of Inscribed Tortoise Shells with Annotations. This six-volume set is a steal at $3000 NTD (US$93) per volume, because of the expense of production: atop every high-quality photographic plate reproducing the rubbings is a tissue overlay page on which are transcribed the kai versions of each graph, along with enumeration of the inscriptions, and arrows showing the general direction (such as down and to the right) in which to begin reading that particular inscription. Extremely helpful, and not cheap to produce. Plus in the rear are additional copies of these transcriptions, along with other analyses. If you don�t have access to any other sources of rubbings, picking up one of these volumes would be quite worthwhile, and then you can decide whether you want to order the remainder of the set, as they are available separately, from the Academia Sinica, Taipei.

You can also look up particular OB graphs and compare the interpretations of various scholars at a glance in Soran (or in Mandarin, Zōnglan ) which stands for 甲骨文字字釋綜覽 by松丸道雄Matsumaru and 高嶋謙一 Takashima. This is a concise listing of various kaishū renditions of individual oracle bone graphs by leading scholars, allowing one to conveniently refer to the various possible readings at a glance, as well as to get a quick feeling for the level of consensus on a particular graph. Scholars who can read difficult, scholarly x�ngshū semicursive Chinese at skimming speed may prefer Gu3l�n, because Soran does not contain any discussion, but for non-natives who read slowly and only with effort, Soran provides a very convenient reference source.

Another useful tool is the concordia, which are ways to look up inscriptions by topic or by particular characters contained within them. So this is a great way to compare a bunch of inscriptions with a particular graph or compound, in order to help understand its usage. The most accessible is 姚孝遂 Y�o Xi�osu� and 肖丁Xi�o Dīng (BTW, the latter is the pen name of the above-mentioned 趙誠 Zh�o Ch�ng, q.v.), 1989?.《殷墟甲骨刻辭類纂》yīnxū jiagu k�c� l�izuan, commonly abbreviated 《類纂》 L�izuan; three volumes. 北京 Bĕijīng: 中華書局 Zhōngh�a Shūj�. Like the earlier and better-known 《綜類》 Sorui, i.e., 《殷墟卜辭 綜類》Inkyo bokuji sorui by 島邦男 Shima, Kunio (which is harder to get, and vey expensive), the L�izuan compendium organizes large numbers of oracle bone inscriptions by content, so that one may compare various instances of a graph, compound, or phrase. Like the 摩釋 M�sh� by the same two authors, it follows each handwritten oracle bone script entry with a rendition in modern kaishū (insofar as is possible). L�izuan�s indexing allows searching by common oracle bone components (similar to the b�shou index in modern dictionaries). Although it contains more examples than Sorui, it also has numerous errors and omissions , and is not sufficient by itself for scholars to make definitive statistical statements about numbers of occurrences. Its OB transcriptions and kaishū renditions are sometimes erroneous as well, and so should also be compared with the rubbings in H�j�, the kaishū transcriptions in Sh�w�n and Soran, and even the analyses in Gu3l�n. But it is definitely suitable for students.

Serious (professional) scholars will want to move up to other reference sets, like Gǔl�n《甲骨文字詁林》jiaguw�n gul�n by 于省吾 Y� Xĭngw�, of 吉林大學 J�l�n University. $760RMB. This contains an authoritative summary (in handwritten Chinese) of scholarly discussion and analysis for individual oracle bone characters, and is thus more useful for gaining a comprehensive understanding than, say, Soran, which merely lists single kaid�ng renditions by a variety of scholars. A second set for pro�s is李孝定 Lĭ Xi�od�ng (Lee Hsiao-ting, 1965). 甲骨文字集釋 jiaguw�nz� j�sh�, (collected interpretations of oracle bone characters), 台北 T�ibĕi, 南港 N�ngang3 (Nankang): 中央研究院歷史語言研究所 Institute of History and Philology, Academia Sinica. This sixteen-volume, 5000-page work (bound in eight volumes) is the primary, authoritative collection of summarized commentaries by various scholars on individual OB graphs, with the author’s own review and conclusions on each. However, the handwriting, the difficult literary style, massive size and correspondingly high price are each so daunting that I cannot recommend them to you.

Another useful reference is the Hanyu Da Zidian, 漢語大字典. H�bĕi Cishu Chūban3she4 and Si4chuān Cishu Chūban3she4, 1992. This massive tome is the most comprehensive dictionary for classical Chinese and obsolete graphs, and as a bonus, contains partial graphic histories of many characters. A Taiwanese edition (converted from simplified to traditional characters) is available from 建宏出版社 Jian4hong2 Publ., ISBN 957-813-478-9.

On early pottery graphs, try Woon, Wee Lee (1987). Chinese Writing: Its Origin and Evolution. Explores in depth the early pottery inscriptions, from the Neolithic Banpo to the excavations at the last Shāng Dyn. capital at Xiaotun (Anyang, Henan Province) in light of OB (oracle bone) and bronze forms. Discusses at length the processes involved in the formation and evolution of characters, with copious oracle bone and bronze examples, some obsolete. Perhaps a bit too scholarly for beginning students, but certainly worthwhile reading for more advanced students and ‘hardcore’ amateur sinologists. Bibliography is mainly Chinese sources, unfortunately in pinyin without characters, and thus of limited value. Also unfortunately not indexed by character, limiting its value as a reference source. Originally publ. by the Univ. of East Asia, Macau (no ISBN); now available through Joint Publishing, fax: 852-28104201; email: (attn: Edith Ho kit-sheung). Note: the Joint Publishing staff can’t seem to handle English titles well; be sure to send the author and title in Chinese by fax to get the right book: 作者: 雲惟利, 書名: 漢字的原始和演變.

Also worth a look is Boltz, William G. (1994). The Origin and Early Development of the Chinese Writing System. American Oriental Series, vol. 78. American Oriental Society, New Haven, Connecticut, USA. ISBN 0-940490-78-1. A fascinating scholarly treatise on the theory behind title subject, with a grammatonomic approach, emphasizing the internal, or linguistic processes, rather than the external, or material aspect (graphic structures). Boltz rejects the notion of the ideograph (graphs which represent meaning without resorting to sound), and emphasizes the graph as representation of the spoken word. He explains the motivations for and problems resulting from the borrowing of characters for other, spoken words, in a four-step model of the development of writing systems, from 1) logographs, to 2) two kinds of loans, semantic loans resulting in po4yin1zi4 (polyphony) and the resultant phonetic ambiguity; and phonetic (rebus) loans resulting in semantic ambiguity; 3) determinatives added to resolve ambiguity; and finally (in other languages but not Chinese) 4) desemanticization of the logographs, resulting in a purely phonetic system of writing.
Boltz patiently guides the reader through the kinds of reasoning needed to infer etymological processes. Particularly interesting is the reasoning on how to infer that some currently monophonic characters must have once been polyphonic, inferring polyphony for the components of a number of supposed hui4yi4 (logical aggregate, compound indicative, or associative compound) characters such as an1 ‘tranquility’ and hao3 ‘good’, thus dispelling mythical etymologies like “woman under roof represents tranquility”. Boltz explains how to determine whether the role of a particular component is as an original semantic or phonetic one (or both), or instead as a determinative resolving phonetic ambiguity of a polyphonic character or semantic ambiguity of a parasemantic or polyphonic graph.
Perhaps too scholarly for the average beginning student of Mandarin, and not arranged or indexed to be an etymological reference works, but certainly worthwhile reading for the “hardcore” amateur sinologist and intermediate to advanced Chinese major.

Well, those should get you started :laughing: . Let me know if you have any questions. I don’t have similar info on bronze references yet, other than what’s in some of the above like Liu2, Xie4, Wu2, and the Hanyu Da Zidian, but give me a couple years and I’ll probably have some to recommend to you.

Prof. Keightley seems to be a good friend of my former University Professor and once came to our seminar to hold a lecture on oracel bones. I’m sure he had prepared it, but nevertheless, he just read the bones like he would read a modern book… :astonished: :astonished: We were all veeeeery impressed!! :notworthy: :notworthy: :notworthy:

Excellent reply, thanks for the references. I met Harbaugh as he taught at my college for a while. He specifically noted that his dictionary was not a scholarly work, and even made some apology for it, but I think it was rather meant to make learning characters easier for beginning and intermediate students.

Do you know about learning anything about early bamboo strip scripts?

Yes, it most certainly is impressive! But if you think about it, any jiaguologist has to be fluent in the inscriptions just as if reading a modern book. It’s a lot like learning to read modern Chinese characters: the non-reading foreigner will be very impressed by the highly literate one, but by the time you become literate, it doesn’t seem like such a big deal. Guess that’s true of any arcane field of expertise. I will say this, though – I have found that it really is harder to learn the OB for a variety of reasons. There is a veritable paucity of textbooks; it can’t be reinforced by your daily interactions; there are a huge number of graphs which have not yet been identified or deciphered, and many which scholars are still disputing; many words have a variety of graphs, or shall we say, many graphs have many variant forms, adding to the complexity of the task; good dicitonaries and reference works are obscure, expensive, and hard to acquire and use; and many of the reading materials are damaged or fragmentary; and the direction of writing changes at random, making 45 to 135 degree turns willy nilly, running into other inscriptions which may or may not be related, and so on.

Note that we can now benefit from the work of all the earlier scholars who did not have the resources I now have, and this to me is the most impressive thing of all about their achievements, that they did it clearing their own trail through these forests of strange glyphs !

[quote=“TaiOanKok”]I met Harbaugh as he taught at my college for a while. He specifically noted that his dictionary was not a scholarly work, and even made some apology for it, but I think it was rather meant to make learning characters easier for beginning and intermediate students.

Hope I wasn’t too hard on him. Actually, I love his book, for the way it is organized. I had long been cross-referencing my own personal dictionary in much the same way. I think it is a valuable book for beginning to intermediate students, for “anti-radical” lookups, if you will, and for mnemonic purposes. What I object to, and strongly, is any implication that it is valuable for etymology, as it implies. Not only are his readings chiefly the out-of-date Shuōw