User talk:Onel5969/rfc draft

Thoughts on the interpretation of "interpret"
I think one of the most significant issues in this saga has been the interpretation of the word interpret. There are two distinct meanings to interpret that apply in this situation (from wiktionary): The second one we do all the time with sources, we translate non-English sources without the need for an existing English translation in a reliable publication. The first is the meaning that is implied in WP:OR. I am of a strong conviction that the second is what is meant by the term "interpreting a map" (although I prefer "reading a map"). So, that being my personal take, perhaps the language in WP:OR could be clarified if "interpret" was changed to "reinterpret", as in "to come to a conclusion distinct from that which was presented by the original author." - Floydian τ ¢ 16:13, 2 December 2022 (UTC)
 * 1) To decode the meaning of a topic and then act, whether to continue researching the topic, follow through, act in opposition, or further the understanding through sharing an interpretation.
 * 2) To explain or tell the meaning of; to translate orally into intelligible or familiar language or terms. applied especially to language, but also to dreams, signs, conduct, mysteries, etc.


 * So that would fall under option 1? How would you word it?  Onel 5969  TT me 17:22, 2 December 2022 (UTC)
 * I guess the simplest method would be to change #4 at WP:PRIMARY from
 * "Do not analyze, evaluate, interpret, or synthesize material found in a primary source yourself; instead, refer to reliable secondary sources that do so."
 * to
 * "Do not analyze, evaluate, interpret, or synthesize new ideas or conclusions from material found in a primary source yourself; instead, refer to reliable secondary sources that do so."
 * Or, perhaps another option is to consider the elevation of WP:MAPCITE to a guideline. - Floydian τ ¢ 18:18, 2 December 2022 (UTC)
 * Take a look at the chart, something like that?  Onel 5969  TT me 18:52, 2 December 2022 (UTC)
 * I'd add it as a "suggested wording" for option 1 rather than making it a distinct option, as the idea behind it would be to remove the language semantics so that reading maps is not considered original interpretation. - Floydian τ ¢ 20:58, 2 December 2022 (UTC)
 * Another concept to consider: we don't require our sources to be in English. Using a source in a foreign language involves much of the same concepts as using a map: there's a similar element of translating the content into English.  Imzadi 1979  →   06:05, 5 December 2022 (UTC)

Proposal 2 has the implicit assumption that all maps are primary sources, which is one of the focal points of the dispute, and such a declaration would have other ramifications. I am not sure this is the way to go. --Rschen7754 18:05, 3 December 2022 (UTC)


 * Is there a similar "do not interpret secondary sources" text? My goal with this change would be to rid the notion that reading (or "interpreting") a map is equivalent to novel interpretation of some raw data. By extension it would make maps acceptable as a source for describing a linear feature (primary or otherwise). - Floydian τ ¢ 18:15, 3 December 2022 (UTC)
 * I think that would be covered by the policy stated in WP:OR: "Articles may make an analytic, evaluative, interpretive, or synthetic claim only if it has been published by a reliable secondary source." In other words, the interpretation or analysis has to be in the secondary source.  Onel 5969  TT me 20:26, 3 December 2022 (UTC)
 * I think that's a very valid point, and as I said in the other page, I think these are two separate issues. How about we change that verbiage from Primary Sources to either MAPCITE or OR?  Which is kind of implied in #3, but we could simply make that stronger and get rid of #3?  Onel 5969  TT me 20:29, 3 December 2022 (UTC)

Another option might be explicitly saying that reading a map is not OR in the "What is not original research" section. Though if we went that route, we might have to draw a line somewhere and say that conclusions should be explicitly represented by the symbols or something like that. There is a difference between saying that road X intersects road Y (obvious through the symbols), and road X passes through a wooded area (sourced to the satellite layer). And a whole spectrum in between. (FWIW I think we need to be careful about using language like "allow" and "prohibit" in the question, because whether most of the current use of maps by the roads projects is allowed or not is disputed. The way I see this potential RFC, the purpose is to clarify by explicitly allowing/prohibiting.) --Rschen7754 01:11, 4 December 2022 (UTC)
 * I noticed that WP:NOTOR includes the following sentence under "Translation and contextualizing": Source information does not need to be in text form—any form of information, such as maps, may be used to provide source information. Interpretation of such media is not original research provided that it is done in a routine manner observing any limitations usually associated with the medium concerned, and such interpretations are readily verifiable by anybody who has access to the same source. Maybe a straightforward answer is to propose moving this into WP:OR. --Rschen7754 04:11, 5 December 2022 (UTC)
 * I think that makes sense to make that one of the suggestions. I've modified selection #1 to read that.  Onel 5969  TT me 01:58, 11 December 2022 (UTC)

Where to add announcement templates when put out there
 Onel 5969  TT me 22:24, 2 December 2022 (UTC)
 * 1) OR talkpage
 * 2) PRIMARYSOURCE talkpage
 * 3) MAPCITE talkpage
 * 4) NPP Reviewers talkpage
 * 5) Road Project talkpage
 * All the country road projects (US, Canada, India, Australia, HK) should be notified. This could have implications on other projects i.e. other modes of transport. I would also include Template talk:Cite map, maybe similar templates. Perhaps WT:FAC, OR noticeboard. --Rschen7754 01:38, 3 December 2022 (UTC)
 * Thanks. Will do that.  Onel 5969  TT me 15:58, 3 December 2022 (UTC)
 * There are 32,000 uses of Cite map . did start a list of possible projects on the other page, but I think someone will have to sit down and go through that list. --Rschen7754 19:04, 3 December 2022 (UTC)

Single maps
This section of MAPCITE is ambiguous and controversial and should be revisited: "When describing the route of a waterway, mountain range, road, railroad, etc., a single map should not be used the sole source used to provide the description."

Does this mean that every part of the description has to be cited to two maps? Or just that some facts can come from one and some from another? Or that some non-map source has to be integrated somehow (if even there is one?)

Someone could also cite both Google and Bing maps and I doubt that really adds anything to the article or would satisfy concerns of those who think map reading is OR. Rschen7754 01:55, 3 December 2022 (UTC)
 * I wrote that text, years ago, and admit it needs improvement and arguably a complete re-write. I intended to make the point that if someone writes a description (not just of roads, but anything) by just following Google maps they are wasting my time. I can just go to Google Maps myself and find out the same info (plus more) faster than I can by reading. My first choice is that prose based sources are also used to source the route description, if available. If none are available, the route description should be compiled from multiple maps, preferably using multiple types of maps (I try to use a topographical, road and a recreation map in mine. That also means by default I'm using both GIS and Satellite image based maps.) I don't think there's value in using two maps of the same type to confirm the details, especially if both maps have been vetted as a reliable source. (Although if the two maps contradict each other it should be handled the same as it would when two reliable prose based sources contradict each other). The intent is more to make the route description of more interest and value to more readers. There are readers that don't give a rats behind about the details covered in a road map, but would be very interested in what a recreational map would show about the highway. You have my blessing to toss that text and replace with better. Dave (talk) 19:51, 3 December 2022 (UTC)
 * I think that if we went with #3 we would need to review the text further. I was not aware of the "single map" portion and it does not reflect working practices in the roads projects. Who knows what else is there. --Rschen7754 00:02, 4 December 2022 (UTC)
 * Agreed. When I first wrote that essay it was about article quality and how to avoid a lame route description, more than compliance with Wikipedia policies. It's evolved in the direction of policy compliance since then. I stand by everything in there, but if it's going to be a policy it probably needs some polish. Dave (talk) 02:14, 4 December 2022 (UTC)
 * I've got a minute. I'll try applying some of that polish today. I'll also make note of this discussion on WT:MAPCITE Dave (talk) 22:36, 4 December 2022 (UTC)

Primary vs. secondary
There's a lot of discussion about the classification of primary vs. secondary sources. Much of it is wrong. WP:MEDRS (specifically WP:MEDDEF) has a good description on the classifications in the context of medical sources, and it bears applying that more widely for a moment to the topic areas being discussed here.

Is a newspaper article a secondary source? In most cases, no, it is not. A newspaper article published contemporaneously to an event is considered a primary source by historians. The reporter observed the event, interviewed participants, photographed the scene, etc. That reporter may not have been an active participant in the event in the same fashion as the medical researcher, but they still documented their perceptions of it.

A secondary source would be something like an article or a book published later as a retrospective into those events. Yet we base whole articles on recent events on the contemporaneous news reporting all the time, even when we're also admonished not to base articles on primary sources.

Just to muddy the waters a bit, but a newspaper article can be both primary and secondary, at the same time. Recently, a new section of freeway opened for U.S. Route 31 in Michigan near Benton Harbor. Articles like reporting on that opening are primary for the opening and the routing that was built, but secondary for the history of that planned freeway conversion dating back to the 1970s and the complications that delayed construction for decades.

WP:GNG doesn't seem to care about any primary–secondary distinction so long as there's "significant coverage". Often that's been interpreted to mean "enduring coverage", or "widespread coverage", meaning the media reports about the event across a wide geographic area, multiple outlets, or for a longer period of time.

I think we need to be honest on what we actually do around here in researching and writing Wikipedia articles. We try to say that we're creating a tertiary source composed of secondary sources. Perhaps in the medical topics we do that because of how strictly WP:MEDRS is applied. History articles are another area that will have a lot of published scholarship that editors can consult. Then we get topic areas like Supreme Court cases where whole sections are frequently written and cited to the case opinions themselves. Or look at articles on episodes of TV shows where the plot summary is implicitly cited to the episode. That isn't to say that there aren't some articles published about the case or TV episode discussing those respective topics that couldn't be cited in place of the original material, but they'd be primary sources as well in many situations due to their contemporaneous nature and lack of objective distance from the events covered.

Classifying a source as secondary does not make it infallible. A magazine feature published in 1994 about a historic hotel may give the wrong opening date when the newspaper in 1930 printed a story the day after the opening along with the names of the first guests, the time they checked in, and their assigned room number. We should objectively cite the 1930 article and not discount it because it's primary.

Given all of the above, and given actual practices, I would say that classification of maps as primary or secondary would be just as nuanced, and yet it doesn't matter. The article on the 2022 West Java earthquake is linked from the Main Page right now. It has 115 footnotes in this revision, which is the one current as I write this. Of those, just 3 are not from 2022, meaning 112 of those sources should be classified as primary sources by academics, at least in part, because they lack the historical distance from the event to be secondary. No one is going to nominate that article for deletion though, nor do I think anyone should.  Imzadi 1979  →   21:30, 4 December 2022 (UTC)  Imzadi 1979   →   21:30, 4 December 2022 (UTC)


 * Hi. I agree with much of what you say above.  One issue is that GNG "doesn't seem to care about any primary–secondary distinction".  Bullet points 3 and 4 clearly indicate that primary sources should not be used: "Sources" should be secondary sources" and ""Independent of the subject" excludes works produced by the article's subject or someone affiliated with it."  Onel 5969  TT me 14:45, 8 December 2022 (UTC)
 * If we are going to apply the primary/secondary source distinction to roads (be it maps, newspapers, whatever) we need to apply it consistently across the site. That is separate from the issue of maps being primary or secondary sources. That distinction generally does not matter because of the SNG (all state/provincial highways and up are notable) and NEXIST, however the SNG is being called into question as well so it may become an issue. So then, do we ask the question about whether maps are primary or secondary? To me that seems like an odd thing to do. Some maps are inevitably going to be primary, but most of the ones the road projects use we would consider secondary. Any question needs to also factor in that some maps are never going to meet reliable sourcing. I can draw a map by hand and put it on my website but it will never be a reliable source. --Rschen7754 02:01, 11 December 2022 (UTC)

The comments left on the other page are enlightening, although I do not necessarily agree with his conclusions. Perhaps we need a second question on this RFC, how to classify maps. I think it is wrong to classify all maps as primary, however that probably does need to be an option since a lot of people apparently believe it. I think a better option would be to set some sort of criteria for how to classify maps, rather than have a RFC for every possible map publisher out there. --Rschen7754 21:06, 18 December 2022 (UTC)


 * All maps cannot be classified as primary. A map in an encyclopedia that sums up different maps published in scholarly articles would be a tertiary source I think. a455bcd9 (Antoine) (talk) 21:10, 18 December 2022 (UTC)
 * We're talking about using maps as sources (i.e. road maps like Rand McNally and Google Maps) rather than about how we make maps for illustrations. --Rschen7754 21:39, 18 December 2022 (UTC)
 * Yes I understand and I said that maps as sources can be primary, secondary, or tertiary sources (depending on the source). a455bcd9 (Antoine) (talk) 21:45, 18 December 2022 (UTC)
 * Yes I'd say the text of this RfC, as currently written, is based on two faulty premises:
 * That all maps are primary sources. That's like saying "all videos are primary" or "all written sources are secondary". A map is a medium for communicating information. What category of source it falls into depends on what that information is and where it came from, not the presentation format.
 * That when current policies refer to sources, they mean written sources. This isn't stated anywhere to my knowledge, and we have well-established systems for citing TV episodes, interviews, podcasts, serials, speeches, audiovisual media in general and–of course–maps suggests they were never meant to be read so narrowly.
 * The problem this RfC is supposed to solve is that editors are using the map and creating their own text, based on the map and this is contrary to WP:V, but if you replace the word "map" with "source" in that quote, you basically have a nutshell for WP:V. –&#8239;Joe (talk) 10:22, 19 December 2022 (UTC)
 * Indeed, I do not think this RFC is necessary at all, however it seems to be the only way forward to prevent editors (not just NPP) from mass blanking articles citing maps. This is also why I am hesitant to start it myself. --Rschen7754 03:05, 20 December 2022 (UTC)
 * I feel like this RfC just means: "Does WP:V apply to maps used as sources?" And everyone will say "yes". A result that won't help us 🤷. a455bcd9 (Antoine) (talk) 08:28, 20 December 2022 (UTC)

Related debate
Please note this ongoing related debate: Wikipedia_talk:No_original_research (conclusion at the end).

I don't know the issue of road articles at all. I would only be wary of unintended consequences. For instance, I think users should be able to use this map made by a RS (see Ethnologue) to write in an article: "Najdi Arabic is spoken around Buraidah".

We should also distinguish two situations:
 * Using maps as a source to create a new map uploaded to Commons and used as an illustration on a Wikipedia article
 * Using maps as a source to write text on Wikipedia

And of course, distinguish whether the map used is a primary source (Google Maps?), secondary source (map in an academic paper or book), or tertiary source (map in an encyclopedia). A455bcd9 (talk) 09:46, 5 December 2022 (UTC)


 * I agree about delineating between the two questions you pose above. My intent is for this RfC to focus on the latter of those two questions.  Onel 5969  <i style="color:blue">TT me</i> 14:46, 8 December 2022 (UTC)
 * Yes but option 3 is "Promoting WP:MAPCITE from an essay to a guideline, and include either #1 or #3 above?" and WP:MAPCITE also deals with user-made maps. A455bcd9 (talk) 08:56, 9 December 2022 (UTC)
 * I see what you're saying, so mentioning MAPCITE would necessitate bringing combining the two concepts. Unless MAPCITE is specifically focused itself.  Not sure how to word that in an RfC.  Onel 5969  <i style="color:blue">TT me</i> 11:09, 9 December 2022 (UTC)
 * I think MAPCITE is far from being ready to be promoted as a policy and I would ditch option 3.
 * Also, I don't know what "include either #1 or #3 above?" means: what are #1 and #3? A455bcd9 (talk) 11:37, 9 December 2022 (UTC)
 * I think those numbers got mixed up in the switcharoo of the various options. I agree that MAPCITE is not ready for policy yet, but perhaps we can fix it up so that it may be less opinion-y and more present-active-tense declarative to make it more worthy of that potential "upgrade". - Floydian τ ¢ 18:57, 9 December 2022 (UTC)
 * It may be easier to start from MOS:MAP instead of MAPCITE. MOS:MAP looks better. A455bcd9 (talk) 08:12, 10 December 2022 (UTC)

Timing
Considering the late date on the calendar I hope this won't start until after January 1. --Rschen7754 20:57, 18 December 2022 (UTC)