Wikipedia talk:No original research

The project page associated with this talk page is an official policy on Wikipedia. Policies have wide acceptance among editors and are considered a standard for all users to follow. Please review policy editing recommendations before making any substantive change to this page. Always remember to keep cool when editing, and don't panic.

This is the talk page for discussing improvements to the No original research page.

Put new text under old text. Click here to start a new topic.
New to Wikipedia? Welcome! Learn to edit; get help.

Shortcuts

If you want to know whether particular material constitutes original research or original synthesis, please use the No original research notice board. Questions about the policy itself may be posted here.

view · edit

Frequently asked questions

I disagree with the definition of secondary source.: Wikipedia mostly follows the definition in use by historians, which requires more than simply repeating information from some other source or rearranging information from the author's notes. The earliest definition of a secondary source in this policy was in February 2004 "one that analyzes, assimilates, evaluates, interprets, and/or synthesizes primary sources".
This published, reliable source is engaging in original research.: We prohibit Wikipedia editors from posting their own original research in Wikipedia. Wikipedia editors can only put in Wikipedia what has previously been published in a reliable source.
I've proven that general relativity is wrong, but the physics journals won't publish my proof. Can I use Wikipedia to publish my ideas about how Einstein was wrong? I can cite lots of sources in the article to support each piece of the puzzle.: No. If you want to put a whole idea in Wikipedia, you need to be able to cite a source that contains the whole idea, not just isolated bits of it.

Archives

Index 1, 2, 3, 4, 5, 6, 7, 8, 9, 10
11, 12, 13, 14, 15, 16, 17, 18, 19, 20
21, 22, 23, 24, 25, 26, 27, 28, 29, 30
31, 32, 33, 34, 35, 36, 37, 38, 39, 40
41, 42, 43, 44, 45, 46, 47, 48, 49, 50
51, 52, 53, 54, 55, 56, 57, 58, 59, 60
61, 62, 63, 64

Rewrite 2004-2005: Draft • Talk
"Primary v. secondary sources" subpages
"Change needed": Proposal • Talk
"Various examples": Examples • Talk
"Transclusion example": Example • Talk
List all subpages

This page has archives. Sections older than 30 days may be automatically archived by when more than 2 sections are present.

Publisher website links and WP:PRIMARY

Moved to Wikipedia:No original research/Noticeboard § Publisher website links and WP:PRIMARY

– 15:34, 10 May 2021‎ (UTC)

Seemingly widespread and contradictory views about what primary and secondary sources are

I've been involved in two different discussions of late that, together, have left me baffled. Each one has involved editors who seem quite experienced. I'm not exactly a new editor myself - I've edited regularly for over 2 years and have over 15,000 edits. Yet, the positions espoused seem to be diametrically opposed, and in fact backwards from what would make sense to me.

Exhibit A (permalink). This is a WP:RSN discussion about a research paper analyzing voting results and presenting conclusions. Most, but not all, respondents say that this research paper and the authors' conclusions are a secondary source, with the primary source being the voting results themselves. Example quotes: A secondary source analyzes primary sources to reach conclusions and create knowledge - like this paper and The author is interpreting the exogenous primary data and secondary sources and is thus making secondary interpretation. There are many more like this.

Exhibit B (permalink). This is a discussion at an article where the parts relevant to this discussion are about this book from Oxford University Press by Farber & Sherry. The book discusses pre-existing scholarship, and presents the authors' conclusions about it. Although there are fewer total editors at this discussion, we get comments like this: Farber & Sherry et al. are primary sources for their own "criticism" of CRT. When I quoted the first four sentences of WP:SECONDARY, another editor replied: As the policy says, "A source may be considered primary for one statement but secondary for a different one." All sources are primary for something elaborates: "More importantly, many high-quality sources contain both primary and secondary material....A peer-reviewed journal article may begin by summarizing a careful selection of previously published works to place the new work in context (which is secondary material) before proceeding into a description of a novel idea (which is primary material)."

Am I missing something, or are these two scenarios demonstrating fundamentally contradictory views about what is primary and what is secondary? In both sources, author(s) of the RS are taking pre-existing material and presenting their conclusions about them, which is being used with in-text attribution in the article. In one case, that exact sort of material is being called secondary, and in another, primary.

And this is not only inconsistent, but precisely the opposite of how I (and at the first discussion, some others) would understand it. In the first scenario, the authors' analysis is on raw data, so it would seem to be equivalent to WP:PRIMARY: a scientific paper documenting a new experiment conducted by the author [which] is a primary source for the outcome of that experiment. In fields like political science, the raw data from the "experiment" is usually pre-existing and not created in the authors' own lab - much like many studies in, say, epidemiology, as well as some studies in physical sciences like astronomy. Although this is not a medical topic, MEDRS's section on avoiding primary sources defines primary and secondary in much the same way and is of interest for comparative purposes. Meanwhile, in the second scenario, the source engaging in analysis is discussing pre-existing academic articles, much like a review paper, and is itself a book by an academic publisher. Yet this is supposedly a primary source.

I bring this up because this widespread of contradictory claims may call for clarification of the policy in one or more places. Crossroads ^-talk- 17:11, 14 November 2021 (UTC)[reply]

Mostly in your second discussion the issue was around due weight (and soto voce ONUS) because a source was said to be out-dated, superseded, or not picked up. Mostly primary or secondary does not matter (or is rather an abstract 'purely academic' debate) as long as one is adhering closely to the source material, not talking about living persons, not talking about scientific/medical experiments, and not 'over-representing' or 'over emphasizing' the source. -- Alanscottwalker (talk) 17:32, 14 November 2021 (UTC)[reply]

That was a part of it, yes, but the array of contradictory views on PSTS is still an issue. If PSTS can be used in the way discussion B says, then the editors at discussion A are wrong. Or if those editors are right, then the editors at discussion B are misunderstanding the policy and adding undeserved weight to their arguments.

I am not interested in pursuing either of those discussions further. I want this policy to be clear so misunderstandings are reduced. These were not consistent positions, and we don't want policy to be so subjective and confusing that it becomes 'stuff I want to include is secondary and stuff I don't want is primary'. Crossroads ^-talk- 17:44, 14 November 2021 (UTC)[reply]

Not 'wrong', speaking in different contexts. Not likely an "aray" that any policy writing can cure. No absolute certitude in every situation and nuance is unavoidable. (example, policy says "base" in basing on sources, it is inevitable that application will cause discussion on that). Writing and editing is not an algorithm. Alanscottwalker (talk) 17:54, 14 November 2021 (UTC)[reply]

I feel your pain, and don't have much amelioration to offer beyond aspirin. WP:PSTS doesn't help a lot. WP:IDPRIMARY, an explanatory supplement to PSTS, says that primary sources have three separate, basic characteristics to identify, the third of which is: "Is it primary?". the Primary source article begins, "In the study of history as an academic discipline, ...". The Secondary source article says, "the distinction between primary and secondary sources is subjective and contextual, so that precise definitions are difficult to make. (with shortened footnote cites of a couple of outside sources, only one of which has a full cite in the article). Wtmitchell (talk) (earlier Boracay Bill) 18:20, 14 November 2021 (UTC)[reply]

PSTS makes more sense if we remember how it ORIGINALLY started… originally, the point was to say that we should not make WIKIPEDIA a primary source for information, especially analysis and conclusions. This is why it is contained within WP:NOR. Unfortunately, over the course of various rewrites, that original point was lost. Blueboar (talk) 18:53, 14 November 2021 (UTC)[reply]

Based on my very similar question above I think this is a bit of a confused mess. When I was publishing/reviewing a common question we asked was what is the the novel contribution of this work? If this work is trying to present something new or present an analysis that was derived from a novel analysis of existing data we would call that a primary contribution. If the work was repeating/summarizing the work of others then it was a review work. I can see where this is confusing when we are dealing with the analysis of existing data sets and we should be careful to not assume a meta analysis is the same thing as taking existing data sets and combining them in new ways. If I took data sets about smoking, commuting times, and voter turnout, put them thought some statistical model and reached a novel conclusion about how they are tied together that would be a novel conclusion based on available data. Thus my paper should be viewed as a primary for a discussion regarding how smoking and commuting times impact voter turnout. However, if in the paper note that the data shows that smoking rates have fallen over the past 40 years (my hypothetical data set says that) then that isn't a novel contribution of mine thus my paper can be a secondary source for such a claim. I think the critical part is if I combine data, regardless of it's origin, in a new/novel way to reach a new/novel conclusion, then my paper must be considered a primary source for that claim. Springee (talk) 19:13, 14 November 2021 (UTC)[reply]

Blueboar hit the nail right on the head and if we could go back to the original intention a huge amount of useless back-and-forth would be avoided. Whether a source is primary or secondary is nearly always irrelevant. What we can't do is publish our own analysis or conclusions that the source doesn't explicitly contain, and that is true for all types of sources. That is what NOR means. Applying it directly to Springee's example is much easier than arguing over primary/secondary labels and reaches the correct conclusion right away. Zero^talk 02:55, 28 December 2021 (UTC)[reply]

The best way I've found to judge primary versus secondary for a source or specific source material relative to a topic is the nature of if the source material is transformative of other primary, secondary, or tertiary sources. Transformative being the application of what we normally consider original research but by what we hope is an expert source, such as doing analysis, synthesis, criticism, etc. Simple reorganization or summary of what is out there is not transformative nor for us original research by this policy. So like said above, your Farber & Sherry example is one that I can having secondary nature in that their opinion of the existing research, but as others point out, it is primary source for their own opinions. Or as another example that had come up elsewhere, US Supreme Court decisions often have portions that are secondary sources for existing law/court cases, but the decision itself is primary for the case it deals with.
The problem is that our pages on PSTS evaluation, while they cover this transformative nature as a means to determine it, focus a bit too heavily on the "one step removed" factor, which is important but more about the independence of a work and not so much its primary or secondary nature. I found it far easier to make the determination sticking to "transformation of information" than any other metric. --Masem (t) 03:09, 28 December 2021 (UTC)[reply]

FYI

I've opened up conversation here as to application there of our policy on primary sources. --2603:7000:2143:8500:34F3:46D5:FCE9:6825 (talk) 23:40, 27 December 2021 (UTC)[reply]

Is the goal to avoid the wikipedia contributor's novel interpretation or the primary sources' novel interpretation?

The 2nd sentence of "Primary, secondary and tertiary sources" has me scratching my head.

  Secondary or tertiary sources are needed to establish the topic's notability and avoid novel interpretations of primary sources.

What do the final 6 words want to avoid?

Is the goal to avoid novel interpretation by the primary source's author?

Is the goal to avoid novel intprepretation of the primary source by the wikipedia contributor?

Both? With what weighting?

2601:1C1:C180:4F40:5C4A:BE6D:F0D9:20B5 (talk) 22:29, 6 January 2022 (UTC)[reply]