Talk:Roman metallurgy/Archive 1

Roman iron production exaggeration
84,750 tons is an exaggeration and the absolute high estimate of Roman iron production. It is based on 1970-1980s estimates of Romano-British iron production being 2000-3000 tons while be proportionally smaller in terms of overall production. Other more modern estimates for Romano-British iron production range include 700-800 tons annually, and Britain playing a larger role in overall production. See the "The Wealden Iron Industry" - or the website listed below that sums up some of its major points. J. S., Hodgkinson. 2008. "The Wealden Iron Industry." (The History Press, Stroud).

Gun Powder Ma, or 'Tibet Libre' of CHF has been discussing the validity of the Roman metallurgical estimates of 80k tons on CHF forums. If you are familiar with the subject and could input some knowledge, that would be great.


 * http://www.chinahistoryforum.com/index.php?/topic/34145-the-pre-industrial-economies-of-western-europe-and-china/
 * http://www.chinahistoryforum.com/index.php?/topic/13206-han-vs-rome-military-comparisons/page__st__1365

Intranetusa (talk) 03:00, 29 June 2010 (UTC)

Dubious comment on iron production figure
Removing a complete table when you disagree with just a single number is disruptive and vandalism. Then, Hodgkinson, the source you claim you quote, nowhere refers to Craddock's 84,750 t for the whole empire as an "high estimate of Roman iron production". It actually does not deal with Craddock at all. It only deals with Roman iron production in the English Weald, that is in only one production center in one province (Roman Britain). So how on earth do you come to your "disclaimer"? Have you checked out WP:OR yet? Gun Powder Ma (talk) 17:40, 29 June 2010 (UTC)
 * The estimated 84,750 t of Craddock 2008 (who btw is one of the foremost expert on ancient metallurgy) conform to WP:Verifiability. However, this edit is not at all supported by the source cited, hence it needs to be removed according to the same guideline. Gun Powder Ma (talk) 20:41, 29 June 2010 (UTC)
 * I added two more references supporting the approximately 80,000 t. There is an half a page long discussion in Healy, but I can't be arsed to quote it here given Intranetusa's bullish bevahiour. Your loss. Gun Powder Ma (talk) 20:17, 30 June 2010 (UTC)

It wouldn't hurt to: Intranetusa (talk) 03:34, 1 July 2010 (UTC)
 * 1. provide some newer data rather than ones that date back to the 1980s, or data by other scholars that confirms the original *figures
 * 2. provide some sources on how these figures were calculated. At this point you're just pulling numbers out of books. I could *quote a thousand books that give crazy estimates and figures too.
 * 3. We don't even know how these numbers were calculated in the first place. There is absolutely no record on the total amount of *annual iron production by the Romans themselves. That's why you don't see crazy figures for other articles - like the ancient *Egyptians in 1000 BCE produced 1000 tons of copper and 10 tons of gold a year, or the Gupta Dynasty produced 500 tons of silver. *Nobody knows what the actual figures are unless the ancients wrote it down themselves!
 * 1. The estimate has been adopted by Craddock in his 2008 in the The Oxford Handbook of Engineering and Technology in the Classical World, so it is considers up-to date by the author who is an international expert on ancient metallurgy.
 * 2. The estimate is referenced by three sources.
 * 3. Blabla

I thus remove your synthesis. Gun Powder Ma (talk) 10:30, 1 July 2010 (UTC)

Dubious iron production figures
The 82,500 tons iron comes from Craddock p. 108, which uses Cleere and Crossley 1985: 57-86. It relies on the rather high estimate of 2250 tons annually for Romano-British iron production. Lower estimates are credible and has to be taken into account. Intranetusa (talk) 18:18, 29 June 2010 (UTC)
 * If the figures are to be used, it is surely important to establish how they were estimated. What method was used to arrive at the figure quoted? Is it based on reliable estimates of extracted ore or archaeological remains or what? I think the exercise is beset with problems, and the least that could be done is to provide error bands to suggest that the figures are guesstimates rather than based on, say, statistical returns to the government in Rome. Peterlewis (talk) 18:44, 29 June 2010 (UTC)
 * Error bands could be useful if they are available in the literature. If not, it is not up to a Wikipedia editor to analyze the data to generate an error band.  That would be original research.  --SteveMcCluskey (talk) 01:36, 30 June 2010 (UTC)
 * Dito. There are unfortunately no error bands; it is an estimate a priori not better, nor worse than any other, but from an international expert in a recent, high-quality publication on ancient technology, so absoluty in accordance with WP:Verifiability. This hysterically going around, deleting whole tables and trying to discredit the estimate by adding fabricated criticism which is not by any means supported by the cited sources is pretty immature (talk about WP crowded by nerds and singles). Gun Powder Ma (talk) 10:50, 30 June 2010 (UTC)

Could you explain the "a priori" method then? This seems to me to be meaningless. Peterlewis (talk) 12:59, 30 June 2010 (UTC)
 * Verifiability: The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material added to Wikipedia has already been published by a reliable source, not whether editors think it is true.


 * Craddock's ca. 80,000 t = verifiable = keep
 * Intranetusa's comment construing some outlandish connection between Wealden output and Craddock's estimate = not supported by source = remove Gun Powder Ma (talk) 13:08, 30 June 2010 (UTC)

The comment is not an outlandish claim. Craddock's estimate of 82,500 tons per year is an outlandish claim based on outdated data. He based this number of an exaggerated amount of 2000-3000 tons of annual iron production in Roman Britain/Wealden Britain. There are plenty of figures out there that aren't that high, so it stands to reason this is an exaggeration and/or the high estimate. The 700-800 tons figure for annual Romano-British production is a figure that is just as credible, and more believable. Take a look at the wiki article on the history of the Wealden Industry that contradicts Craddock's data. You currently have two wiki articles with contradictory statistics: "Total iron production has been estimated at 700-800 tons per year, but under one third of that after 250 AD." Intranetusa (talk) 23:49, 30 June 2010 (UTC)
 * http://en.wikipedia.org/wiki/Wealden_iron_industry
 * http://en.academic.ru/dic.nsf/enwiki/543912
 * This amount only refers to one iron production center in Roman Britain (the Weald). How this should "contradict" Craddock's estimate of ca. 2,000 t for the whole of Britain and of ca. 80,000 t for the whole empire is beyond me: Your http://en.academic.ru/dic.nsf/enwiki/543912 has nothing to say whatever about total output of Britain or the Empire. I know it rattles your Han national pride big time that modern scholars find the Roman Empire outproducing all other ancient empires by two-digit factors, but if you cannot even stomach harmless estimates move back from Baltimore, USA, to where you've been making cheap propaganda for all along. Gun Powder Ma (talk) 00:14, 1 July 2010 (UTC)

Sorry to disappoint you, but I'm not a Han nationalist. If you're going to play low ball like that - why, are you an Italian nationalist or something? And if you really want to bring that up, then we can discuss you constantly creating ridiculous and ill-mannered threads on CHF forums and elsewhere. I know you hate the modern Chinese communist government and want Tibet to be free, but that's obviously cloudy your judgment regarding ancient history. Literally every topic and every post you've made on CHF is about how East Asia sucks. You should understand that modern events is not representative of ancient history. The fact that you edited Roman metallurgical topics immediately after you had an argument with on CHF makes me highly suspicious of your motives. And you haven't bothered responding to the criticisms of your flawed statistics either.

Sinocentrism is bad, but bashing entire cultures and promoting other forms of biased centrisms like you are now is just as bad. That's just utter hypocrisy.

Now moving on to the topic: Intranetusa (talk) 03:13, 1 July 2010 (UTC)
 * 1. provide some newer data rather than ones that date back to the 1980s, or data by other scholars that confirms the original *figures
 * 2. provide some sources on how these figures were calculated. At this point you're just pulling numbers out of books. I could *quote a thousand books that give crazy estimates and figures too.
 * 3. We don't even know how these numbers were calculated in the first place. There is absolutely no record on the total amount of *annual iron production by the Romans themselves. That's why you don't see crazy figures for other articles - like the ancient *Egyptians in 1000 BCE produced 1000 tons of copper and 10 tons of gold a year, or the Gupta Dynasty produced 500 tons of silver. *Nobody knows what the actual figures are unless the ancients wrote it down themselves.
 * Talking about low ball: although you might not have even noticed yet, but in most places on earth, and I subscribe to that view heartily, people who do renrou sousuo are considered as morally defective. Gun Powder Ma (talk) 10:39, 1 July 2010 (UTC)

Unfortunately, I'm not exactly well acquainted with the communist terminologies that you seem to be so intimately familiar with. And after reading that article, I don't see how that term even applies to this situation either. Considering how nationalistic and vitriolic your posts have been on other forums, your assertions would be the pot calling the kettle black. Intranetusa (talk) 20:05, 1 July 2010 (UTC)


 * I do not normally deal with this period, but have intervened after observing of Intranetusa's talk page a warning about edit-warring here. I went there because he asked me on my talk page to participate in a discussion on another forum, which was considering iron prodction in China.  I suspect that all the estimates for iron, about which you are fighting, are in fact based on one by Henry Cleere in about 1976.  This estimates annual output in the eastern Weald at up to 750 t pa.  This is based on actual data.  This is a good sample figure, but the more diffcult question is how to relate this sample even to Roman Britain as a whole let alone the Roman Empire as a whole.  A further comment on this will be found on my own talk page.  I do not regard the guess (I am not clear whose) that 750 t was a thord of British production as outrageous, given that I have estimated consumption in England and Wales (much the same area) as 4500 or 5000 t in the early 16th century.  I have not removed all the preceding text after editing, as i am not sure of its basis.  I leave that to others.  Peterkingiron (talk) 18:07, 1 July 2010 (UTC)

I see, thanks. I will check it out. Intranetusa (talk) 20:06, 1 July 2010 (UTC)


 * I am not an expert on the Roman economy (read a few short clips and Gibbons), but I find the figures for iron production a bit dubious to say the least. Considering that iron production in Song China, a thousand years after the Roman Empire, was just 125,000 tons, it is highly dubious that the Romans, who did not have access to cast/wrought iron/steel making and other Song technologies, could have an output 2/3 that of the Song (Even Western European iron production did not surpass this figure until six or seven centuries later). Using the production of one iron center to estimate the output of an entire empire is extremely dubious; that area may have been an extremely specialized area whose production was far higher than the empire as a whole. For example, there is a town in China that makes something like half of the world's buttons- if we were to use this figure to extrapolate button production for earth in 2008 we would end up with ridiculus numbers. Regards. Teeninvestor (talk) 13:28, 5 July 2010 (UTC)
 * My difficulty in commenting is that I have not seen what Craddock said. Cleere's estimate for the eastern Weald is based on his measurement of the size of slag heaps.  I regard the Weald as 1/3 of Britain as credible.  Unfortunately his methodology cannot be apploied in some other regions as cinders were removed for resmelting in the blast furnace era.  There was certainly a significnat Roman iron industry based on Forest of Dean ore with iron smelting cites including Ariconium and Worcester; at least that is the most probable origin of the ore used at Worcester.  Additionally, there is evidnce of smelting in other areas, but perhaps on a smaller scale.  Craddock's 82500 would seem to be on the basis that the Weald was 1/110 part of the whole Empire.  This is all based on multiplying up a single estimate.  It may even be worse that that, as Cleere may have estimated the whole Weald based on slag at one site.  On the other hand, I suspect that the Chinese estimates are also based on a very small sample.  These are probably the best estimates available, but they have to be regarded as extremely tentative.  Peterkingiron (talk) 14:45, 6 July 2010 (UTC)
 * I believe the Chinese samples were based on the iron mining tax, in which the government took a fixed percentage of the output of iron mines (say 1 or 2%), and not based on extrapolation.Teeninvestor (talk) 16:59, 13 July 2010 (UTC)

Removal of this figure
Because concerns above have not been addressed, I have removed the figures. We cannot put an iron production figure on here with just extrapolation from one already shaky estimate. That's ridiculus; you might as well extrapolate average US per capita income based on bonus estimates at Goldman Sachs.Teeninvestor (talk) 20:11, 6 August 2010 (UTC)
 * Nice try, but all three cited sources comply to WP:Verifiable. One of the sources, The Oxford Handbook of Engineering and Technology in the Classical World, published by Oxford University Press, is even the most recognized, rcently-published single volume on ancient technology in English, so try harder to discredit the number which your sinocentric bias evidently cannot stomach. Gun Powder Ma (talk) 22:26, 9 August 2010 (UTC)
 * GPM, the scholars clearly described their method and it is fallacious. If some paper claims the GDP per capita of Africa is higher than US, are you going to accept that? You are claiming Roman iron production per capita is higher than Europe in 1700, as European iron production in 1700 was lower than Song levels and their pop was about equal (and this paper claims Roman iron production per capita was higher than Song). That is completely wrong. This figures are eminently falsifiable; extraordinary claims require multiple extraoridnary sources. Do not edit war against consensus.Teeninvestor (talk) 23:15, 9 August 2010 (UTC)
 * Craddock and Sims are acknowledged international authorities on ancient metallurgy and Healy's book is among the most widely cited in this field. We both know that you know nothing about this area, so it is up to you showing why this figure should not be included in the article. I have enough of your edit-warring. Gun Powder Ma (talk) 14:14, 10 August 2010 (UTC)
 * Really, GPM? Do you know any thing about iron smelting and iron technology? This figure is implying that Roman iron per capita production was higher than Europe in 1700 and contemporary China, as well as Song China. This is contested by Wagner (2001) and Needham, which represents the standard view that Chinese and later European iron technology was far superior to Roman iron production technology; Romans had only access to the highly inefficient bloomery apparatus, while Chinese and Europe in 1700 was making cast iron and steel. So tell me GPM, how did the Romans produce more iron per capita even though their technology is 1000 years behind? If Romans were producing fifteen times as much iron per capita as Han, how come they were still using wooden farm tools while iron farm tools were standard in China?Not only that, the discussion above proves that the method used was extremely faulty; first it extrapolated iron production for one center based on slight evidence, and then used this already shaky figure to extrapolate production for the whole empire. That's like talking the CEO of GS's bonus, using it to extrapolate compensation for all of Wall street, and then using that to extrapolate GDP per capita for the U.S. completely unacceptable.Teeninvestor (talk) 14:26, 10 August 2010 (UTC)
 * Blabla. Your reasoning, if it may called any, is original research and synthesis: You cannot dump the Roman iron production figures on the basis that it cannot have been higher than Chinese or later European. So where is your source which explicitly refutes the Roman numbers? Where is your source whhich refutes these international experts? Gun Powder Ma (talk) 14:34, 10 August 2010 (UTC)
 * Did you read what I wrote? I've already cited several scholars whose research directly contradicts your forty year old estimates.Teeninvestor (talk) 14:36, 10 August 2010 (UTC)
 * Do you understand elementary Wiki guidelines such as WP:Synthesis and WP:OR? And Craddock is actually from 2008. Gun Powder Ma (talk) 14:41, 10 August 2010 (UTC)
 * Final comment: We have an overwhelming consensus of scholars saying that Chinese iron technology and European iron technology in 1700 was far ahead of Rome. You have an estimate based on an extrapolation based on an extrapolation that was first made in the 1980's which says that Romean iron per capita production is higher than both, even though academic scholars have repeatedly said that the Roman iron technology and production was far inferior. We have several editors above contesting your statement. There is no way from a technological point of view that Roman iron production per capita could be higher, from both common sense and what the scholars have said; it would be like saying that Zimbabwe has a higher car per capita production than US or China even though it has no automobile factories or capability to manufacture machine tools. Other editors above have explicitly refuted your claim, showing that Britain produced only 1/3 of what you alleged and that it played a bigger role in overall production, which would drastically decrease the iron production you allege. What else do we need? Teeninvestor (talk) 14:43, 10 August 2010 (UTC)
 * Who are these editors? Name them please, I am going to notify them of your attempt at speaking on their behalf. Gun Powder Ma (talk) 19:25, 10 August 2010 (UTC)
 * Look at their comments above, that's all I have to say. Your statistic has been attacked repeatedly by other editors besides myself, who used different sources but reached the same conclusion.Teeninvestor (talk) 19:55, 10 August 2010 (UTC)
 * None of your comparative experts are experts on ancient Rome. Also "inferior" iron technology does not mean lower production numbers - one civilisation can produce a lot of low grade iron and the other just a little high quality steel; who has the better technology? Arnoutf (talk) 14:27, 14 August 2010 (UTC)


 * Please see my comments at the foot of this page. The works cited are by leading scholars in the field of ancient metallurgy.  The methods used with different (and ones abandoned in Europe at the Renaissance, but perhaps because they were fuel-inefficient or couldnot be scaled up.  The products may have been inferior (but not necessarily so).  However, that says nothing at all about tghe sceler of output.  Historical metallurgists frequently have the problem that scholars in other fields do not understand the subject.  Peterkingiron (talk) 16:59, 14 August 2010 (UTC)

Last ruler of a unified Roman Empire
The chronology derived from Shepard (1993) in the lead of this article says that the institution of the Roman Empire began with Augustus and ended with Constantine I. This is simply false. A monumental figure he may be, Constantine was not the last man to rule both halves of the Roman Empire; that can only be said of Theodosius I. Yet there were others who ruled over all of the Roman Empire after Constantine; Julian the Apostate is an example. And for that matter, the institution of the Roman Empire did not die with Constantine. Even after the fall of the Western Roman Empire almost two centuries later, the Eastern Roman Empire not only existed, it rebounded under Justinian I. So please, someone check Shepard (1993) on this! If he truly said this about Constantine, clearly we need another reliable source for the time line.-- Pericles of Athens  Talk 22:42, 1 July 2010 (UTC)

The estimates on Roman iron output are rock-solid
There has been a recent concerted effort by two users, Intranetusa and more recently Teeninvestor, to discredit a scholarly estimate of Roman total iron output as "dubious". The employed methods have been loud clamour (always effective in misguiding other, uninformed users), unbelievable pieces of self-righteous synthesis and the usual edit-warring by Teeninvestor.

Let's take a sober look at the sources and the authors (Craddock 2008, p. 108; Sim, Ridge 2002, p. 23; Healy 1978, p. 196):

The archaeological record shows that iron was a very familiar low-cost material. Estimates have been made for iron production of 2,250 tons per annum in Roman Britain, and 82,500 tones per annum through the rest of the Empire.

Source: Craddock, Paul T. (2008): "Mining and Metallurgy", in: Oleson, John Peter (ed.): The Oxford Handbook of Engineering and Technology in the Classical World, Oxford University Press, ISBN 978-0-19-518731-1


 * The author Paul Craddock of the British Museum is an international expert on ancient metallurgy (credentials),
 * The editor John Peter Oleson is one of the most renowned scholars specialized on ancient technology (personal website)
 * Oleson's The Oxford Handbook of Engineering and Technology in the Classical World, published by the Oxford University Press, is the most important work on ancient technology of recent years: see reviews on Amazon by the peer-reviewed scholarly journals American Journal of Archaeology, Journal of Roman Archaeology and Technology and Culture.

In addition, this number of around 80,000 t per annum is supported by further experts:
 * Sim, David; Ridge, Isabel (2002): Iron for the Eagles. The Iron Industry of Roman Britain, Tempus, Stroud, Gloucestershire, ISBN 0-7524-1900-5. "Dr David Sim is a skilled professional blacksmith and archaeologist who did his PhD at Reading on Roman iron working (credentials on page 3)."
 * Healy, John F. (1978): Mining and Metallurgy in the Greek and Roman World, Thames and Hudson, London, ISBN 0-500-40035-0. This is a fundamental work in the field and listed as such in the bibliography of the current Oxford Roman Economy Project (page 1).

Include: All this shows that the sources giving the estimates absolutely comply to WP:Reliable and that attempts at discrediting the numbers are completely unfounded. The bitshould be included again and the edit-warring over it has to stop immediately and permanently. Gun Powder Ma (talk) 20:08, 10 August 2010 (UTC)


 * Oppose:You haven't addressed what me and intranetusa have pointed out to you repeatedly. Facts in wikipedia not only have to be verifiable, they have to be the mainstream view and not that of a tiny minority. If a renowned scholar suddenly declared that Africa's GDP per capita was higher than US would you include this claim? I can also find 6 "scholars" that would otherwise comply with WP:RS and who thought the Soviet Union collapsed because it wasn't Marxist enough. Should we include that? Your estimates are all based on an old estimate in 1976 which first extrapolated the iron production of one center based on very shaky evidence and than extrapolated iron production for the ENTIRE empire based on that figure. This method is equivalent to talking the CEO of Goldman Sach's bonus, using it to extrapolate compensation for all of Wall street, and then using that to extrapolate GDP per capita for the U.S. Besides this, me and intranetusa have already pointed out major errors with your approach using academic sources. Firstly, an overwhelming consensus of scholars saying that Chinese iron technology and European iron technology in 1700 was far ahead of Rome. Your estimate says Roman iron production was far higher than both, even though almost all academic scholars have repeatedly said that the Roman iron technology and production was far inferior; Rome couldn't make cast iron, it couldn't make steel, and its iron industry was universally regarded as far less developed than Han China (per Wagner and other sources) and 1700 Europe. According to your estimate, Roman iron production would be 15 times more per capita than contemporary China, yet Han China had widespread use of iron farm tools and Rome did not; unless the Roman government was building huge iron coloseums every two years, wheres the iron? To believe your statement would be like saying Iraq's automobile per capita is 15 times that of the US even though there is 1 automobile factory in the whole country. Secondly, Intranetusa has shown that modern estimates believe that Britain produced only 1/3 of what you alleged and that it played a bigger role in overall production, which would drastically decrease the iron production you allege. Assuming that Britain's iron production was 750 tons and it played a role in iron production 2x more than previous, this would scale your production figures by a factor of 6. What else do we need?Teeninvestor (talk) 20:21, 10 August 2010 (UTC)
 * You are again totally evading my point (WP:IDIDNTHEARTHAT) and your attempt at construing a connection between the Roman and Chinese output by which you then refute the Roman numbers is WP:Synthesis and WP:OR at its core, but I'll leave it to other users telling you this. Suffice to say that even your OR and SYN discussion is ridden with irremediable errors. Gun Powder Ma (talk) 20:30, 10 August 2010 (UTC)
 * GPM, you can evade my point all you want, but the final point is clear;

Firstly, There is overwhelming consensus among scholars that Rome's iron industry and output was far below that of Han/Song China and 1700 Europe's, and thus support a much lower figure than you are letting on, since you claim that Rome's iron industry had a higher iron output than those nations. Secondly, Intranetusa has shown that modern estimates utterly contradict the estimates you have put. Thirdly, the source you cited has a method which all of us here (except you) agree is utterly wretched and unreliable. The method is equivalent to talking the CEO of Goldman Sach's bonus, using it to extrapolate compensation for all of Wall street, and then using that to extrapolate GDP per capita for the U.S. Fourth, If something is utterly wrong and a complete minority view, we're not gonna put it in the article (See WP:UNDUE). If some paper says that Africa has a higher GDP per capita than US, it's not going in even if its a harvard professor saying that.Teeninvestor (talk) 20:35, 10 August 2010 (UTC)


 * This strikes me as a quite peculiar debate. Gun Powder Ma argues his case on sources. Teeninvestor, you are of course free to dispute the figures given by Gun Powder Ma's sources; I assume your disagreement is founded on your readings of other sources. If so, I suggest you provide them in this section. When (or if) you can offer alternative figures from equally relevant and specialised sources, they can be used in the article alongside those provided by Gun Powder Ma. The figures provided and cited by Gun Powder Ma should in any event remain in the article – no matter how implausible you find them in light of your own readings of other estimates for iron production in other empires, epochs and economies, or conclusions you have drawn from your readings or your own writings in wikipedia or other articles on the same. Personal opinion and original research cannot be used to counter or remove accurately cited content. Haploidavey (talk) 22:26, 10 August 2010 (UTC)
 * My position is not my own "original research" but the mainstream academic position. First of all, Intranetusa has above suggested that recent modern estimates disputes the source that GPM uses and holds that the production at the site alleged was barely one third of what was previously alleged and that Roman britain produced a far larger share of iron output than previously imagined, which would decrease the iron production alleged by a factor of six. Secondly, my comparison of the Roman and Han iron industries is also supported by the vast majority of schoolars; See for example this quote about Wagner, Donald B. (2001b)'s book, The State and the Iron Industry in Han China:

"In his conclusion, Wagner offers an interesting reflection. The iron-production technology of the Roman world resulted in small-scale, localized industry, which was of no political or financial interest to the state. In China, a more advanced technology, which gave significant economies of scale, encouraged a powerful state to take a direct interest in iron production and thereby further increase its economic and political power" Not only that, but sinologists Joseph Needham and Robert Temple have also supported this position. It is well known that China's iron technology and production (and that of Europe in 1700) was far ahead of Rome; this is a mainstream position. In addition, the way this estimate was arrived at is completely unreliable. It is based on an extrapolation for empire-wide iron production from an extrapolation for an entire iron center from one site.Teeninvestor (talk) 23:09, 10 August 2010 (UTC)

Han Chinese iron output
What Teen. conveniently forgets to mention is that the same Donald Wagner gives a precise number for contemporaneous Han Chinese iron output:


 * Wagner, Donald B. (2001): "The Administration of the Iron Industry in Eleventh-Century China", Journal of the Economic and Social History of the Orient, Vol. 44, No. 2, pp. 175–197 (191):
 * "For the Han period I have suggested elsewhere that iron production might have been on the order of 0.1 kg per capita per year (Wagner 2001a: 73)."


 * Wagner, Donald B. (2001), The State and the Iron Industry in Han China, Copenhagen: Nordic Institute of Asian Studies Publishing, ISBN 8787062836, p. 73:
 * "If we were to assume an average annual production of 100 tonnes per Iron Office, then total annual legal production in the han Empire as a whole in AD 2 would have been about 5,000 tonnes, or about 0.1 kg per person."

Hence we have reliable estimates on both Roman and Chinese iron output. What I find curious is that Teen. is actually absolutely aware of this figure which he has quoted himself in Economic history of China (pre-1911). Still, he acts as if Han production dwarfed the Roman output. I wonder who he is trying to fool? Gun Powder Ma (talk) 23:40, 10 August 2010 (UTC)
 * He suggested that is the output of the state iron monopoly when Emperor Wu of Han nationalized the iron industry in 117 BCE against the will of the Confucian scholars. This figure does not take into account illegal production (note he said legal), the possibility each iron office produced more than 100 tons, and the vast expansion of the iron industry after it was privatized during the Eastern Han. Conveniently Wagner argues against your thesis that Roman iron output was much larger than Han China; so if we accept Wagner's view which is the consensus (and theres no reason not to), Roman iron production would be much less than Han China. If we take the late Tang figure of 25,000 tons as equal to the iron industry of Eastern Han (due to the intermediate disruptions, this is reasonable) that would give a figure of 0.5 kg per capita. Teeninvestor (talk) 23:58, 10 August 2010 (UTC)
 * I'm sorry, but I'd advise Gun Powder Ma not to be drawn into competitive comparisons. It might well be true (or debatable, or false) that China's iron industry was technologically well in advance of/ more productive than/ State invested than contemporary Rome's. In the end, what has any of that to do with Roman metallurgy? Haploidavey (talk) 00:25, 11 August 2010 (UTC)
 * Nothing, GPM's figure alleges that Roman iron production per capita was 15 times that of Han China and higher than Europe in 1700, Song China and slightly below Ming China. As I've shown above, the mainstream consensus violently disputes this figures.Teeninvestor (talk) 00:51, 11 August 2010 (UTC)
 * Nothing. And that has exactly been my point all the way: To dismiss the Roman numbers, Teen. uses a comparison with the contemporary Han Chinese production. But not only is this comparison a thorough piece of synthesis, as I argued above, it is also factually wrong even if we would allow such a comparison to be made, since the Han output was actually far lower than the Roman one. So, on either account, Teen. is wrong. Gun Powder Ma (talk) 00:53, 11 August 2010 (UTC)
 * My logic is very simple; Mainstream consensus (Wagner, Needham et al) says Han, Song and 1700 Europe's iron production is higher than Rome- GPM says the opposite- mainstream consensus is the main view, should be included. It's that simple. The empirical evidence cited by these scholars, such as the much more widespread use of iron farming tools, cooking pots, etc, among Han China than Rome puts another nail in the coffin. Unless all that iron was produced and then dumped in the sea to stimulate the economy your thesis doesn't hold up. And that's assuming that your estimate holds up under modern estimates; as intranetusa has shown above, Jeremy Hodgeskins, a scholar who actually studied the weald iron site, has suggested a far lower annual iron production figure.Teeninvestor (talk) 00:56, 11 August 2010 (UTC)


 * Teeninvestor, could you please cite and quantify the "far lower annual production figure"? Regarding the remainder of your post: GPM seems to have produced no theses, developments or syntheses from sources as far as I can see; nor should any of us when it comes to provision of article content. And we don't develop article content and conclusions from disparate sources using our editorial logic; article content summarises the reasoning of our sources. Nothing more. And I can find nothing in the article - information, statements or assertions, cited or otherwise - to justify the insistent debate; unless we're merely looking at the "if ever" of the previous version? If so, I have to say I find "if ever" a little unsatisfactorily; vague, more than anything. Is it the exact phrase used by the source? Haploidavey (talk) 01:28, 11 August 2010 (UTC)
 * No, that was just an attempt of mine to properly reflect the lead and copper figures which remained unsurpassed until the IR. Scratch that if you like. Gun Powder Ma (talk) 01:37, 11 August 2010 (UTC)
 * That helps; its current form allows for a combination of some lower and some higher figures. Haploidavey (talk) 02:18, 11 August 2010 (UTC)
 * So we're going to throw academic mainstream consensus to the winds? Let me repeat again. 1. GPM's iron production figure has been disputed by people who worked at the site. 2. GPM's iron production figures have been disputed by the mainstream consensus that Roman iron production per capita was lower than Han, Song or 1700 Europe, while this figure contradicts that. How more clear can we get? And in case you think I'm some sort of synthesis/OR warrior, I'm not; Every word I've said here has been said by some academic.Teeninvestor (talk) 01:36, 11 August 2010 (UTC)
 * It's about time you provide proof for 1. and 2. or give up your edit-warring at last. Gun Powder Ma (talk) 01:40, 11 August 2010 (UTC)
 * Wagner is proof for 2. and Intranetusa above has the source for 1; but he seems to be unavailable on wikipedia right now. On the article mining in Roman Britain, however, it says 2/3 of iron mines were in weald, not 1/3 as you allege; this would already cut your figure in half. Teeninvestor (talk) 01:42, 11 August 2010 (UTC)
 * Teen, no one has suggested we throw academic consensus to the winds. You'll make little headway by restating yourself. You need to justify your statements with sources and citation. And in case you're wondering, I don't doubt the sincerity of your case; but you only undermine it if you omit hard evidence on which it's based. (Inserted comment: by the way, you know we don't cite wikipedia articles. We can use their sources, if they're reliable. So never mind what the mining article says).


 * On a change of tack; I see Peter Lewis has edited this article. He might be able and willing to represent current scholarship on the matter, and I see GPM has invited him to discussion. I suggest its editors hold off until then. Haploidavey (talk) 02:00, 11 August 2010 (UTC)

Agreed, though I believe I have cited books such as Wagner 2001 which has made the exact same case I am making. And yes, I acknowledge I have no hard evidence for 1), but we'll wait for other editors to see if they have anything to say.Teeninvestor (talk) 02:01, 11 August 2010 (UTC)
 * 1. I've read the sources Intranetusa refers to and they are completely irrelevant as neither gives an estimate of the total iron output of the Imperium, but only of a tiny area, the English weald! Intranetusa's attempts to step up these numbers to the whole empire are then classical OR stuff and can be totally discounted.
 * 2. Repeat an untrue comment often enough and there will be others who fall for it. I gave you Donald Wagner's estimate for Han Chinese production which are far below the Roman level. Therefore, you cannot use him for questioning the Roman output (which, in the absence of a clear reference by Wagner to the 80,000 tons, is anyway synthesis and original research). Gun Powder Ma (talk) 08:57, 11 August 2010 (UTC)
 * Wagner says Roman iron production is below Chinese. Chinese iron production is estimated at 5,000 to 25,000 tons or so. You say the reverse. Conclusion; you contradict Wagner and need to be removed.Teeninvestor (talk) 13:13, 11 August 2010 (UTC)

Clarification requested
I'm a little confused by the debate here. Teeninvestor's reasons for questioning the figure seem sound. However, if Gun Powder Ma's figures are incorrect (even though his sources look good), where are the figures to replace them? I don't see any figures on iron in the table under "Output."

Somewhere above it was claimed that the Romans didn't use iron for farm implements. This is quite untrue, so perhaps I misunderstood the statement. The Romans had a great number of iron implements for various purposes, including agriculture (not to mention surgical instruments, which might be made of "steel"). According to Varro, De lingua latina 5.31, the ploughshare (vomer) was made of iron, and the agricultural treatises of Varro, Cato, and Columella mention several iron tools, or tools that have iron heads or blades. Iron agricultural implements were found at Pompeii. ANRW: "Archaeology has accumulated large numbers of iron implements." See also K.D White's Farm Equipment of the Roman World, on uses of iron.

The Romans used an awful lot of iron, in fact. Iron nails are abundant in the archaeological record, and were used also for shipbuilding. Smiths, farriers, masons, carpenters, shipbuilders, surveyers, these all used a wide range of iron tools (hammers, saws, anvils, axes, chisels, tongs, compasses), not to mention the legionaries' mattocks, shovels, weapons, and armor. Linchpins, terrets, horse-bits and harness fittings, braziers, andirons (fire-dogs), lampstands, bucket handles and hoops, shears, door hinges and handles, chains, ladles, knives, all manner of everyday objects might be found in iron. These are found throughout the Empire, a very large geographical area, so the demand would've been great. (Gaul may be a bit underrepresented in the article: Celtic metallurgy is a major influence in the development of Roman metal technologies, particularly for weaponry and armor, during the Republic; in the early Empire, Gaul was known for technological innovation in farm equipment, such as harvesters.) As for comparisons with China, these seem to be getting in the way, and are not really decisive in determining Roman output, for which one can rely only on numbers published in the scholarship. Cynwolfe (talk) 06:55, 11 August 2010 (UTC)
 * In what way do you think are Teeninvestors' reasons "sound"? I don't even detect a trace element of a plausible and consistent argument. Gun Powder Ma (talk) 08:57, 11 August 2010 (UTC)
 * I see Gun Powder Ma's case for including his statistic quite clearly; it's well sourced and well presented above. Once I started reviewing the number of iron objects in use by the "Romans" (more on that in a minute), the figure didn't strike me as too high. A ton of iron is really not that much iron in volume, when you consider the many uses it was put to throughout the Empire (hence my point about Gaul). My question was why GPM's figure wasn't still in the article. (Am I just overlooking something?) In trying to see it from Teeninvestor's perspective, I was imagining coming to the article and saying, 'hm, based on what I know about Han output, that figure seems way too high.' His reasons for raising the question initially seem plausible, even sound.


 * What I'm missing, however, is what alternative figures Teen's providing. If he has alternative figures, why aren't they in the article? The figure could be given as a range: "XX,XXX tons(footnote here to cite Teen's source for his stat)–84,750 tons(footnote on GPM's source)." Even if Teen has a specific figure to provide, it doesn't exclude GPM's figure, which is well documented. If Teen has no figures to provide, I'm at a loss to understand how GPM's could be deleted; at most, the figure could be qualified as "up to 84,750 tons." The note would then cite GPM's sources, but with a qualifying statement from Teen's source that So-and-So has argued that Han output exceed that of Rome but has been estimated at a per capita figure that is 15 times less. So I agree with Haploidavey. My position is in keeping with WP policy on neutrality: to describe the scholarship in a way that explains the information provided. Deletion is never the answer on contentious subjects. I haven't taken time to search through the edit history to find out what the chart used to say about iron, and feel that it should be restored and tagged as "dubious" for now, until outside editors who know the scholarship better can comment. The absence of the data makes it harder to discuss. Cynwolfe (talk) 13:57, 11 August 2010 (UTC)
 * Other users have used sources to contest GPM's estimates, but they are unavilable right now. What makes this even more disturbing is that GPM has been revealed to be editing this article to support his side of the argument on a Chinese history forum, as shown above.Teeninvestor (talk) 14:00, 11 August 2010 (UTC)
 * I don't doubt that GPM's figures can be contested; I know of few figures of this kind from antiquity that couldn't be contested. The point is, if he gives a figure and ascribes it to a valid source, that passes verifiability. You can qualify it with an explanation, or add your own figures to create a bracketed range. But I see no reason to delete his figure. Tagging that cell in the chart with "dubious" would direct interested parties to this talk page, which might provide additional perspectives and sources. What GPM does or says elsewhere is of no relevance to the validity of his sources. I can see, Teen, why you looked at this number and said "huh?" I take no position on what the correct figure should be. GPM feels confident about his data and his methodology in obtaining it, and I appreciate that it might be difficult for him to see a "dubious" attached to it, or to see it modified as I've suggested. But if the scholarship is noted and described accurately, that should save both of you a lot of pointless head-butting time that you could spend on making other contributions you might enjoy more. The Cambridge economic history of the Greco-Roman world might have something; limited preview here. Cynwolfe (talk) 14:45, 11 August 2010 (UTC)
 * As been now widely realized, Teeninvestor has no alternative estimate on Roman iron production to offer. His rejection of the 80,000 t rests solely on the notion that an expert on Chinese metallurgy, Donald Wagner, rates Han China to be more productive than Ancient Rome. However, Wagner does only offer an estimate of Han Chinese total iron output, but not for Ancient Rome (where he has no actual expertise), and since his number of 5,000 t per annum is far below the estimate of other authors for the Roman Empire, Teeninvestor has actually presented no argument at all to dispute the Roman figure. I hope this has has now become clear enough. Gun Powder Ma (talk) 18:43, 12 August 2010 (UTC)

Roman estimate is conservative
I'll briefly give some background info on how the Roman figure of 80,000 t was arrived at to show that it is actually a quite conservative estimate: iron consumption in Roman Britain has been estimated at 1.5 kg per capita, not more than a third of the 4.5/head estimated for 17th century England. Projected on the population of Roman Britain of 1.5 million this amounts to 2,250 t for this Roman province or the annual output of no more than 500 furnaces of the Ashwicken type. This figure is then stepped up to around 80,000 for the entire empire, given a conservative estimate of 55 Mio. Since more recent population estimates for Roman Britain assume 3.6 mio and for the whole Imperium 60-70 up to 100 mio., the figure of 80,000 is actually pretty much on the low side. Gun Powder Ma (talk) 18:43, 12 August 2010 (UTC)


 * If you read what I said carefully, GPM, you'll see that I'm asking you to put iron back in the table. I don't see any reason to delete it. I do see why Teeninvestor questioned it based on his own knowledge-framework, and that's what a talk page is for. If Teen wishes to tag it as "dubious" till more editors weigh in, or to revise it along the lines I suggest above, he's free to do so. However, there is no consensus on this page that I see for deleting your contribution, even if some qualifier such as up to 80,000 must be appended for the sake of including it at all. You can't just leave iron out of a table on metals output. I see your sourcing of this as impeccable; other figures may exist, and they could be bracketed with yours. But under no circumstances short of Craddock recanting do I think the figure you got from him should be excluded. I question stuff in articles all the time from within my own knowledge basis; I discuss it on talk pages, or rewrite it, or replace it with text I hope to be accurate, or as a last resort delete on the basis of sources that directly contradict the statement. In view of the ubiquity of iron in the everyday lives of Romans at almost any historical period, your tonnage doesn't seem that great to me. A ton of iron is not so large in volume as one might think. (Teen may not be as aware of that ubiquity, since I believe it was he who thought the Romans had no iron farm implements, when in fact the archaeological record is thick with them, as are the extant agricultural treatises; I actually sit around and read Varro, so this seems familiar to me, but it may not to him.) Clear? Your work is good. Please put iron back in the table. Teen should tag, not delete, because this is an ongoing discussion of sourced material, and the presence of the material in the article may attract interested but dispassionate parties. This is of course only the opinion of a single editor. Cynwolfe (talk) 19:17, 12 August 2010 (UTC)
 * I did read your contributions carefully. I just wanted to be careful to secure consensus, since there has been recently much reverting on the figures. Since both User:Haploidavey and you find the figure unproblematic, I'll restore the bit. Regards Gun Powder Ma (talk) 19:33, 12 August 2010 (UTC)
 * (edit conflict)You certainly have consensus; I entirely go along with Cynwolfe. As you say, the per capita estimates you've provided are based on the lower, if not the lowest end of range. That's a pretty safe position to take; perhaps even overcautious, so nothing to defend here; put the figures in the tables, where they belong. If I find other, equally well-sourced figures (well, who knows?), I'll post them here on the talk-page for your disposal, not in the article. And I'll put this on my watchlist. Just as an aside, never mind hypothetical dumping at sea (somewhere way above): no iron-impoverished culture casually abandons the several tons of used and unused iron nails found at Inchtuthill. Haploidavey (talk) 19:42, 12 August 2010 (UTC)
 * Gun Powder Ma is a careful reader, and I didn't mean to imply otherwise. I thought perhaps my recently acquired diplomatic cloak, which I wear uncomfortably, obscured my meaning. For the sake of neutrality, let me quibble slightly: I don't know whether the figure is unproblematic, because I haven't researched it; I find its inclusion on the table unproblematic, because it's properly sourced. Thanks to Haploidavey for mentioning Inchtuthill, which I'd been snapping my fingers to try to remember. Several tons of iron nails in one spot, you say? Cynwolfe (talk) 20:55, 12 August 2010 (UTC)
 * Thereabouts. For some reason, I remember it as "seven to ten tons". Damned if I can find a reference, but the main excavations were through the 1950's; the nails and other iron utensils had been concealed through burial after the fort's dismantling and abandonment. I suppose the smaller advance forts along the Gask Ridge line would also have been slighted and their iron could have contributed to the Inchtuthill deposit. But that's just my guess. Haploidavey (talk) 21:14, 12 August 2010 (UTC)
 * According to this site on the nails (not sufficient for scholarly citation, but informative): There were over threequarters of a million of them, ranging in size from 2 inches up to 16 inches. They weighed nearly seven tons. There's an article Inchtuthill. An interesting digression. Cynwolfe (talk) 21:21, 12 August 2010 (UTC)

Comment
I have returned to this subject becasue I was asked to, though it is not the period of the iron industry that I normally deal with:
 * Any figure is an estimate. It is not a fact, and its accuracy will depend on the methodolgy used in creating the estimate.
 * The figure for Roman Britain is probably based on an estimate by Henry Cleere of the size of certain slag heaps in the Weald, multiplied up by a perceived number of similar sites, and again by from Britain to the whole Empire.
 * This is a valid means of making an estimate. The question is whether the multipliers were right.
 * As far as I can see (though, as stated, this is not my area of expertise) all the estimates go back to this one. It does not particualrly matter whether it is quoted by one scholar or 100.
 * Cleere is certainly WP;RS. Quotation of him by subsequent scholars strengthens that status.  Other scholars are of course entitled to question that, but for a WP editor to do so in the way he edits an article is a variety of WP:OR or even WP:V.
 * If there are rival views, it is legitimate to set them all out, ideally with an assessment of which is most probable (but such a synthesis may also come close to WP:OR).
 * Editors, who persist in putting forward their own WP:OR or who delete information based on WP:RS without relying (and citing) a more reliable source, are liable to adminstrative action withdrawing theri editing rights. Peterkingiron (talk) 21:13, 13 August 2010 (UTC)
 * I am glad to see that I agree on all important points (the figure, though, does not originally come from Cleere). Regards Gun Powder Ma (talk) 21:32, 13 August 2010 (UTC)
 * Perhaps I should ahve said "based on". The estimate of 750 for the Weald is Cleere's.  I do not know who multiplied it up.  Peterkingiron (talk) 23:44, 13 August 2010 (UTC)
 * I see that the estimate is allegedly based on population, but the estimated consumption per head must have a source. I suspect that investigation of the sources will show that it does go back to Cleere's estimate for the Weald, estimated to be one-third of British output (a credible guess - but still a guess).  This has been divided by population (again estimated and perhaps litlte more than a guess, perhaps based on later periods) to give an estimated consumption per head.  This has then been multiplied up using the estimated populatiopn of the Empire.  Whether or not the estimate is right, it does come from WP:RS and should be able to appear in the article.  Any statements against this appear merely to be the opinion of the editor (which WP calls WP:OR).  If there are rival views, the appropriate course would be to have an article section on iron output and to discuss the rival views in it.  I do not have access to the works cited and thus cannot do this myself, but perhaps some of the warring editors can cooperate in producing such a section.  Peterkingiron (talk) 17:09, 14 August 2010 (UTC)