Talk:Primary source

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject History (Rated Start-class, High-importance)
WikiProject icon This article is within the scope of WikiProject History, a collaborative effort to improve the coverage of the subject of History on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
Start-Class article Start  This article has been rated as Start-Class on the project's quality scale.
 High  This article has been rated as High-importance on the project's importance scale.
WikiProject Libraries (Rated Start-class, Mid-importance)
WikiProject icon This article is within the scope of WikiProject Libraries, a collaborative effort to improve the coverage of Libraries on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
Start-Class article Start  This article has been rated as Start-Class on the project's quality scale.
 Mid  This article has been rated as Mid-importance on the project's importance scale.

Incorrect definitions[edit]

Comment by Birger Hjørland (Marts 30, 2005)

The way in which primary source, secondary source and tertiary source has been defined in Wikipedia is not in accordance with the way in which is has been defined by UNISIST in 1971 and taken over, by among others, the present writer. (See Fjordback Søndergaard; Andersen & Hjørland, 2003).

Also, the normal language in Library and information science is to speak of bibliografies and the like as secondary sources (or secundary literature).

There are variations in the use of these concepts between the humanities and the sciences, however, if a general terminology should be established, we recommend the UNISIST terminology.

Fjordback Søndergaard, T.; Andersen, J. & Hjørland, B. (2003). Documents and the communication of scientific and scholarly information. Revising and updating the UNISIST model. Journal of Documentation, 59(3), 278-320. Available at: <<Link is broken - is this it? —Preceding unsigned comment added by (talk) 17:57, 9 November 2009 (UTC)


This article needs a rewrite along the lines of the recent changes to Secondary source. It needs more coverage of non-event-oriented history, and more on the strengths and weaknesses of primary sources. Also, the rather patronizing comment that one needs an advanced degree to use primary sources should go. It's true that interpretation of some primary sources calls for advanced training; it's also true that grade-school children read primary sources all the time without knowing it. Wikipedia can do better. I might do this in a week or two, but I wouldn't object to someone else taking up the task. Katherine Tredwell 01:11, 15 May 2006 (UTC)

This was completely rewritten a while ago, and I think the previous version was much better. You might consider taking a look at that. (I only made a very minor edit to the old version, so I don't have any personal pride that was injured through the rewrite.) up+land 03:37, 15 May 2006 (UTC)
Fact is there are several hundred graduate programs in history in the US alone--and many more around the world--that focus on how to interpret primary sources. To say anybody can do it is meaningless. Yes elementary school students can look through telescopes and see the rings of Saturn but that does not make them astronomers who can write the Wiki article on Saturn. Rjensen 11:04, 15 May 2006 (UTC)

I am not entirely unacquainted with the existence of graduate programs in history. Sure, some texts are difficult, but the article goes too far. "In general, primary sources are difficult to use and advanced college or postgraduate training is normally required..." is simply untrue. A university freshman or bright high-school student can be given a research project and taught to ask questions about a source and its context. In the US at least, we don't do it, so nobody knows how to do it even with the simplest sources. Katherine Tredwell 18:43, 15 May 2006 (UTC)

The freshmen are given a preassigned text and carefully taught how to handle it. They are NOT allowed to pick any primary source they want and plunge in. They do not learn general skills for dealing with primary sources--that is only taught in advanced college or grad school courses. Keep in mind that Wiki in historical articles strongly frowns on original research with primary sources. Rjensen 19:31, 15 May 2006 (UTC)

I'm not claiming that anyone can do any level of historical research on anything without any training, just that the current article overstates the case. Wikipedia frowns on original research in articles on any subject because it's an encyclopedia. Most generalist encyclopedias these days contain little or no original research. I won't discuss the matter further right now; I have better ways to spend my time. Whenever the article is revised we can take it up again. Katherine Tredwell 19:46, 15 May 2006 (UTC)

I work with students in grades 4-12 on a program that is mostly devoted to students at those grades taking on research projects and supporting their thesis with primary and secondary sources. To say you need graduate training is completely inaccurate. —Preceding unsigned comment added by (talk) 20:10, 30 June 2008 (UTC)

I don't think the article says you need graduate training - "The interpretation of primary texts is typically taught as part of an advanced college or postgraduate history course, however advanced self-study or informal training is also possible." There are a few primary sources that are widely used below post grad - census returns for example. Nonetheless, AFAIK, formal training in "the interpretation of primary texts" is unusual in schools. Rjm at sleepers (talk) 14:15, 1 July 2008 (UTC)
Hmm, I think you're right that it does overstate the case. But I do think that it should be emphasized that most historians feel that some training is usually needed, or, at least, that the interpretation of primary sources is often quite difficult and requires a good deal of contextual knowledge (and one can interpret it as they wish -- perhaps it is just job security, after all!). --Fastfission 21:36, 15 May 2006 (UTC)
This is an age of specialization. Most people need special training to diagnose a medical condition, except that 200+ million people diagnose and treat themselves every day. The biggest problem in teaching historiography is that students come into the course with a lot of preconceptions about "truth" and "facts" that are seriously wrong-headed. Students assume there is one "truth" and the idea that historians have multiple competing explanations for the same events is a tough one. Happily these skills are not really needed on Wiki since we don't allow much original research on history. Rjensen 22:41, 15 May 2006 (UTC)
I have attempted to re-write part of this section. I was not convinced about some of the assertions. For example, some accurate histories can be written without use of a primary source. In any case, interpretation of modern primary sources is less difficult than earlier sources, although some earlier sources (eg parish registers) are relatively straight forward. Schools tend to avoid using difficult primary sources, but amateur historians may have no choice; they cannot wait until an appropriately trained professional takes an interest. Rjm at sleepers 10:20, 9 March 2007 (UTC)

Hello. primary source is by definition more authoritative than secondary source. I think there is a bit of confusion here. Yes, the seconary source could mean academic reseach. It could also refer to reference to contemporary social commentaries or journalism. Yes, one single primarly source is not authoritative. This miss that point. To establish a historical fact, primarly source have to be corraborated. You don't built historical fact from hearsay (i.e secondary source). Vapour

It doesn't miss the point to think about how to regard primary sources on the individual level, as well as the aggregate. On the individual level, many well-researched secondary sources are more authoritative than some primary sources. Katherine Tredwell 23:23, 25 July 2006 (UTC)
A primary source is by no means more authoritative. "Primary" refers only to the way it is used, not to its veracity, its applicability, or its usefulness in understanding anything at all. Secondary sources are not "hearsay" at all (perhaps you mean something else by that than the literal definition)—in fact many primary sources are hearsay in the literal definition (Joe Doe writes in his diary that he heard Jane Doe say something). --Fastfission 23:29, 25 July 2006 (UTC)

Is the Bible a source?[edit]

There are many articles purporting to be History which are just Bible stories. We don't even know who wrote most of the bible or its provenance. If I ask for better sources I am ignored. If I delete the bible stories, I get blocked without any stated reason. Don't know why I bother. Fourtildas 06:11, 18 September 2006 (UTC)

To answer the question in your heading, yes, the Hebrew Bible and New Testament are both primary sources for ancient history, and the many other "bibles" out there can also be primary sources. They should be subjected to the same scrutiny as any other primary sources. What I think of as "Bible stories" are relevant to encyclopedias; they might be referred to in literary sources, for example. Instead of deleting them, just be sure the source is made clear in the article; e.g., if an article on an ancient war or city is based entirely on a biblical book, make sure that is mentioned in Wikipedia. Katherine Tredwell 00:31, 19 September 2006 (UTC)
Well it it already is mentioned. I could replace the article with one sourced to the Scientology bible(or whatever they call it). At least we have a pretty good idea who wrote that. But I suspect you would use your Jimbo-given admin powers to revert it to the version based on the Judaeo-Christian story, without bothering with any justification or explanation.
All this could be avoided if we adopted the standards of peer-reviewed academic journals: If nobody there supports the flat earth or the earth centred universe or the parting of the red sea or the burning bush in the peer reviewed literature then these ideas are toast. But these ideas still live on in WP as long as "some people believe" in them.

Papers in Scholarly/Academic Journals...[edit]

... should often be classified as primary sources, IMHO. The classic case is the publication of results of experiments or fieldwork, especially when statistical or other forensic methodologies are involved. That is, the paper is not establishing facts, it is reporting evidence of some kind, all the more so when the research was motivated by some as yet unresolved issue. The basic problem here is the misuse of primary sources, already mentioned in WP policy. The paper being published in a reputable journal may meet the minimal requirements of a reliable source, but this does not automatically make the paper a secondary source suitable for citation in support of supposedly established facts. Put another way, the paper is original research, and its presence in the journal only makes it a primary source from a reliable source.

Failing to make this clear allows fringecreep, where an article on a controversy gets swamped by all sorts of tangential references, defended on the grounds that they are reliably sourced! rudra 22:53, 10 March 2007 (UTC)

Currently this article is mainly about the use of primary sources by historians - presumably because that is the main interest of those who have edited it. It is addressed to the casual Wikipedia user. I don't believe it is intended as a contribution to a debate on WP policy. It is extremely rare for a paper in a scholarly or academic journal by a historian to be a primary source. There are a few exceptions - for example a scholarly edition of a primary text - where some of the content can legitimately be treated as primary. Most historians would treat the Philimore edition of Domesday as primary, particularly as access to the original is not permitted. None the less, even in such a scholarly edition, interpretation by the author is secondary. Scientists (of course) are not concerned about primary vs secondary sources, nor about verifiability as defined by WP. Their touchstone is whether a result can be confirmed by repeating the experiment. Rjm at sleepers 07:44, 16 March 2007 (UTC)


It is hard to tell if anything in this article is sourced, because of the almost complete lack of footnotes.--Sefringle 03:33, 16 March 2007 (UTC)

Following your comments, the article now has 9 footnotes as well as references and external links. As far as I can see, the only thing controversial on the page is what level of training is needed to use primary sources. If you think it needs further citations, please note the specific assertions which you regard as unsourced. Rjm at sleepers 07:27, 16 March 2007 (UTC)


How do people feel about:

"Although virtually every type of document that has been used by historians as a primary source has been forged at one time or another, primary sources in a reputable record office are likely to be authentic".

Rjm at sleepers 07:34, 23 March 2007 (UTC)

What is the meaning and purpose--to warn people about what? Online forgeries? There are very few forgeries I can think of in last 100+ years that made any difference in historiography. How about this: "Historians dealing with recent centuries rarely encounter forgeries of any importance." Rjensen 08:29, 23 March 2007 (UTC)
The purpose I had in mind was to indicate to amateur family and local historians (probably the majority readers of this article) that documents in a county record office or in the National Archives are likely to be authentic. I'm happy with your wording, but I'll wait for other comments before putting it in. Rjm at sleepers 09:17, 23 March 2007 (UTC)
I think a positive statement of reassurance makes more sense. It's very hard and rare to forge a historical document (usually it's to "authenticate" a forged art work or swindle.) . Rjensen 10:56, 23 March 2007 (UTC)
Historians dealing with earlier centuries also rarely encounter major forgeries. Unless one counts pseudonymous works, most of which may not be "forgeries" in the modern sense. Katherine Tredwell 00:39, 27 March 2007 (UTC)

@ Rjm at Sleepers: The source you're using (Everyone has Roots) for the mention of Edward Dering's forgery, do you own it or have easy access to it? I can find no mention of him forging "monumental brasses" anywhere, but plenty of evidence that he forged the so-called Dering roll to enhance his own ancestry. --Labeet (talk) 10:14, 23 September 2008 (UTC)

Page 104: "The Kentish antiquary Sir Edward Deering who invented a Saxon descent for himself in the seventeenth century went to the lengths of having brasses engraved for his pseudo ancestors and placed in Pluckley church." Camp gives as his source Peerage and Pedigree by J H Round (pages 113-116). There is a little more (apparently also from Round) here - [1]. This appears to be a direct quote from Wagner's English Genealogy. Dering apparently also had false arms carved on the font - [2]. Rjm at sleepers (talk) 06:56, 24 September 2008 (UTC)

@ Rjm at Sleepers: According to Wagner's book (as confirmed by History Centre Surrey in an e-mail) it does say that he forged brasses at Pluckley Church. But I think that a reference to his main forgery, namely that of the Dering Roll would be more appropriate, since the roll has recently been acquired by the British Library for a substantial sum and since that is what Dering is most "famous" for. Source: British Library: The story can be found on several other websites, but they all seem to have copied the information from the BL press release --Labeet (talk) 10:37, 20 October 2008 (UTC)

I myself find the forged brasses more interesting. There are lots of examples of forged documents, but far fewer examples of forgeries of other things used as primary sources. But there's no reason why the roll should not be included as well. Rjm at sleepers (talk) 13:13, 20 October 2008 (UTC)

I can't help but feel that the primary motivation of the section on Forgeries is to express an anti-church agenda rather than explain to wikipedia readers the significance of forgeries as it relates to primary sources. I'm not in a position to debate the accuracies of the forgeries themselves, nor am I inclined to. The second paragraph which contains the examples of forgeries currently cites

   * "Donation of Constantine"
   * Anglo Saxon Charters
   * 11th/12th century monasteries
   * monumental brasses in a local church
   * Hitler Diaries
   * unspecified documents placed in UK National Archives

The last "example" is no example at all since it gives no specifics and is not verifiable. That leaves 4 of the 5 examples relating back to the church. Considering the wikipedia discussion under the general term, "forgery", you will see a broad set of examples which does not appear to take aim at any single entity. If you want the section on "Primary Source"/"forgery" to be helpful to the wikipedia reader, a broader historical cross-section of examples and fewer that take aim at a single source would go a long way to removing bias and improving the quality of this wikipedia article. —Preceding unsigned comment added by (talk) 13:48, 3 April 2010 (UTC)

I'm sorry that user feels the section on forgeries is there to express an anti-church agenda. For myself, I have no such agenda and I don't think other contributors do. I believe that an article on primary sources must include some consideration of the possibility that what purports to be a primary source may in fact be a forgery. The Donation of Constantine was important because of the development of diplomatics which it gave rise to. The analysis of Anglo-Saxon charters produced by 11th and 12th century monasteries (a single example, not two) to determine authenticity is an important part of studies of this period and this type of document. These are the only forgeries in the article attributable to the church. The false brasses were undoubtedly perpetrated by a layman and have nothing to do with the church other than being placed there in an attempt to establish false provenance. The last example (forged documents placed in the UK National Archives) is easily verifiable by following the cited link and there are a number of other sources that could be used. The documents in question were forged typescripts apparently intended to support a particular historians view about the relationship between Churchill and Hitler. My feeling was that the specifics would be too great a level of detail, but more can be added if you feel it would help. I will certainly add another reference to provide additional details for the interested reader. I have had a look at the forgery article and couldn't see anything there that seemed to me to be more suitable than the current examples. All that having been said, I'm sure the community would be happy to consider any additional or alternative examples of forgery of apparently primary sources. Rjm at sleepers (talk) 17:06, 3 April 2010 (UTC)

Is the discussion issued with a court ruling a primary source or a secondary source?[edit]

When a court issues a written ruling, with a statement of facts, followed by the discussion, the discussion has an obvious status of a primary source but also has an appearance of being a secondary source. I see a synthesis, where the judge(s) analyze prior rulings and prior history and synthesize it into a new opinion. Yet, I do not see a 'peer review'. Also, I am not sure the judge is playing the role of a scholar or a researcher, but perhaps some role close to this. I see a gray area, and I am curious of other editors opinions as to whether the discussion that accompanies court rulings is considered by Wikipedia guidelines to be a primary source, or a secondary source. SaltyBoatr 04:10, 15 May 2007 (UTC)

I don't see how a "discussion has an obvious status of a primary source". It seems to be after the fact and by non eye witnesses. This excludes it from being primary. Rjm at sleepers 05:45, 15 May 2007 (UTC)
I wrote obvious because Wikipedia:Attribution/FAQ#Types_of_source_material (second paragraph) says that "historical documents (of) ... trials" are primary sources. It seems conflicted to me that on one hand: the written court opinion is primary, but somehow on the other hand: what they write in the court opinion is secondary. How can it be both? SaltyBoatr 16:45, 15 May 2007 (UTC)
I think there is a difference in usage for the term primary source in different contexts. In historiography, a trial record is a primary source for the trial itself, but not necessarily for the events being described. The Eichmann trial for example took place so long after the event, that the evidence of the witnesses could not be considered primary. Similarly, the evidence of an "expert witness" would not usually be primary. Wikipedia seems to use the term differently. I would not regard newspaper reports as primary since they pass through the hands of a reporter and a sub-editor before appearing in print. Rjm at sleepers 06:01, 16 May 2007 (UTC)
Thanks for your helpful reply. I should have been more clear with my question. I was thinking in context of the Wikipedia usage and meaning of the term. SaltyBoatr 16:14, 16 May 2007 (UTC)

Use of primary sources[edit]

During a spate of vandalism, we have lost the section on the use of primary sources. I'm not sure whether this was deliberate. I'll add it back shortly if no one argues for its removal. Rjm at sleepers 06:16, 18 May 2007 (UTC)

Types of primary source[edit]

A large chunk of this section has twice been deleted and twice reverted. There may be a legitimate objection to the deleted text, but if so the deletion deserves a better explanation. At the moment, it looks to me that a constructive edit clarifying the objection would be a better way of proceeding. Rjm at sleepers 07:33, 12 June 2007 (UTC)

Veracity of Primary Sources[edit]

Perhaps the 9/11 conspiracy could be used as an example of the veracity of a primary source. Yes, the live accounts at the time only mentioned a small plane instead of a commercial jetliner, or that there was a bomb going off in the buildings, but that's because nobody could believe it was even possible at the time. When the unbelievable happens, it should always be taken with a careful grain of salt. MMetro 20:28, 8 September 2007 (UTC)

I'm not sure this would be helpful in an article that is about the use of source material in historical studies. Any mention of 9/11 conspiricy is likely to attract unconstructive edits. I suspect there are less contoversial examples you could choose if you want to make a point about primary sources - Canute commanding the sea to retreat, Columbus discovering America or George Washington chopping down the cherry tree. Rjm at sleepers 06:36, 9 September 2007 (UTC)
I can see the point about Columbus, how our maps conflict a bit with his account, but Washington chopping down the cherry tree is secondary, isn't it? -- Originating in a history book for kids? Maybe, just a regular court case of he said/she said. The transcripts of the testimony in the case is primary, even if they contradict each other.
MMetro 04:16, 12 October 2007 (UTC)
Ah, your point is that primary sources can contradict each other and can be innacurate. A point worth making, although I think it is already covered to some extent. I would still advise against using examples from 9/11 (or the word conspiracy for that matter). Rjm at sleepers 06:38, 12 October 2007 (UTC)

Primary source in scientific literature[edit]

Director Re has added "in the scientific literature, a primary source is the first report of a scientific result by the researchers themselves, and a secondary source is a synthesis of information from primary sources, usually published in review journals." I am not familiar with the usage in scientific literature, but I would have expected the publication of a result to be a secondary source whereas the primary source would be the researchers notebooks. Rjm at sleepers 05:45, 11 October 2007 (UTC)

Hmm, I suppose it is a usage thing. In science we say "the primary literature" more often. In science we generate a lot of data which is never published. Data means nothing unless is is statistically significant. We write articles which are rejected after peer review. We live to get published. So in our way of thinking, if it wasn't published, it didn't happen. I imagine that historians are constrained by lack of data, and value the sources at the level of notebooks. So our systems are out of phase, with scientific notebooks being 0.5-ary sources (zeroary in science), articles being 1.5-ary (primary in science), review articles being 2.5-ary (secondary in science) from a historical perspective and only the tertiary being agreed on as for laypeople. Director Re 02:09, 12 October 2007 (UTC)
Their notebooks are also primary sources-- the key word is synthesis. If they're presenting the findings of their experimental/theoretical research, what they show to the public: patent, book, journal article, thesis, science fair project, etc., is also a primary source. If they're writing for a mainstream medium, let's say Stephen Hawking writes an article for Time, he can't put the raw idea with its complex equations out there, so he is forced to synthesize and interpret his data into something the public can comprehend. He becomes like someone else looking at his work and writing about it, and is therefore creating a secondary source. He can also draw further conclusions from his previous research, perhaps annotating his original text. Without presenting new original research, his annotations would be creating a secondary source-- an analysis of his primary source.
That's also why the history of Fermat's Last Theorem has an annotation as a primary source.
MMetro 04:16, 12 October 2007 (UTC)
IMO, the key word is interpretation. Virtualy no serious scientific paper simply presents the results of an experiment. The results are probably selected and certainly interpreted to extend, refute or support a particular hypothesis. In anycase, published scientific papers are not written at the time of the experiment so they fail to be contemporary. I'm not sure about "primary literature", but I have my doubts. Surely only historians (of science) go back and look at Mendel's original paper (or Newton's Principia or Galileo's Starry Messenger). Rjm at sleepers 06:29, 12 October 2007 (UTC)
Look, I'm just stating how scientists use the terms primary, secondary and tertiary when talking about scientific articles. Director Re 21:58, 12 October 2007 (UTC)
RJM, contemporary does not mean an instantaneous moment in time. If two 18th Century researchers collaborated trans-Atlantic, they'd be contemporary but not simultaneous with the experiments. Take Franklin's kite experiment. So any reasonable delay in publishing results doesn't matter. And of course results are interpreted to support or disprove a hypothesis. That's the point of any experiment. And as for not going back to the original text, the information is much more readily available from a tertiary source such as a textbook, and presented at a level appropriate for the reader, or updated with modern information, such as including relativity with Newtonian physics in a science textbook.
The reason the notebooks and first publication are both primary is that they are the original work of the scientist. Supposing the scientist had made an error in setting up the controls of the experiment, another scientist can look at either materials and say, "This needs to be reexamined". I think calculating the speed of light, spontaneous generation, and Piltdown Man are probably examples. Mmetro 20:38, 14 October 2007 (UTC)
The issue is not delay in publishing results. A modern edition of Domesday is regarded as a primary source despite being more than 900 years after the fact. The delay in writing for publication is significant. There will inevitably be selectivity and interpretation. It is similar to the difference between a diary written up each day and an autobiography written some time later - even if the autobiography uses the diary as a source. The deficiencies of a scientific paper as a primary source are highlighted by reading The Double Helix. However, as Director Re points out, there is a usage of the term primary literature in Science that is different from the usage of primary source in history. So a published paper by a modern scientist may belong to the primary literature as far as a scientist is concerned whilst not being a primary source for the historian. Rjm at sleepers 07:39, 15 October 2007 (UTC)

Using primary resources[edit]

I have removed the following: "When using primary source material, care must be taken to not include copies of primary source material. In certain circumstances, there is a place for this source material, but usually not as part of an article."

I believe it is about Wikipedia articles and Wikipedia's no original research policy and IMO is not relevant or indeed true for using primary sources in historical studies. It probably should go in the relevant Wikipedia guideline. My apologies if I have misunderstood. Rjm at sleepers 08:00, 21 October 2007 (UTC)

Latest revision[edit]

I don't feel comfortable with the most recent revision. How about the following:

In historiography and other areas of scholarship, a primary source (also called original source) is a document, recording or other source of information that was created at the time being studied, by an authoritative source, usually one with direct personal knowledge of the events being described. It serves as an original source of information about the topic. Primary sources are distinguished from secondary sources, which often cite, comment on, or build upon primary sources.[1]

Rjm at sleepers 08:23, 4 November 2007 (UTC)

Sounds fine to me. You got rid of some of the weasel words. Director Re 21:39, 6 November 2007 (UTC)


I have changed the order of the intro, so that the definition precedes the statement about characteristics of a primary source. This is to avoid the inference that the first mention of a particular "fact" is by definition a primary source. Rjm at sleepers 09:55, 15 November 2007 (UTC)

Unreliable sources[edit]

There are a couple of citations to some library websites, which I'm deleting because they are not reliable sources. Also, the idea that library science has some special definition of primary source that differs from that of historiography is unverifiable, probably original research, and simply wrong. Library science uses the same definitions. Also, I resored a citation to a reliable source (a published book). COGDEN 17:30, 15 November 2007 (UTC)

These repeated attempts to change the meaning of "Primary, secondary and tertiary sources" by removing the library and information science usage are inappropriate, and based on an agenda that was repeatedly made quite plain on WT:NOR, where the issue has been under debate by a number of dissenters about the use of source classification in WP:NOR. Post-secondary (college-level) library website pages are quite reliable summaries of PSTS in general library science usage. And that is centrally important in this article at present, because WP:NOR was linked by you (Cogden) to the main namespace articles such as this one. The vital distinction here is that libraries and other information resources use PSTS to facilitate any type of research, very analogously to Wikipedia, while other areas of scholarship classify sources as primary, secondary and tertiary in order to facilitate original research in their respective disciplines. Thus, the library science usage is a crucial part of this article, every bit as much as more specific areas of scholarship such as historiography. ... Kenosis 18:22, 15 November 2007 (UTC)
I don't see how these sources are unreliable. An explanation of why precisely they are not reliable would be appreciated. JoshuaZ 18:28, 15 November 2007 (UTC)
I fail to see how they are reliable either. In fact, I fail to see how any of the edits COGDEN has made to this article or the Secondary source article are of any value. But, perhaps he can explain why I'm mistaken in my assessment, and why his edits are not just of value, but are superior to any of those made by Kenosis and I shall declare with words of soberness that I was in error.
The question is not which edits are of "more value." The question is whether they are verifiable. I have yet to see a reliable source stating that the library science definitions are different from that used in historiography. First, they are unreliable because they are transient websites that are not subject to a fact-checking process. Second, the cites themselves do not say that their definitions are different from those used in historiography. Any such inferences are original research. Besides, all the definitions on these websites are consistent with the definitions used by historiographers. If you want to say something, make sure it is verifiable and isn't just your own interpretation of the website. COGDEN 23:08, 15 November 2007 (UTC)
Transient websites? Are they going to be locked up on vagrancy charges?
Second, the cites themselves do not say that their definitions are different from those used in historiography. What does that mean? Really, what does it mean? I don't recall disclaimers in the OED along the lines of "warning: this definition is different from that of Merriam-Webster".
If the defs aren't different, why do you argue that the sources of the defs are unreliable? Or is the difference you perceive (or maybe don't perceive) OR? Or are you somehow saying that stating that chicken nuggets and filet mignon are different is OR? I'm sorry, but the entire argument you've raised makes no sense and there's really no logical way to address it. &#0149;Jim62sch&#0149; 08:47, 16 November 2007 (UTC)
All I'm saying is that you need a reliable source that says the library science definition of primary/secondary differs from that of historiography. These sources don't say that. In the secondary source, there has been suggested some anonymous librarian webmaster's definition that sort of says something related, but not really, but that's not a reliable source, for the reasons discussed on that page. And it doesn't even really say that, so if you say it in the article, that constitutes WP:OR.
The reason I'm arguing against including these citations, even though the definitions in the citations are the same as that in the published peer-reviewed sources, is that we have this policy called WP:RS, that says we have to use reliable sources, not just some high school teacher's syllabus or some library's "help page". COGDEN 17:45, 16 November 2007 (UTC)
Sorry, dude, but you don't have consensus. I reverted. &#0149;Jim62sch&#0149; 19:21, 16 November 2007 (UTC)
You are reverting relevant, verifiable information that is cited to reliable sources. If you disagree, the correct approach is not to censor what you disagree with, but rather to find your own sources and incorporate contrasting points of view, if you can find them in reliable sources. COGDEN 20:17, 16 November 2007 (UTC)
Believe it or not, this isn't a pissing contest. You are trying to change policy based on arguments that are bereft of reason, unless one defines reason as a burning desire to add OR to a specific article. If you wish to make changes to policy, see frigging WP:Consensus and try to work on achieving it. &#0149;Jim62sch&#0149; 22:02, 16 November 2007 (UTC)
The content here is independent of the policy. This is about the definition of a primary source, for which Wikipedia is totally irrelevant. We can't go beyond what is said in the reliable sources. This article needs verifiability, which is a fundamental Wikipedia policy. We can't ignore Wikipedia policy just because one of the policy pages is related to the definitions discussed herein. We need to get this page right, and if someone thinks the policy is not compatible with the term primary source, then we'll use some other term there. But we can't sacrifice WP:V just for this page. COGDEN 22:29, 16 November 2007 (UTC)
What? I don't even know what this "for which Wikipedia is totally irrelevant" means. Hell, I don't know what your entire screed means. &#0149;Jim62sch&#0149; 22:08, 17 November 2007 (UTC)

Discussion at RS noticeboard[edit]

FYI, I've brought up the issue of whether library help pages are reliable sources on the Reliable Source noticeboard, so that we can have some outside comment. It's obvious where I stand, but if there appears to be a Wikipedia-wide consensus that Wikipedia should accept such websites as reliable sources, then I'll accept that. Ultimately, it doesn't matter as far as the content is concerned, since the definitions are the same, but it's the principle of the thing—I think that when peer-reviewed journals are available, we have no business citing some library nerd's website (which is all these are, for all we know). COGDEN 22:37, 16 November 2007 (UTC)

Uh, right. &#0149;Jim62sch&#0149; 22:08, 17 November 2007 (UTC)

Earliest source primary[edit]

I have modified the comment "the primary nature of a source may also derive from the fact that no copy of an original source material exists, and it is the oldest extant source for the information cited". I have not checked the reference, but I assume it does accurately reflect the sentiments expressed. Nonetheless, there are many historians who would not accept this idea. For example in The Archimedes Codex (page 32), Netz points out that the earliest source for Archimedes age at death was in the 12th century AD. He describes the source as "a gossipy, fanciful poem". Rjm at sleepers (talk) 07:58, 21 November 2007 (UTC)

Original research[edit]

This article's lede, while well-sourced, commits original research. It presents definitions of primary sources as authoritative and unified, which couldn't be further from the truth. It pushes a particular (synthesized) POV of primary sources, instead of objectively describing the many varied and often contradicting definitions of primary sources. Vassyana (talk) 18:11, 28 November 2007 (UTC)

If you think a particular POV is not adequately represented, then please add it. What is it, anyway? You should be more specific. COGDEN 01:39, 4 December 2007 (UTC)
The issue is that a composite definition based on the POV of an editor is what is presented. A synthesized definition is provided, rather than accurately presenting the issue. Vassyana (talk) 10:20, 6 December 2007 (UTC)
I'm having some difficulty understanding what you regard as original research or POV. There are 6 citations in 7 lines. The only significant statement without a citation is "a document, recording or other source of information that was created at roughly the time being studied, by an authoritative source, usually one with direct personal knowledge of the events being described." Is this what you regard as POV? Rjm at sleepers (talk) 10:51, 6 December 2007 (UTC)
No. See WP:SYN. You can have tons of citations, but it doesn't mean the article is fine. The lede takes a particular POV and justifies it by weaving together different definitions. There are several common definitions of "primary source" and they can be quite different both between and within fields. The article should not be presenting them as a unified definition and should not be presenting information limited to certain fields and subfields as universal. COGDEN damn well knows better because he's discussed this issue repeatedly and been shown how his approach is flawed. Vassyana (talk) 11:08, 6 December 2007 (UTC)
I've re-written and removed tag. Rjm at sleepers (talk) 08:42, 3 April 2008 (UTC)
The rewriting is quite appropriate, since the term "primary source" not only has historiographical meaning but also medical meaning and other meanings. The original lead was quite misleading to novices. --RekishiEJ (talk) 12:14, 4 August 2008 (UTC)

original recording[edit]

I've never heard a primary source be called an original recording - this may be ignorance on my part. Most internet usage of the term original recording seems to be about audio. Anyway, I'll put a fact tag on it. Rjm at sleepers (talk) 08:24, 29 December 2007 (UTC)

No source, so I will remove it. Rjm at sleepers (talk) 17:43, 28 January 2008 (UTC)

Removed Olden-Jørgensen blockquote[edit]

I removed the second paragraph, a blockquote by a Danish writer named Olden-Jørgensen. I gather that the translator (BH) was trying to correct what he saw as an inadequate distinction between primary and secondary sources, but this passage failed to make the point. — ℜob C. alias ᴀʟᴀʀoʙ 18:31, 16 April 2009 (UTC)

Primary, secondary or tertiary source?[edit]

[[Image:Nihon_odai_ichiran_waseda.jpg‎|thumb|right|80px|This illustrative page from the Waseda University's copy of the 17th century Nihon Ōdai Ichiran shows part of a widely-used Edo period reference book about Imperial Japanese history.]] [[Image:Nihon Odai Ichiran sample page.jpg|thumb|right|80px|This sample page from the 19th century Nihon Ōdai Ichiran illustrates the book's section layout, merged composition of Japanese kanji and French type-face text, and rare pre-Hepburn transliterations in the context of the original published paragraphs.]] The following text was added to the article. Then I solicited feedback from contributors to this talk page. There was some uncertainty about whether this was helpful; and so I moved the text here pending consensus about how best to proceed.

In Japan's Edo period, the imprimatur of the Tokugawa shogunate was given to a standardized history text prepared by the head of the Yushima Seidō and the national, state-supported educational system. This compilation was developed from extant historical records and earlier histories; and it would have been considered a tertiary source when it began to be distributed amongst government officials nationwide. An illustrative page from the original 17th-century Japanese text of Hayashi Gahō's Nihon Ōdai Ichiran (above left) is considered a "primary source" in the context established by the first 19th-century European translation of this chronology of Japanese emperors.
Isaac Titsingh's 1834 book, Annales des empereurs du japon, was the first Japanese history text to be published and disseminated in the West. A page from Titsingh's book (above right) would be considered a "secondary source" in relation to the Hayashi text; but it would be identified as a "primary source" in terms of works by later writers. For example, Wikipedia articles about Nihon Ōdai Ichiran or about Emperor Kenzō, which would each be categorized as a "tertiary source" in relation to the Japanese original or as a "secondary source" in relation to its first publication in Europe.
The introduction to Timon Screech's 2006 book about Titsingh's posthumous Illustrations of Japan (1822) establishes the Dutch scholar's work in relation to what Japanese scholars compiled in the 17th-century. This modern British scholar's prose becomes an illustrative example of a tertiary text which discusses its sources. Screech explains that Titsingh relied on nine specific Japanese sources, including Nihon Ōdai Ichiran:
The work was published ... in 1663 (it was reissued in 1803), perhaps because it was a necessary reference work for officials, and did not evaluate, but merely listed rulers .... It is a dry affair, and Titsingh acknowledged that translating it was 'a most tedious task.' In the end, the Table was excised from the posthumous version of Titsingh's book, although it was eventually published separately, by a British press, though in French, in 1834, under the title of Annales des empereurs du japon. It does not figure in the present edition."
-- Timon Screech in Secret Memoirs of the Shoguns: Isaac Titsingh and Japan, 1785-1812.[2]

If there is a way that these paragraphs can be modified so that they enhance the quality of the article, I would be pleased to see this text restored.

  • Kenosis focused on the very thing that worried me most:
"Truth be told, I would be very reluctant to confuse unfamiliar readers with a debate or example as "nit-picky" as the example you give. The concept of a "primary source" is very much relative to the scope of one's inquiry .... I think it's preferable to place this content you added at the bottom, and see where it goes in terms of other WP editors' responses to the example. I stopped short of removing it outright."
  • COGDEN offered good critical commentary:
"I think it's a good example, but to ensure complaince with WP:V and WP:NOR, there ought to be some other source that explicitly says that Nihon Ōdai Ichiran is a primary source, that Annales des empereurs du japon is a secondary source, and that an encyclopedia article about it is a secondary or tertiary source. I think the classification of sources in the example is true and valid, but I think there ought to be supported by citations that actually make those characterizations."
"I'm not sure the way in which you have used the term "primary" is consistent with the definition at the start of the article - at least for historiography. I'm not qualified to comment on usage in other disciplines. I concede there are occasions when historians treat transcripts and facsimiles as if they were primary, although this is a slippery path. Since I'm not comfortable with the whole section, I'm reluctant to make minor changes.

It may be that this effort -- despite my good intentions -- simply doesn't work in the context of this article? --Tenmei (talk) 21:49, 17 April 2009 (UTC)

I think you should clean up the opening paragraph of your comment, -- clarify and define what you are talking about. When you write: "The following text was added to the article." ...I'm guessing you mean; add to THIS article?

If so, while I think the points and questions you raise are critical, no I don't think those quotes belong here. They are too long, complex, and require too much inside expertise to be used as examples or illustrations. Examples should illuminate, not raise further questions. But again, those and similar questions need to be discussed.

Here is one question you may have in mind: Should the term Primary Source be defined strictly according to technical definitions such as proximity to the events in question, or should the term also imply that it's a good or reliable source -- as far as we know? For example, the king's personal friend who is also his official propaganda arm, town crier, and biographer? And what of the king's dire enemy? A thousand years later, those may have the maximum proximity to the events in question. Or as you ask, What is the source of the source? Should those things be considered before we call something a Primary Source? In my mind, the king or an unbiased observer is a primary source, and a vested interest is not. Since the town crier is being told what to write, he's a secondary source. You seem to be asking; Should a thousand years have the power to turn a secondary source into a primary source?

To over-simplify, I guess a question I'm asking here and below is, should the historian's job be as seeker and teacher of truth, or as a stenographer?

Doug Bashford-- (talk) 22:36, 12 September 2010 (UTC)


Would anyone object if I set up some auto-archiving? There are some very old threads here. Yaris678 (talk) 01:06, 27 December 2009 (UTC)

What about in Current Events?[edit]

Today, huge segments of the population are being trained that all the past respected inputs from reality (news, science, textbooks, academia, etc) are lying to them. (Think Rush Limbaugh etc, that's his main teaching.) So? How does one find truth? It's a question of huge and growing new importance.

: primary source

Definition: an original fundamental and authoritative document pertaining to an event or subject of inquiry; a firsthand or eyewitness account of an event

This article's emphasis says primary sources are almost exclusively about history. Somehow I'd always thought of primary sources in terms of current events: say; the news maker himself, as opposed to the reporter or commentator. But more than just current events, in all epistemological endeavors: Where is the truth?

What is the term I'm looking for?

As an aside, by this article's definitions, I'm concerned that Rush Limbaugh's world view and depiction of current events will be given nearly as much weight by future historians as say, the president of France, or the CEO's (or other movers and shakers) who's actions made the history. Shouldn't this be more plainly discussed?

Doug Bashford-- (talk) 19:15, 12 September 2010 (UTC)

Primary is not first-person or first-party[edit]

I think it would be good to have this article directly address the differences between a "primary source" and a "first-person" source (an issue of the grammatical person), as well as a first-party source (an issue of involvement).

For example, imagine a person who witnesses a motor vehicle accident and writes a description in his (or her) diary, like these:

  • The red car ran the stop sign and hit the blue car. This is a third-person, third-party, primary source for what the red car did.
  • I called the police. This is a first-person, first-party, primary source for what the witness did.

These statements would both be considered primary sources, but they are are not both first-person, first-party sources. WhatamIdoing (talk) 17:54, 14 November 2010 (UTC)

Is the term "first-person source" a common one? SlimVirgin talk|contribs 18:04, 14 November 2010 (UTC)
The overwhelming majority of autobiographies are first-person sources. WhatamIdoing (talk) 21:49, 14 November 2010 (UTC)
What I think SlimVirgin was asking, was whether the distinction between the two types of sources you listed, is common enough to be notable. --Saddhiyama (talk) 10:44, 15 November 2010 (UTC)
As far as I can tell, no reliable source ever uses them interchangeably. Whether they bother to explicitly say something like, "Now I am going to tell you what a historian calls a primary source; please note that this has nothing to do with what a grammarian calls a first-person source, although it is possible for a source to be both primary and grammatically written in the first person" is not something I know.
But Wikipedia provides a lot of disambiguation and context-setting information, and I see this as more of the same. It would also help build the web to get people to the article that they actually want, if they guessed wrong in their search.
As for SV's question about commonness, if it applies to merely whether people use the exact terms "first-person source" or "first-person sources", then I find about 50K ghits between them, so they are certainly not phrases of my creation. WhatamIdoing (talk) 20:45, 16 November 2010 (UTC)

Is the earliest source primary - reprise[edit]

The following has been removed

If created by a human source, then a source with direct personal knowledge of the events being described.

This was accompanied by the edit summary "ancient documents relating second-hand stories are still primary sources according to historians". I was intending to revert this removal, but the removed text was grammatically poor so I will raise it on the discussion page. As I pointed out above, there are many historians who would not accept the idea that a second hand story can be a primary source. To repeat my example in The Archimedes Codex (page 32), Netz points out that the earliest source for Archimedes age at death was in the 12th century AD. He describes the source as "a gossipy, fanciful poem". For many (most?) historians, qualification as a primary source requires direct personal knowledge written down at the time. I propose the following.

Information for which the writer has no personal knowledge is not primary, although it may be used by historians in the abscence of a primary source.

Rjm at sleepers (talk) 07:44, 22 March 2011 (UTC)

First line of intro[edit]

Ntox has recently changed the first line of the intro. I have to say, I don't particularly like what he wrote, but it was probably no worse than what was there before and there were refs to support it. I have changed the singular to the plural which seems to me to be better English. I have also added a clarification from one of the sources to the effect that a primary source is not written with the benefit of hindsight. The problem with the current first line of the intro is it fails to get across the need for direct personal knowledge, written down at the time. An autobiography written by a politician after he or she leaves office may be the original source of a slur about another politician, but it is not a primary source. Rjm at sleepers (talk) 07:31, 13 March 2012 (UTC)


For the moment, I have reverted a deletion by Staszec Lem. The referenced source clearly says that accounts written after the fact with the benefit of hindsight are secondary (and therefore not primary). On the other hand, it is a rather clumsy sentence, so we should be able to improve it. Rjm at sleepers (talk) 06:43, 20 July 2012 (UTC)

For the moment I have moved it to later in the intro and reversed it so that it says something is a secondary source rather than something is not a primary source. :::: — Preceding unsigned comment added by Rjm at sleepers (talkcontribs) 06:50, 20 July 2012
I deleted one word as the Rjm's version seemed to not quite be English.
< Generally, accounts written after the fact with the benefit of hindsight primary are secondary.
> Generally, accounts written after the fact with the benefit of hindsight are secondary.
I believe Rjm had been trying to collapse two sentences into one. The cited source has:
Generally, they are accounts written after the fact with the benefit of hindsight. They are interpretations and evaluations of primary sources.
I suspect "interpretations and evaluations of primary sources" is an important part but don't see a clean way to work that in. I'd do the cleanup now the overall lead is confusing in that there at several negative logic statements. The article starts out with "Primary sources are original materials." That's fine.
Next is "Information for which the writer has no personal knowledge is not primary, although it may be used by historians in the absence of a primary source." Would people be horrified if the first part was reworded as "Information for which the writer has personal knowledge is primary" or to tighten it up "If a writer has personal knowledge of a subject then the writer is considered a primary source." I'm trying to get rid of that double "not", particularly as this is the second sentence in the article.
"Although it may be used by historians in the absence of a primary source" is an unrelated concept and should not have been in the same sentence. I'm not sure what that part is trying to say. The object "it" in that sentence is "Information for which the writer has no personal knowledge is not primary." I'm also wondering why "historians" is in that sentence. If you reworded it as "Information for which the writer has personal knowledge is primary, although it may be used in the absence of a primary source" then you can see the logic of the article is falling apart.
In summary, the lead, and much of the article needs attention. Unfortunately, the reason I arrived at this article is because it's wikilinked from Wikipedia:No original research#Primary, secondary and tertiary sources. Thus it's being used to clarify one of Wikipedia's core policies. --Marc Kupper|talk 21:08, 21 December 2012 (UTC)


I am about to remove the diagram that was added recently. It looks like a Venn diagram and therefore implies that primary sources are a subset of secondary sources which is not true. Rjm at sleepers (talk) 10:52, 20 March 2013 (UTC)

I'm not excited about the diagram, but I don't agree that it looks like a Venn diagram or that it implies that the inner items are subsets of the outer items. WhatamIdoing (talk) 20:03, 20 March 2013 (UTC)
I am also unexcited by the diagram for other reasons. If anyone regards it as helpful, perhaps they could re-draw it with three non-overlapping circles. Rjm at sleepers (talk) 06:30, 21 March 2013 (UTC)


I have removed JSTOR as a repository for primary sources. There may be a few transcripts and documents that are regarded as primary in some fields, but it is almost exclusively journals and books and therefore secondary. Rjm at sleepers (talk) 08:40, 25 November 2013 (UTC)

Investigative Journalism[edit]

The Wikipedia article says - "Investigative journalism is a form of journalism in which reporters deeply investigate a single topic of interest, such as serious crimes, political corruption, or corporate wrongdoing. An investigative journalist may spend months or years researching and preparing a report."

There may be specific cases where an investigative journalist is an eyewitness or a participant, but the essence of investigative journalism is that the journalist talks to those who participated and reaches a conclusion based on many sources. It is a virtually textbook example of a secondary source. The current citation for the claim that it is a primary source says "Primary sources provide first-hand testimony or direct evidence concerning a topic under investigation. They are created by witnesses or recorders who experienced the events or conditions being documented." Despite the use of the word investigation, this does not come close to saying that investigative journalism fits the definition of a primary source. In the absence of a specific citation, I propose to delete the claim. Rjm at sleepers (talk) 05:40, 27 September 2014 (UTC)

Agree. Good call. GrindtXX (talk) 12:32, 27 September 2014 (UTC)
No objection or alternative citation, so I have removed it. Rjm at sleepers (talk) 06:26, 29 September 2014 (UTC)
    • ^ Handlin (1954) 118-246
    • ^ Screech, Timon. (2006). Secret Memoirs of the Shoguns: Isaac Titsingh and Japan, 1785-1812. p. 65.