Wikipedia talk:What SYNTH is not

Illustrations

Added, "Objective straightforward basic descriptions of a painting or diagram is not SYNTH." This is similar to what is in Wikipedia:Identifying_and_using_primary_and_secondary_sources#You are allowed to use primary sources... carefully stating, "An article about a painting: The painting itself is an acceptable primary source for information about the colors, shapes, and figures in the painting... It is not an acceptable source for claims about the artist's motivation, allusions or relationships to other works, the meaning of the figures in the painting." - Sidelight12 ^Talk 10:00, 15 June 2013 (UTC)

First sentence

Is this really necessary?

SYNTH cautions against committing "original syn", i.e. original research by synthesis

It should just be

SYNTH cautions against committing original research by synthesis

- Sidelight12 ^Talk 04:53, 19 June 2013 (UTC)

Done. --Dan Wylie-Sears 2 (talk) 05:49, 18 April 2014 (UTC)

SYNTH is not useless

Moot point, complaint based on a misunderstanding

I am seeing a major problem with this section. It seems as if the author is presenting a quote which (depending on your interpretation) may contain a synthesized claim, then throwing up OR to argue against it. Check WP:SYNTH, which says: "A and B, therefore C" is acceptable only if a reliable source has published the same argument in relation to the topic of the article. In this case, a reliable source (the source for the Major's quote) has in fact published the argument used. The Major said, in no uncertain terms "Any man who is missing after a battle(A), and not captured by the enemy or found among the dead(B), is presumed to be a deserter(C)." I understand that there can be some debate over this example, as there is no source to say the men were actually listed as deserters, but my point remains. In short, I don't think an example which can be argued not to be synthesis is a good example for how useful the synthesis policies are. In fact, I think it's better as an example of the shortcomings of the policy. I'm reluctant to make such a bold change to this essay, but unless I'm missing something, I plan to change that entire section to a more clear example of why SYNTH is a bad thing. Inside the collapse is what I have in mind. If you disagree, by all means let me hear your thoughts. SYNTH is not useless "SYNTH" refers both to a policy forbidding original research by synthesis, and to such synthesis itself. The policy is not useless. You should not be able to use Wikipedia to publish a crackpot theory, even if you can cite sources for all the premises of your arguments. You should not be able to count something as verified when it really isn't, even if you can cite sources for all the premises of your arguments. Here's a hypothetical example of SYNTH: N The theory that the earth might be hollow has its supporters in academia[citation needed]. Independent researcher Tom wrote "Barring any expedition which drills some 50 miles below the surface to prove the existence of an Oxygen/Magnesium/Silicon mantle, it is impossible to state conclusively that this model is accurate."[1] Prof. Dick wrote "The precise contents of the Earth's core can not be proven by analysis of seismic activity alone. To say that research in this field alone is enough to demonstrate a nickel-iron core is erroneous."[2] Geologist Harry of Jones University wrote "Continental drift was once seen as a fringe theory. We cannot discount more modern theories -especially ones stemming from serious professionals- without evidence to back that up. Serious inquiry into alternate models is a fundamental part of the scientific method."[3] These quotes show that serious academic research into the hollow earth theory is still happening.[citation needed] The citation needed tags are well-placed. The first one questions whether any of the statements given constitute support in academia for the hollow earth theory, because none of the sourced quotes explicitly support it, nor are all the quoted parties clearly members of academia. To call these statements support for this theory could be highly misleading. The second tag points out that the conclusion (that '...serious academic research into the hollow earth theory is still happening') is also SYNTH. None of the sources state that they are studying the theory, and while all of them might be questioning the scientific consensus, that is not the same as researching an alternative theory. The purpose of the policy is to ensure the accuracy of Wikipedia. The above quote pushes the view that the hypothesis of a Hollow Earth is a serious subject of scientific inquiry, when the the truth is that it is not. Absent this policy, articles about controversial subjects or fringe theories could easily devolve into uselessness.

MjolnirPants Tell me all about it. 15:11, 16 April 2014 (UTC)

Thank you for discussing first. Boldly changing it would have been perfectly acceptable imo, but consensus is preferable.

The A, B, and C I had in mind are A, the Major's rule; B, the set of numbers; and C, that there were 500 deserters. It's quite reasonable to apply the rule to the reported fact. But if no reliable source has ever done so, then the editor who does it is synthesizing. As I say, perhaps the Salamander Valley Critic and the Confederate record-keepers were using different criteria for their lists from those used by Major Jones, or different criteria from each other, or from the source for reference 2. Maybe a subsequent part of Major Jones's order, not included in the excerpt in source #1, stipulates that you have to check not only Confederate records for the particular battle but also records from Confederate internment camps, before concluding that a man was not captured. Maybe the Critic published an addendum the following day, saying that additional men among the dead had been identified as Union soldiers. Maybe Major Jones specified that "missing" refers only to men who have not been reunited with their units within five days of a battle, but source #2 was counting men who were not present for roll call the morning after the battle. Since this is a hypothetical example I made up, we can make up more details if it will make the usefulness clearer. In fact, I think I'll edit the essay to do so.

The inference was prima facie reasonable. But the editor didn't have a reliable source who had considered those possibilities. The usefulness of the SYNTH rule is that it settles questions where both interpretations of what should be included are reasonable -- and does so in accordance with the principle that information included in Wikipedia should not merely be probably accurate, but actually be verifiable by reliable sources.

And now that you've heard where I was coming from with that example, if you still like your example better and want to replace it, I won't object. --Dan Wylie-Sears 2 (talk) 19:22, 16 April 2014 (UTC)

I read your response, and I agree with everything you said there, though I don't think you touched on my concerns. Rather, it seems that you and I had different ideas about what was meant by referring to the usefulness of the policy. I saw the section as an example of why Wikipedia needs the policy to be a good encyclopedia, whereas you seem to have seen it as an example of a specific situation where having such a policy provides a clear solution to any content dispute. Now that you've edited it, I believe I understand exactly what it intends to portray and I don't think my example would express that nearly as well. Thank you. MjolnirPants Tell me all about it. 21:13, 16 April 2014 (UTC)

They're both ways that the policy is useful. The main point is that the essay is not trying to advocate getting rid of the policy. Either example would suffice for that. --Dan Wylie-Sears 2 (talk) 05:59, 18 April 2014 (UTC)

That's true. I've added my example to the page under a new section 'SYNTH is not unnecessary', and re-written parts of the original section to make it clear that they're two different points. Feel free to revert me if you don't like it, or think we should discuss it further before adding. I just didn't think it would help the readability of this section to have another proposal within it. MjolnirPants Tell me all about it. 14:01, 18 April 2014 (UTC)

SYNTH vs simple logical implications

From a discussion elsewhere involving among others Osplace, the issue of whether simple DIRECT logical implication constitute original research (specifically Synthesis of published material) has arisen. This question is related to WP:SYNTH is not the word thus and to a lesser extend to SYNTH is not numerical summarization. In other words, does applying well-known simple rules/laws of logic (in a direct way) to sourced statements constitute OR?

A very simple examples of simple logical implication follows. Suppose that a reliable source states that "A is taller than B". The same or a different reliable source states that "B is taller than C". It is a trivial (direct) conclusion from the rules of logic that A must be taller than C. In other words, it is flatly impossible for A to be taller than B (which is presumably sourced) and for B to be taller than C (also presumably sourced), while A is NOT also taller than C.

Would the unsourced conclusion that "A is taller than C" be original research?

There are many other similar possibilities in which from known (sourced) facts we can reach a conclusion (that is not explicitly sourced) by applying directly the rules of simple logic. Other examples:

1a) Every member of group G is also a member of group H (sourced). John is NOT a member of group H (sourced). Logical (unsourced) conclusion: John is NOT a member of group G.

1b) Every member of group G has condition C (sourced). John does NOT have condition C (sourced). Logical (unsourced) conclusion: John is NOT a member of group G.

2a) Every member of group G was destroyed on the date YYYY-MM-DD (sourced). M1 still existed after YYYY-MM-DD (sourced). Logical (unsourced) conclusion: M1 was NOT a member of group G.

2b) A Tsunami wiped out the population of region R on the date YYYY-MM-DD (sourced). Tom, a resident of region R, was alive after YYYY-MM-DD (sourced). Logical (unsourced) conclusion: Tom had left region R before YYYY-MM-DD.

Please, notice the emphasis on DIRECT application of simple rules of logic. If one makes an intricate lengthy sequence of logical implications to reach a conclusion that constitute a theorem, which SHOULD be sourced.

Please, comment. Virgrod (talk) 17:34, 7 December 2014 (UTC)

Unfortunately natural language and natural life are not mathematics, and simple rule of logic must be applied not only to statements made, but also to context, to primary sources, to secondary sources; how the statements were rephrased during a chain of reference (Chinese whispers). Simple example: your case 2a" Every member of the group was destroyed (sourced). M still existed. Solutions: (s0) M was not a member (s1) M pretended he was not a member so he was spared; (s2) M hid in the bush and the author of the source did not know about this; (s4) M was deadly wounded and put into the common grave later crawled out and kept silent for years in fear of being killed,.... and so on.

Same with other your examples, with the exception of the first one, which is pure math. And even the first one is still questionable: the height may be compared in different conditions. A common example is comparison of buildings. Source S1 compared A and B by rooftops, and source S2 compared B and C including aerials. Or S1 compared the building proper and S2 included pigeon huts built on top without permit. And some other variants. Note that both S1 and S2 give correct answers within their context of question posed (easily guessed), but their answers are not directly comparable.

Therefore wikipedia have rather limited, fool-proof list of exceptions to WP:SYNTH. And even "simple math" is sometimes questioned. -M.Altenmann >t 16:41, 10 December 2014 (UTC)

Thanks, >t. To clarify, I suppose the operating assumption when dealing with RELIABLE sources (and in my examples) is that the source is in principle correct. Do you agree? If reliable sources contradict each other, directly or indirectly, then the article should let the reader know that such contradictions exist. I point this out because in your comment you seem to touch on cases where a source is wrong (the source says ALL were destroyed, but in fact some weren't...which is of course possible, but not the focus here). So, in the examples above, please, consider that the sources are reliable and correct. The question remains whether the DIRECT application of simple everyday logic to specific statements that have been reliably sourced is appropriate (i.e., not OR). In the "taller than" case, you raise a valid point (the measurements may not be comparable) but again, suppose we know they are; the issue remains whether the implication "A is taller than C" would be OR. Also notice that this kind of logic isn't really mathematical logic, but rather "common sense" logic applied to everyday "natural" language, the kind aspiring lawyers, for example, should master before entering law school. Like when someone says "Chicago is in Illinois, and John was NOT in Illinois; hence John could NOT have been in Chicago"... even a person without formal education should be able to follow that kind of logic (and it IS logic). The "whisper" example wouldn't apply since we are only concerned with direct inference from two (properly sourced) statements to another, as in the examples above.Virgrod (talk) 20:17, 10 December 2014 (UTC)

re Chicago: Please keep in mind that this kind of simple logic is always performed when we are summarizing the text from sources. "John lived in Cupertino, then in San Jose, then in Los Gatos". We may just as well summarize "John lived in Silicon Valley (or in California, or in the USA; depending on the context of the summary). Sometimes people jump to wrong conclusions, apply wrong logic, misinterpret sources. There issues are resolved case by case in talk pages. -M.Altenmann >t 03:38, 11 December 2014 (UTC)

>tAgree with all the above. What the rules of inference try to do is to formalize common sense deductions we all do semi-automatically on a daily basis. One can definitely misapply the rules of logic: "Chicago is in Illinois, John is not in Chicago, therefore John is not in Illinois" is wrong (John could be in Urbana, for example). The rules of inference should help us identify/avoid such deductive errors.Virgrod (talk) 08:22, 11 December 2014 (UTC)

re: "you seem to touch on cases where a source is wrong". Not exactly so. Picking on my examples; they are just that, examples, serving to demonstrate that the is no way to 100% formalize real life. I could have continued fine tuning case 2a. there may be not the case that "source is wrong". It is just may be a wikipedian slightly took it out of context, and this again may be only resoved in talk page. Refined cases kind of chicago/illinois will not be objected/discussed, of course, and we don't need policies for common sense. (Wait, we do have one:-) ) And in complicated cases yoo think the policy will help, but in fact the issue will be swept under the rug, and we still will not avoid talk page discussion, only the discussion will be not about sources, but about whether a wikipedian correctly applied the policy. And it will be even longer :-) -M.Altenmann >t 03:38, 11 December 2014 (UTC)

If a source does say that "something happened to ALL members of a group" when in fact there were exceptions, then the source is wrong (should have said "some", "most", instead of "all"). If we didn't read the source correctly then that is our problem. In this discussion we presume that statements are accurate and non-ambiguous and we have interpreted them correctly. The only issue if the inference. Virgrod (talk) 08:22, 11 December 2014 (UTC)

Finally, answering on the original question: does applying well-known simple rules/laws of logic (in a direct way) to sourced statements constitute OR? -- Yes, if one may reasonably demonstrate that for "If A and B then C" there may be reasonable contexts for statements A and B that make them incompatible. And this may be established only during a talk; and even simpler things have led to curiously long discussions. Therefore I say, skip the policy, go right into the discussion of the article text. -M.Altenmann >t 03:38, 11 December 2014 (UTC)

In the examples above the statements are presumed comparable. If they aren't or if the statements are ambiguous we have a separate problem (source ambiguity could affect the article even without any "deduction"). Anyhow, we actually started at the talk page of the article in question, and when the OR issue arose, we came here, since it appeared that the issue fell within the scope of Wikipedia:What SYNTH is not, and perhaps was of general interest. The Tsunami example above is closely patterned after the specific issue affecting the article. Virgrod (talk) 08:22, 11 December 2014 (UTC)

Virgrod, what you were trying to do in both articles (Ulises Heureaux and Joaquín Balaguer), is WP:OR. Please stick to the point, find the aproppiate reference or decline your claim. Thanks, Osplace 17:05, 11 December 2014 (UTC)

Osplace So far, only one person besides myself has commented on this thread. Perhaps we can leave the discussion of the specific issues in the articles for the talk pages of the respective article.Virgrod (talk) 21:23, 11 December 2014 (UTC)

Are you going back there because this option failed to you? Osplace 02:47, 12 December 2014 (UTC)

SYNTH is not providing a brief, neutral in-text definition of a new term on first mention

Lately I've been running into some odd reversions of edits performed to conform with WP:LINKSTYLE, reverts justified with an edit summary of WP:SYNTH. Perhaps from partisans of the "a wikilink is enough" approach to writing hypertext. I've added a brief item to the list to clarify the lack of conflict between WP:LINKSTYLE and WP:SYNTH. Thank you. Hugh (talk) 20:40, 16 March 2015 (UTC)

Yes, good. I encounter this problem frequently myself, though mostly from specific editors with an axe to grind, and a WP:OWN grip over a certain range of articles they try to keep me out of. It's happened in less predicable circumstances, however. There seems to be a common mistaken belief that nothing can be defined without direct quotation.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  22:35, 27 May 2015 (UTC)

SYNTH question

I understand the SYNTH rule against having A (cited to source), B (cited to 2nd source), C (cited to 3rd source), and then D (editor's own synthesis of ABC, in the case where ABC does not appear in any of the cited sources), or even more obvious cases, such as A, B, C, then D (author's completely OR theory about the subject). But what about A (cited to source), B (cited to 2nd source), C (cited to 3rd source)? To give an example:

As of 2015, Metro City has 50% male inhabitants and 50% female inhabitants.The 2015 Census indicates that Metro City's inhabitants are 95% African American and 5% Caucasian. The most recent survey of sexual orientation in 2015 indicated that 90% of Metro City residents are heterosexual and 10% are homosexual.
Does this paragraph constitute improper synthesis? I have been accused of doing SYNTH in a paragraph I proposed for a popular music genre article, in which I penned a paragraph which was similar in concept to the one above. Thanks for your input!OnBeyondZebrax • TALK 20:48, 26 August 2015 (UTC)

I don't understand what anyone would see in that which would constitute synthesis. Could you link to the talk page where the issue is being discussed? I wouldn't mind weighing in on the actual subject of discussion. MjolnirPants Tell me all about it. 21:04, 26 August 2015 (UTC)

I would like to ask for your input on that other article, but given that an RfC is ongoing on that page (requested by me), I think it violates WP:CAN, the rules against canvassing. Thanks for your answer! As far as the alleged SYNTH in the sample paragraph, a critic could argue that by presenting A, B, then C, there is an implied synthesis of ABC. With the sample paragraph, a critic who views it as SYNTH would argue that the paragraph implies the conclusion "Metro City residents are half male, half female, mostly African American and mostly heterosexual", a statement that cannot be found in any of the individual cited sources. If these critics are right, then substantial portions of WP--where different points are made about a topic in different sentences, using different sources--constitute SYNTH and should be deleted!OnBeyondZebrax • TALK 23:04, 26 August 2015 (UTC)

Well, it's not canvassing if you're seeking outside opinions, it's canvassing if you're targeting editors whose opinions are likely to align with your own. The reason I asked for a link is that I suspect I'm only getting one side of the story from you (no offense). That being said, an implied synthesis is impossible to avoid in any article. You simply cannot put verifiable statements together in any form without implying some relationship between them. MjolnirPants Tell me all about it. 13:47, 27 August 2015 (UTC)

Thanks for your answer. Well-said! The article is heavy metal music.OnBeyondZebrax • TALK 21:27, 28 August 2015 (UTC)

You asked for Consensus at a featured article "Heavy Metal Music talk page", no one agreed with your addition, you came here to drum up some approval. You need and needed consensus. You don't have it. have a nice day.CombatMarshmallow (talk) 21:49, 28 August 2015 (UTC)

I think this discussion needs to stop now. It can be continued there, as there's no real relation between this discussion and this article. MjolnirPants Tell me all about it. 13:09, 31 August 2015 (UTC)

References

SYNTH is not background information

Occasionally I've seen editors saying there's a blanket ban on using pre-event sources in documenting the background to an event. However, there's no prohibition in SYNTH against doing so, provided as usual that the content does not "reach or imply a conclusion not explicitly stated by any of the sources", and that the content is relevant and non-controversially explains the background. If others have encountered this as well, we should add a section saying "SYNTH is not background information" or words to that effect in order to clarify.

For example, IMO this is OK:

"According to officials, Hawaiians would have 12 to 20 minutes to reach shelter in the event of a missile attack from North Korea.(ref pre-event source) The lack of time to seek shelter increased the sense of fear during the false alarm.(ref post-event source that doesn't explicitly say 12 to 20 minutes)."

Of course editors can argue that a piece of content is SYNTH, OR, or NPOV, but shouldn't base the argument solely on the date of the sources.

Rolf H Nelson (talk) 19:44, 14 January 2018 (UTC) revised 19:51

I don't see an issue with that. My advice would be to pose this question at RSN; I doubt there are a lot of watchers on this page. Get a consensus there and it will be easy to hash out how to phrase it, here. ᛗᛁᛟᛚᚾᛁᚱPants Tell me all about it. 19:49, 14 January 2018 (UTC)

Like everything else we put in articles, Background information is whatever sources about an event describe as background to that event. WP:SYNTH comes into play when wikipedia editors are choosing a certain set of information to add as background, which may or may not really have any relevance to that event. This is a very easy way to accidentally slide into SYNTH. If Fact "A" is actually relevant to Event "X", then reliable sources about X will include mention of Fact "A" and can be cited as such. What happens, though, is during coverage of Event "Y", a wikipedia editor who strongly believes Fact "B" is relevant will (with good intentions) include sources which mention Fact "B" published either before "Y" or after "Y" but with no mention of a connection to "Y". Citing sources on X & Y will very naturally give you what is appropriate Background content, without having to guess. -- Netoholic @ 00:58, 15 January 2018 (UTC)

Hi Netoholic, thanks for joining me on this page, since this is exactly what I want to discuss. "Background information is whatever sources about an event describe as background to that event" is not the current WP:SYNTH policy to my knowledge; the policy is not to "reach or imply a conclusion not explicitly stated by any of the sources". I agree with you that citing a source that isn't directly about the event (or isn't about the page topic, in the case of non-event pages) is often a warning sign of SYNTH. I personally disagree that 'If Fact "A" is actually relevant to Event "X", then reliable sources about X will include mention of Fact "A" and can be cited as such' is always true; for example, for length reasons or to avoid repetition, news articles will often provide a one-sentence summary of the background and link to a previous news article published before the event to provide full background information, rather than include details of the background information in the news article. More importantly, there is no policy to my knowledge that's a blanket ban on citing pre-event policies, and I would like to clarify that that is the case. So:

1. Do you or others disagree that "There's no blanket prohibition in SYNTH against using pre-event sources in documenting the background to an event"?
2. Is there alternative wording you'd propose for how this page can clarify that there's no such prohibition?

Rolf H Nelson (talk) 19:42, 20 January 2018 (UTC)

The way I've always seen the practice of citing a source about Y in an article about X is; does the mention of Y imply something which is never directly stated in any reliable source? For example, if a celebrity were to mention having had sex with a teenager, and the sources all characterize this as advocating pedophilia, is it acceptable to point out that the celebrity was actually discussing pederasty while citing a source giving definitions of pederasty and pedophilia, if no sources explicitly mention that? Apologies that the example was not about background information particularly, but this is actually a real example, which was discussed extensively not too long ago. My feelings are that if there's a source that makes the connection, it's acceptable, but we must cite that source along with the off-topic source.

Another acceptable use would be if a source is discussing Y, but gives X as an example, we could use that source in our article about X, despite it being a passing mention. ᛗᛁᛟᛚᚾᛁᚱPants Tell me all about it. 21:58, 20 January 2018 (UTC)

@Rolf h nelson: I think a distinction needs to be made. Lets take Fact A: "The sky is blue." and Event X:"The stock market crashed today.". Fact A has extensive, veritable sources, but how do we know this is a relevant fact to include when writing about Event X? The only way we could properly include this would be if sources about Event X themselves mention Fact A as background. Otherwise, the editors are simply including Fact A on their own presupposition. This is a WP:SYNTH violation. In this case, if a source about Event X mentions Fact A, then only the source for Event X is needed... the preexisting information about Fact A cannot be verifiably connected to Event X. Now, lets add Fact A-1: "The air scatters short-wavelength light more than longer wavelengths" as extended information found in a preexisting source for Fact A. Even if Fact A is described in sources about Event X, they don't mention the extended detail, so should it be included along with Fact A in an article about Event X? No. Sources about Event X don't mention this extended information, so how do we know that it is appropriate to include. Doing so might imply some sort of connection or conclusion that no sources for Event X have raised. This is also WP:SYNTH - editors choosing to include extra information, many steps removed from sources about Event X, which could falsely imply a conclusion or connection. This is the danger with using sources from before Event X to talk about Fact A - that extended detail can be injected into the Background which makes false implications. We should therefore stick only to sources about Event X and follow where they lead. (added) Oh, to answer your questions directly #1: I disagree,. There should be a blanket prohibition on using preexisting sources, to avoid the Fact A-1 problem with extended information being injected. We can avoid them, and so we should, and we have the wikilink mechanism for extended information anyway (though, that can be misused also and has to be watched for). #2 I'll get back to you after we hear from others, but I think this is actually already covered if you read the current wording. -- Netoholic @ 07:28, 21 January 2018 (UTC)

I agree the specific example is a SYNTH violation, as it "reaches or implies a conclusion not explicitly stated by any of the sources". I also agree "there is a danger" of SYNTH when you use any "background" source. However, SYNTH states "Do not combine material from multiple sources to reach or imply a conclusion not explicitly stated by any of the sources". It does not say, "Do not combine material from multiple sources 'that might conceivably' reach or imply a conclusion not explicitly stated by any of the sources". Therefore, agreeing "this particular content does not reach or imply any unsourced conclusion" would a valid reason for inclusion of an item under current policy. Rolf H Nelson (talk) 19:52, 21 January 2018 (UTC)

I would say that the very act of including any information which is not mentioned in sources that cover Event X would be implying a connection or conclusion within that article about Event X. Again, it would be editors choosing to include information that no source for that event ever brought up. -- Netoholic @ 02:49, 22 January 2018 (UTC)

Is there a separate guideline than SYNTH that can apply in some of these situations? As a toy example, let's say the article for elephant has these statements, all well-sourced to a single slightly tongue-in-cheek but accurate WP:RS NYT article mostly about elephants: "An elephant is bigger than a mouse. Mice like to eat peanut butter. Marcellus Gilmore Edson obtained a patent for peanut butter in 1884." We can't win an argument that this is SYNTH, because most people would agree it doesn't significantly "reach or imply a conclusion not explicitly stated by any of the sources". But it's also getting off-topic; the reader came to the page to read about elephants, not the history of peanut butter. Is there a content guideline that can be invoked in that case? Rolf H Nelson (talk) 19:52, 21 January 2018 (UTC)

Any reader would assume that information in an article about elephants has a connection to that topic. Its a de facto implication of a conclusion that Marcellus Gilmore Edson has something to do with elephants eating peanut butter. Now, that's a silly example, real world examples are far more subtle and harder for an editor to catch or a reader to question. This is the danger of allowing any information from preexisting sources to be included and why we should stick only to sources about the relevant topic to provide us relevant background. -- Netoholic @ 03:10, 22 January 2018 (UTC)

Maybe that's the core disagreement between the two of us; I would not assume, and I do not believe anyone would actually assume from that example, that "Marcellus Gilmore Edson has something to do with elephants eating peanut butter". Rolf H Nelson (talk) 04:14, 23 January 2018 (UTC)

Then why would that fact be included in an article about elephants? Such supplemental facts are either irrelevent, WP:UNDUE, or to some readers imply a connection (WP:SYNTH). You've made no case why such a fact should be included nor provided any guidance for avoiding these potential problems, so why include it? This is why its safer to just stick to sources about the main topic and let them provide us the background, not for editors of wikipedia to try and decide what level of background information is necessary, because doing so inevitably leads to these problems. -- Netoholic @ 04:25, 23 January 2018 (UTC)

Let me keep an eye out for a good example of when we should use a "background" source in the wild, and if I do I'll post it here. Maybe scenarios where we need to use a background source are rare. Rolf H Nelson (talk) 06:37, 31 January 2018 (UTC)

RfC on content concerning illegal fetal tissue dealers

https://en.m.wikipedia.org/wiki/Planned_Parenthood_2015_undercover_videos_controversy#/talk/8#/talk/8

I'd love some comments on this. Maybe I misunderstand the Wikipedia policies, but at least some of the content seems relevant. natemup (talk) 16:32, 21 September 2019 (UTC)

SYNTH is not NPOV, when it is point-by-point

The section, SYNTH is not NPOV, when it is point-by-point, relies on a previous wording of WP:SYNTH, which prohibited synthesis of published material that advances a position (emphasis added). Current versions of WP:SYNTH do not include the emphasised text; and the section is no longer relevant to the policy.

Unless there is an objection, I will remove the section. - Ryk72 ^talk 23:32, 14 October 2022 (UTC)

I think the section is still very relevant to the policy even though the wording it relies upon is not still exactly the same. The first sentence might need to be updated to reflect the changes, but removing the whole section based on a small word change is an extremely drastic reaction that takes out the whole concept the section is trying to explain when the concept itself could exist even without the first sentence present so it doesn't really rely so much on the previous wording of synth so much as it does NPOV. Huggums537 (talk) 20:50, 15 October 2022 (UTC)

@Huggums537: Firstly thanks for your kind reply. Greatly appreciated. Reviewing the section again, I see the outdated phrasing present in both the first and third sentences, and as fairly fundamental to the logic of the section; so am, personally, struggling to see how that phrasing could be removed, but the core message retained. Could you outline the key concept that you see as important? We might then easily reach alignment. - Ryk72 ^talk 00:27, 16 October 2022 (UTC) - reping, typo - Ryk72 ^talk 00:29, 16 October 2022 (UTC)

I think the third sentence is fine since it is not specifically referencing any outdated policy like the first sentence does. There isn't any rule that says we should ban an entire concept just because it uses a small portion of language that was changed elsewhere, nor should there be. Huggums537 (talk) 00:59, 16 October 2022 (UTC)

Again, I thank you for the reply. For clarity, I think there are issues with the concluding sentence as well: So if a single statement, taken in isolation, is NPOV, then it's not SYNTH. This seems inherently false, neutrality & original research are orthogonal; and seems contradicted by several of the examples at WP:SYNTH.

As the concluding sentence, it seems to me that it is the key concept of this section. But I am open to other thoughts. Can you outline the concept of this section, as you understand it, please? - Ryk72 ^talk 01:04, 16 October 2022 (UTC)

I am somewhat dumbfounded as to why you would be advocating for the removal of the entire section if you don't even understand the concept? The reason you might not understand the concept or the last sentence based on the examples given at SYNTH is because they only give examples of wrong and no examples of right. If they had included this example:

Y The United Nations' stated objective is to maintain international peace and security. There have been 160 wars throughout the world since its creation.

Then you can see that each individual sentence taken in isolation is a NPOV statement since it doesn't advance any position, but only supports facts by reliable sources that don't combine together with each other to imply a new idea so it is not SYNTH. That is the concept. Huggums537 (talk) 01:26, 16 October 2022 (UTC)

• Comment: This discussion does have me thinking about something though. I like the general concept of this section, but I don't really like the language I was advocating for earlier. This idea that a statement can't be NPOV because it advances a position actually mystifies me. The whole point of NPOV is so that each position gets a chanced to advance their statements in due proportion. So, this really needs to be reworded so that the concept of individual NPOV statements not being SYNTH is still intact, but without the inference that NPOV statements are ones that don't advance positions... Huggums537 (talk) 01:49, 16 October 2022 (UTC)

Arithmetic mean, expected value, and other averages

This doesn't seem right:

It is common to confuse the arithmetic mean (summarization) with the expected value (mean and standard deviation calculated by a reliable source).

Either can be reported with or without a standard deviation, and either can be calculated by the source or not. The distinction is whether the values are weighted (e.g. by probability), right? Except that you can also have a weighted arithmetic mean so I'm a bit lost.

Anyway, one possibility is this, which lists all three terms, and asks editors to respect the source's choice to include standard deviation (which is the point I think the original author was trying to make).

It is common to confuse the arithmetic mean, weighted arithmetic mean, and expected value. If the source reports a value together with a standard deviation, the figures probably be reported together on Wikipedia as well.

And maybe add something about medians just before it:

Be careful about the distinction between arithmetic mean, median, and other "averages"; use the precise term.

Cheers Jruderman (talk) 05:10, 7 July 2024 (UTC)

Btw I wonder if some of this should be in the MOS pages Words to watch or dates and numbers (near #Uncertainty) in addition to, or instead of, here. (As always, care should be taken when promoting sentences from an essay into a guideline.) Jruderman (talk) 05:28, 7 July 2024 (UTC)