Talk:Shang dynasty

Radiocarbon dates

I have reverted a recent change of the dates based on Liu et al (2020), on the basis that it gave undue weight to a single source. Moreover the source itself does not claim that these are the exact dates. Their method was to define a range for each period as the union of the 68% confidence intervals for individual bones from that period, yielding this table of reign dates (their Table 2):

Kings	cal dates (BC)
Wu Ding	1254–1197
Zu Geng and Zu Jia	1206–1177
Lin Xin and Kang Deng	1187–1135
Wu Yi and Wen Ding	1157–1110
Di Yi and Di Xin	1121–1041

The date ranges overlap, and the authors say they have an uncertainty of about 10 years and up to 20 years in extreme cases. They do not claim to contradict the XSZ project dates, but rather say the XYZ dates are within these uncertainty ranges. Kanguole 13:45, 15 March 2022 (UTC)

We are sidelining the latest peer-reviewed radiocarbon dating in favor of XSZ? This is a transparently nationalist project that includes dates for mythical events like the accession of the Xia. Trying to date events based on Chinese astronomical records is a fool's errand. For many years, the supposed solar eclipse of 776 BC was one of the best-known dates of Chinese history. Later, it was shown that there was in fact no eclipse that year and that the ancient records were in error. The Shang used three or four different calendars, each of which used a different date as the first of the month. The record of the supposed lunar eclipse of 1065 BC that XSZ is relying on goes like this: “at the ceremony paying homage to the full moon ... the king announced, ‘The many[...] eclipse(s) is/are untimely; you shall begin planning succession’.” This is not a record of an eclipse on a particular day, or even a particular year. See “Astro-Historiographic chronologies of early China are unfounded.” by Douglas Keenan. Fairnesscounts (talk) 20:17, 15 March 2022 (UTC)

"the latest peer-reviewed radiocarbon dating" is a single source, whereas the XSZ project is essentially the consensus of modern Chinese academia. There has always been a split between the West and China on these matters, and we can certainly not favor a single Western paper over the entirety of Chinese scholarship. Your personal evaluation of the so-called "transparently nationalist project" is pointlessly rude and largely WP:OR. Aza24 (talk) 03:43, 16 March 2022 (UTC)

These authors were part of the XSZ project, and continue in the same tradition. They have the same uncritical acceptance of a 59-year reign from the Wuyi chapter of the Shangshu, even though the same text is known to be wrong on reign lengths for the last Shang kings, which would have been very recent history if it was really written in 1030 BC. They also present their results as compatible with the project. In any case, many of the dates adopted by the XSZ project had already been proposed by authors outside of China.

But the key issue is, as noted above, that they do not claim these are precise dates for various events, but have an uncertainty of 10–20 years. The reason for this is clear from the detail of the paper. For example, the source of the 1041 BC date is the 68% confidence interval 1118–1041 BC for one of the three objects from period V that they dated. That's an uncertainty range of 77 years. Identifying the end date of a confidence interval for a single object with the end date of the Shang is quite arbitrary. Kanguole 10:10, 16 March 2022 (UTC)

As I read the report, the conclusions are not based on any individual radiocarbon reading. The report explains that by using data from consecutive phases, they could apply the Bayesian method, narrow the date range, and increase reliability. Liu is claiming a "fall-in ratio" of 80 to 90 percent. The FIR represents the author's confidence that the divinations in a given phase all occurred within the given date range. In short, the 1041 BC date represents an upper bound for phase V divinations. This would mean that whoever wrote the abstract misunderstood the study since error bars are obviously not equivalent to regnal dates.

Cambridge gives the Zhou accession date as 1045-1046 BC. This is within the error margin that Liu gives for phase V (1121–1041 BC).Fairnesscounts (talk) 20:54, 17 March 2022 (UTC)

I interpreted "superposition of the 68% date ranges of all the samples in that phase" as meaning the union of the 68% date ranges of all samples for a phase. That explains the ranges they give, except for the end of period III and the start of period V, which are off by 2 and 1 years respectively, but I don't think an adjustment method is described. So 1254 BC comes from the start of the earliest range, and 1041 BC comes from the end of the latest one.

It's not surprising that the Cambridge History and the XSZ project came up with much the same date for the Zhou conquest, since both dates are based on the same analysis of the same astronomical events, though the XSZ project also used archaeological evidence and C14 dating. Kanguole 22:39, 17 March 2022 (UTC)

Here's a timeline of the 68% ranges of various objects in their Table 3:

CASS HN M99(3):1

Heji 2140

Heji 21565

Heji 22116

Heji 19779

Heji 22086

Heji 22184

Heji 6846

Heji 6774

Heji 10410

Heji 9816

Heji 3186

Heji 3089

Heji 4122

Heji 31997

Heji 3013

Heji 34120

Heji 21739

Heji 6883

CASS HN M99(3):2

Xiaotun 910

Heji 22594

M5 (Fu Hao)

Heji 13329

Heji 21784

Heji 2869

Heji 23340

Heji 25015

Heji 26766

Heji 23536

Heji 24610

Heji 27616

Xiaotun 1011

Xiaotun 2209

CASS K271A

Xiaotun 2996

Xiaotun 2370

Xiaotun 2315

Heji 27633

Xiaotun 173

Xiaotun 2557

Xiaotun 2294

Heji 35249

Xiaotun 2263

Xiaotun 647

Xiaotun 2281

Xiaotun 3564

Heji 36512

Heji 35641

M1713 (Western Zone)

│

−1260

│

−1240

│

−1220

│

−1200

│

−1180

│

−1160

│

−1140

│

−1120

│

−1100

│

−1080

│

−1060

│

−1040

│

−1020

Diviner groups:   Wu,   Shi,   Zi,   Bin,   Chu,   Unnamed,   He,   Huang.

It is wierd that the ranges are clumped within each of the five periods, especially II–V. I can't think of a good explanation for that. Kanguole 13:39, 18 March 2022 (UTC)

Ah, I see that they discarded the results for over a third of the samples, many because of an "unacceptable low agreement index", i.e. being out of the sequence implied by the periods, even though they went down to 50% in some cases. Earlier they say that "On average, 1 in 20 agreement index values to drop below 60% might be expected, but if the index values are substantially lower or a large proportion fall below 60%, something internally inconsistent between the model and the data could have occurred". Kanguole 16:10, 21 March 2022 (UTC)

For completeness, here are their 68% ranges for the Li-group bones:

Heji 34240

CASS T8(3):148

Heji 32764

Xiaotun 994

Xiaotun 601

Xiaotun 1115

Xiaotun 2707

Xiaotun 1090

Xiaotun 1128

Heji 32780

Xiaotun 636

Xiaotun 1116

Xiaotun 503

Xiaotun 2366

│

−1320

│

−1300

│

−1280

│

−1260

│

−1240

│

−1220

│

−1200

│

−1180

│

−1160

│

−1140

│

−1120

│

−1100

Diviner groups:   Li type 1,   Li type 2.

Again the odd clumping. Kanguole 16:10, 21 March 2022 (UTC)

Establishment year 1555/1554 BC

With reference to this edit;
The dates of the establishment of the Shang dynasty proposed by Pankenier and Nivison (1555/1554 BC) are I far as I see it not in the conflict with the result of the more approximate dating 1600 BC from the Xia–Shang–Zhou Chronology Project. I would say that the Pankenier and Nivison dates are trustworthy and is a further development of the research. I propose to use 1554 BC as the main date for the establishment of the Shang dynasty in this article. (Same goes for the corresponding dates for Xia dynasty). BR --Bairuilong (talk) 09:34, 22 July 2022 (UTC)

Why are these dates trustworthy? Nivison agrees with Pankenier, but no-one else seems to. Indeed most other workers regard the "current text" Bamboo Annals, on which these dates are based, as unreliable. Perhaps the more approximate date is an accurate reflection of the state of the field. Kanguole 10:06, 22 July 2022 (UTC)

You write that no-one seems to trust Nivison and Pankenier. Can you please give some references that support that credible historians distrust 1555/1554 BC? Thank you in advance! BR/--Bairuilong (talk) 10:53, 22 July 2022 (UTC)

That's the wrong way round. References are needed to show that these dates are accepted.

As for the "current text" Bamboo Annals being distrusted, see e.g. Keightley "The Bamboo Annals and Shang-Chou Chronology", Barnard "Astronomical data from ancient Chinese records: the requirements of historical research methodology". Kanguole 11:28, 22 July 2022 (UTC)

You are referring to old publications that I don't think give support against 1555/1554 BC. However, there are a huge number of publications that support 1555/1554 BC. First of all of course several publications by Nivison and Pankenier, and besides that for example: Maisels "Early Civilizations of the Old World", Feng "Early China: A Social And Cultural History", Hägerdal "Kinas historia". It's easy to list several more. (Interesting that you mention Keightley as argument against 1554 BC, since he actually has accepted 1554 BC.).

I can not see any relevant argument against 1555/1554 BC, but on the other hand a lot of support. BR --Bairuilong (talk) 12:55, 22 July 2022 (UTC)

If you are surveying third-party sources, you'll find that even more of them say 1600 BC.

Incidentally, although Li Feng says that Keightley has accepted this date, if you look at the citation he gives you'll see that Keightley is reporting Pankenier's calculation, but not endorsing it. (Which is not surprising, since Keightley's focus was what can be found in excavated texts, which start centuries later.)

There is a diversity of views, so there is no justification for presenting the 1555/1554 BC date as a bald fact. Per WP:ATTRIBUTEPOV, we should present the significant alternative views and say who holds them: Pankenier and Nivison say this and XSZ says that. Nor is it appropriate to call it the latest research, since Pankenier's work (early 1980s) and Nivison's agreement (around 1990) pre-date the XSZ project. Kanguole 16:18, 22 July 2022 (UTC)

Of course all relevant alternative should be presented in the article. I agree about that for sure. The point is what should be in the infobox. You have not shown anything that support your statement that no-one seems to trust Nivison and Pankenier, and the opposite is easy to find - there are a lot of reliable sources that support 1555/1554 BC, and both Nivison and Pankenier are heavily credible sources on this subject. Of course it is possible to find sources that still write 1600 BC (and also 1766 BC, 1675 BC and 1559 BC), but that is not an argument for us to forever stick to 1600 BC. Specially since it is not so hard to find criticism of the XSZ project and its result. BR --Bairuilong (talk) 16:59, 22 July 2022 (UTC)

In saying "still write 1600 BC", you are clinging to the idea that the Pankenier/Nivison proposal is the most recent research, when in fact it is the other way round: the XSZ project was conducted after their proposal.

Both proposals have been criticized, and both are cited (but as I said, the XSZ project dates are cited more often). So both should be presented as the views of their proposers.

As you say, that leaves the field in the infobox. It is inappropriate to present 1554 BC as the definitive answer here. You might well reply that it is also inappropriate to present 1600 BC as the sole answer. So a less definite statement seems to be called for, say "early 16th century BC". Kanguole 17:31, 22 July 2022 (UTC)

Thanks for a constructive reasoning. The essence of my point is that Pankenier/Nivison hold the most credible alternative according to my understanding.

I am also aware that there are margins of error to indicate exact years for event several thousand years back in time with this level of sources. I guess we will never be totally sure, and this also applies, for example, to 1046 BC and several other year for this millennium. Nevertheless we need to select the information from the most credible source. It seems that your opinion is that XSZ project is the most credible. I agreed with that for a long time, but have changed my mind the more I read about the Pankenier/Nivison research.

I appreciate that you make a compromise proposal, but I want to make a counter-proposal: "1550s BC". How about that? BR --Bairuilong (talk) 17:54, 22 July 2022 (UTC)

The 1554 BC date is based on a five-planet conjunction that supposedly occurred in 1576 BC. This was centuries before writing was introduced to China, so there is nothing resembling a contemporary record. Venus was 40 degrees away from conjunction on the night in question, so it was four-planet conjunction. These are too common to be used for chronology. What are the chances that both the beginning and the end of the Shang dynasty were predicted by five-planet conjunctions? The ancients loved cyclical history and astrologers had an interest in promoting the idea that events can be predicted by astrology. IMO, a round number is best when dating a possibly mythical event. See "Astro-Historigraphic Chronologies of Early China are Unfounded:"

"Several researchers have made proposals for a chronology of ancient China that rely on records of astronomical events—solar eclipses or five-planet conjunctions. It is shown herein that either the events did not occur or there is no reliable record of them, or both, and that such problems are unrectifiable." While XSZ is based on eclipse data, Cambridge History uses a dating system based on five-planet conjunctions, according to the article.

As for the Bamboo Annals, it records an eclipse in 776 BC that did not occur. This supposed eclipse was at one time quite well known to historians, but it was later debunked. Spring and Autumn Annals records solar eclipses on 3 April 645 BC,15 May 592 BC, 19 September 552 BC, and 18 July 549 BC. None of these supposed eclipses actually occurred.

1600 BC is by far the most commonly given start date for the Shang dynasty, according to this ngram. This has been the case going back to the 19th century, long before XSZ. Fairnesscounts (talk) 01:20, 23 July 2022 (UTC)

Fortunately, we will certainly not base the dating in our articles on Google-statistic. Add (2001 - 1046 BC) to your statistic, and you have a winner, and a food for thought what your graphics show. Your statistical graphics have very little in common with the state of research for the dating of the Shang Dynasty. BR --Bairuilong (talk) 03:24, 23 July 2022 (UTC)

To also give a more nuanced answer and comment on the Keenan-article:

Keenan places great emphasis on the fact that he judge that Venus was to far away from the other planets 1576 BC in order to be counted as a true five-planet conjunction. However, it is quite extreme to make that assessment of the exact meaning what "五星錯行" meant when was written thousands years ago. It is also important to add the fact that not least Nivison add a important layer of chronology synchronisation on top of the years for the astronomical event. The 1576 BC planet conjunction is only one part of the background to the 1555/1554 BC dating. BR --Bairuilong (talk) 08:49, 23 July 2022 (UTC)

I don't believe the XSZ project is particularly credible, though I do appreciate their modesty in choosing a round number. The truth is that there can be very little basis for dating events from centuries prior to the invention of writing in a given part of the world. All we can do is assess which dates are the most used in the sources. It is possible to trick ngrams, but it is still true that 1600 BC is the most commonly used date.

Astronomical events offer very precise dates, but using them for this period means using the "current text" Bamboo Annals, which most (but not all) scholars believe to be a Ming-era forgery – even champions of the "current text" like Nivison and Shaughnessy acknowledge that they are in the minority. Even the original Bamboo Annals was compiled in the Warring States period, more than a millennium after the dates at issue. Pankenier then has to interpret opaque passages as referring to five-planet conjunctions in certain parts of the sky. Nivison's dating is based on a complex re-interpretation of the reign lengths given in the "current text" Bamboo Annals; he acknowledges that almost no-one agrees with him on this. They build elaborate structures, on completely unreliable foundations.

"1550s" is a less precise version of 1555, but it is still unjustifiably precise. Kanguole 09:57, 23 July 2022 (UTC)

First: We should definitely drop the discussions of what is most frequent published. We need to focus on judge which credible source publishes what.

Having said that; If we should find a dating formulation that cover result from both XSZ project and Pankenier/Nivison I think we need to look at the background to "1600 BC". The way I understand this date is that the XSZ project could not agree if Erlitou Phase III belongs to Xia or Shang. Rounded worded I think we can say that the joint between Phase II and III is about 1600 BC and the joint between Phase III and IV is about 1560 BC. My conclusion of this is that the XSZ project-intention is that Shang begins about 1600 BC or about 1560 BC, and the round date "c. 1600 BC" could be imprecise enough to cover both alternative. I think the expression: "First half of 16th century BC" cover the intention of XSZ project and the Pankenier/Nivison dates. Could this be something? BR --Bairuilong (talk) 11:25, 23 July 2022 (UTC)

• Keightley (1999), David Keightley's essay for Cambridge History, is the most cited serious source on Shang China, including 14 times in this article. It is also recommended by Wilkinson (2013), a manual for historians of China. Keightley dates the dynasty as "ca. 1570-1045 BC." Fairbank & Goldman (2006) gets 12 citations. (This is harder to justify since it has only a few pages on the Shang.) In any case, it dates the dynasty as 1750–1040 BC. Lee (2002) gets four cites. His dates are 1600-1046 BC. He cites XSZ.

Encyclopedia.com returns Gale Encyclopedia of World History, which gives 1766 to 1122 BC, and Oxford World Encyclopedia, which gives c.1523–c.1030 bc. Oxford Reference returns Berkshire Encyclopedia of China, which gives 1766 bce–1045 bce.

Since our sourcing consists primarily of Cambridge History authors and editors such as Keightley, Bagley, and Shaughnessy, I would conclude that we should be using the Cambridge History date for the beginning of the dynasty, namely c. 1570 BC. Fairnesscounts (talk) 15:12, 26 July 2022 (UTC)

As far as I see, the year 1570 BC is only mentioned within parenthesis in the Cambridge History. When Keightley in chapter "Chronology / Absolute Dating" goes in to details regarding dating of the beginning of the dynasty he actually referring to Pankenier and the dating 1554 BC, and he does not give any other dates more weight. I still would say that "First half of 16th century BC" is the best interpretation of the existing source state, if we decide to not go for 1554 BC. BR --Bairuilong (talk) 03:41, 9 August 2022 (UTC)

What source says "first half of 16th century BC"? Surely Cambridge History is better than using no source. Fairnesscounts (talk) 08:28, 16 August 2022 (UTC)

You can find the background to the proposal if you read my text above from 11:25, 23 July 2022 that it is a way to include both the intention from the XSZ project and the Pankenier/Nivison dates in an inclusive expression. I think "Cambridge History of Ancient China" is a good source. The way I understand the Shang dating chapter (D.N. Keightley, chapter "Chronology / Absolute Dating". page 248) is that the only relevant year mentioned for the establishment of the Shang is 1554 BC (with reference to Pankenier). The "1570" within parenthesis is not mentioned and left without explanation in the text. If we strictly should follow Cambridge History of Ancient China, then I think we should go for 1554 BC. That's also fine - and that was my original suggestion in this discussion. BR --Bairuilong (talk) 16:48, 16 August 2022 (UTC)

What you are proposing is called WP:SYNTHESIS. The most common solution is to use 1600 BC as the start date of this dynasty. It's a nice round number that confesses lack of precise knowledge. Dates based on Chinese astrology are notoriously inaccurate. This one seems particularly suspicious since it is centuries before the introduction of writing. Fairnesscounts (talk) 18:50, 16 August 2022 (UTC)

Point taken about synthesis. So you don't any longer support that we should follow Cambridge History? XSZ project just became the most credible source, or how should I understand your sudden reversal? BR --Bairuilong (talk) 03:55, 17 August 2022 (UTC)

I am not a fan of XSZ. The inclusion of the mythical Xia dynasty shows that the project is driven by nationalism. But the 1600 BC date was widely used before XSZ. Violence, Kinship, and the Early Chinese State (2018) by Roderick Campbell has a good take on this. Instead of dynasties, he has periods based on archaeology. There's an Erlitou Period (ca. 1800–1600 BC), an Erligang Period (ca. 1600–1400 BC), a Xiaoshuangqiao-Huanbei Period (ca. 1400–1250 BC), and an Anyang Period (ca. 1250–1050 BC). Fairnesscounts (talk) 09:51, 17 August 2022 (UTC)

I see you propose new sources in every post. Can you please come back to my question above why you suddenly changed your mind about Cambridge History and Keightley? Have you actually read the Absolute Dating-chapter? BR --Bairuilong (talk) 14:19, 17 August 2022 (UTC)

I gave the start date as "1570 BC" because IMO this is the most regular way to do it under our guidelines. It is the way the founding of the dynasty is given by our principle source, namely Cambridge History. There are certainly problems with the date. If you look at Campbell, he doesn't mention 1570 BC at all. Keightley doesn't explain where the date comes from. Fairnesscounts (talk) 05:59, 18 August 2022 (UTC)

You are correct that it is not explained where 1570 BC comes from, and it is also just mentioned briefly within parentheses, so we don't need to put so much attention to that year. On the other hand, the section about the first year of Shang in the Absolut dating-chapter (page 248), Keightley quote: "[...] Using the absolute date of such astronomical events it is possible to conclude that the first year of Chang Tang would have been 1554." (with reference to Pankenier). This is the only specific starting date for the Shang that Cambridge History of Ancient China point at as far as I have seen. It is not easy to get around the fact that Cambridge History supports 1554 BC. Best regards, --Bairuilong (talk) 14:52, 18 August 2022 (UTC)

too-detailed scholarly discussion of dates in lede?

Reading the first paragraph of this article, I am surprised by the in-text inclusion of the names of three scholars and a scholarly project (Pankenier, Nivision, Shaughnessy, and the Xia-Shang-Zhou Chronology Project) in debate with each other on the dating of this dynasty. This is valuable for certain, but perhaps too much detail for the first paragraph. I suggest instead, after the sentence on how the current text of the Bamboo Annals suggests rule dates of 1556 to 1046 BC, something like: “Modern scholarship supports this chronology, and radiocarbon dating from archeological sites has suggested a founding date as early as c. 1600 BC.” Interested readers could turn to the Absolute Chronology section for more detailed information. Thoughts? Resaliest (talk) 04:39, 7 March 2023 (UTC)

100% they need to be discussed as is, but yes the place for that is in their own section and not the lead. The lead can just begin with the generally accepted dates and acknowledgement of dispute or simply say that the specific dates are disputed within a range of ~. — LlywelynII 03:44, 2 September 2023 (UTC)

I don't see any reason to mention the traditional dates at all in the lead, considering there is practically no use of them in contemporary academia. Aza24 (talk) 06:41, 2 September 2023 (UTC)

Sources for future article expansion

• Li Xueqin (1991), "Chu Bronzes and Chu Culture" (PDF), New Perspectives on Chu Culture during the Eastern Zhou Period, Princeton: Princeton University Press, pp. 1–22.

has more on the Shang's relationship with the south in the area that later became Chu, Wu, and Yue. — LlywelynII 03:44, 2 September 2023 (UTC)

Copyvio paragraph

Given recent incidents in articles around early Chinese history/archeology, I've been more wary of copyvio—this revision by Porcinipal appears to plagiarize a paragraph pretty much verbatim from the source it's attached to, which is unacceptable. It's only one, but I suppose I've been seeing it too much in this space to be happy about it. Remsense留 02:15, 15 December 2023 (UTC)

Spot-checking their other contributions, they seem to do this with approximately one paragraph per article. Remsense留 02:19, 15 December 2023 (UTC)