Talk:Strategy/Wikimedia movement/2018-20/Recommendations/Innovate in Free Knowledge

From Meta, a Wikimedia project coordination wiki
Jump to navigation Jump to search

A working prototype: Wikispore[edit]

Wikispore logo.svg

A few of us in the Wikimedia movement have been working on a prototype very much along these lines since a semi-launch at Wikimania 2019, called Wikispore. Our concept especially synchs with the "Expected outcomes" in this recommendation:

  • Create policies for continual experimentation with projects of various scales, measured by content usability for various audiences and obtaining more equitable coverage of knowledge.
  • Create spaces for continual experimentation with projects based on multiple types of knowledge, content formats, and devices.

The aim of Wikispore is to lay the groundwork for a fruitful wiki farm where contributors can germinate anything that is interesting, factual, and non-privacy-violating. If a topic can have a Wikidata item, it can have a narrative page too, and the multiple “spores” on the project are experimental places to implement new standards on sources (including oral knowledge) and notability for different subject areas. And we have a live experimental space too, currently hosted on Wikimedia Cloud Services, at https://wikispore.wmflabs.org --Pharos (talk) 21:13, 21 January 2020 (UTC)

More like this! We need a wiki sphere in which people can innovate at the Project level. –SJ talk  17:58, 4 February 2020 (UTC)

I'd add that it also creates spaces for experimentation with different workflows, community norms, conversation norms, policies, tools; which, if successful, might get reused by the established projects. There is clearly some frustration with some of the established norms, but it's also clear that changing the norms of large, successful projects to ones that have not been battle-tested is a dangerous idea. Our stagnation is largely a consequence of our failure in providing a testing ground for new ideas. --Tgr (talk) 19:59, 22 February 2020 (UTC)

Community Decision, not Consultation - and local communities[edit]

Consultations are sometimes listened to, and sometimes not - this recommendation has one of the best (indeed, one of only a couple) set of improvements over the previous consultations. The "Knowledge as a service." section is great.

The notability issue is the killer - I, and almost all others, will be firmly against it if we think notability rules might be changed despite the community's wish. I used a small "c" there deliberately - local Wikipedian communities (I can't speak for other projects) have always been able to choose their own rules on notability, reliable sourcing etc.

So two suggested changes:

  1. Change it to a "Community decision", not consultation
  2. State that local communities will retain control of their local sourcing rules, as they have for 19 years.

As a side note, I'm not sure why, but the conversation about this aspect has merged onto the previous recommendation's talk page (the one that wants the WMF to specify which content is less important and which is more important to humanity), worth a read. Nosebagbear (talk) 11:18, 23 January 2020 (UTC)

@Nosebagbear: Thanks for your message. Notability is an essential rule for Wikipedia to work, but at the same time, we know certain interpretations of it are limiting specific kinds of content (in certain areas or about certain topics, e.g. women biographies). It is important to raise this issue because the solution to it must come with consensus, whether it is a different interpretation of Notability or a side-project dedicated to specific topics. Your two other suggestions about Community decision/Consultation make sense. I think it is a matter of semantics, as every change must be implemented by the community (there is no overrule). On the other point about Prioritizing Topics, I believe it is not the WMF role to dictate what content is important or less important, but to have tools and research that illuminate the content gaps and the effects of content on society. Any affiliate can use them to create programs if they wish. --Marcmiquel (talk) 00:07, 28 January 2020 (UTC)
  • People have a tendency to think that English Wikipedia has particularly difficult notability standards. In fact, enwiki's general notability standards are significantly lower than those of many other Wikipedia projects; there are tens of thousands of enwiki articles that don't even come close to meeting notability standards on dozens of other projects. There are widely different notability standards on non-Wikipedia projects that bear no relation to (any) Wikipedia notability standard. This is one of the areas where imposed global expectations conflict with community self-governance and self-determination. On more than one project, the "global" pressure to include more biographical articles of women has resulted in communities going for the low-hanging fruit and translating articles of Western celebrities and porn actresses, rather than the much more complicated articles about women scientists, businesswomen, philosophers, and so on. We haven't done ourselves any favours there. There is also a pretty big problem with lowering notability in areas such as health-related topics, where folk remedies, fraudulent products, and "alternatives" have little or no scientific backing and are dependent on word-of-mouth and sales pitches; they do not belong in a contemporary encyclopedia and we would be putting both the public and the project at risk if being pressured to include some of these things because we've lowered notability standards. (Example: there are hundreds of herbs, roots, barks and other "natural products" that are reported to "cure" venereal diseases. None of them actually do, except - cynically speaking - for the ones that are likely to kill the consumer.)
    There is a lot to be said in favour of developing one or more new projects that accept oral histories, personal recollections, historical legends used to explain natural phenomena, and so on. They aren't necessarily encyclopedic, but they have value, and projects that include this kind of information shouldn't be pressured to meet encyclopedic notability standards. But trying to force Wikipedias to include that kind of information does not even meet the expectation of "knowledge as a service". Encyclopedias are supposed to be based on facts, not legends. Risker (talk) 19:53, 11 February 2020 (UTC)
    • Hi Risker, thanks for your feedback. Input from seasoned Wikipedians like you and others who have participated on these talk pages is appreciated and captured to help refine the recommendations and frame them better. I know that when the writers were working away at this recommendation and Prioritize Topics for Impact, there were no thoughts or inklings of things being imposed. Rather a desire to systematically study our impact across the world, especially in regions where we want/need to grow, and to make space and reach out to those we have left out. Notability and reliable sourcing would not change without considering what's best for a project community with that community. It will mean different things in different contexts and at different times for a 10 year movement strategy, and without conflicting with self-determination. We really need to make the principles more prominent. It's been a long time since new projects have been discussed and I am glad to read your thoughts around that. I know for a fact that the ethos of these recommendations encourage analyzing our impact so we don't focus on quantity, but also quality and reach. Although, we shouldn't underestimate the sway female and male porn actors hold globally, especially around venereal diseases ;) I'm being cheeky. If such articles already exist, then studying our impact around the world coupled with evaluation can help improve the situation and our approach. There is also much to be said for mentorship and the potential for encouraging complex work and improving quality, to make it all less complicated. Your knowledge of the movement and shared accountability for encyclopedic standards resonate with many who have contributed to these recommendations. Nothing will be forced on projects, no such intent ever existed in the discussions or among these pages. We should make that more explicit and thanks for helping us do that. Thank you very much for this feedback. MPourzaki (WMF) (talk) 02:03, 15 February 2020 (UTC)

New kind of project[edit]

Hello all. As a co-founder of Lingua Libre, contributor about french minority languages (around 90+) and French Wiktionary Admin, I support the part of this recommendation that could lead to create en adapted project for small linguistic communities. For 7000 languages (estimation in decreasing) on Earth, only 3500 are described, 1500 are written (maybe more) with 150 kinds of alphabets (include ideograms, etc.). Lots of these communities are small and giving them a Wikipedia project is like to give a international liner to a fishermen's village. We need to create a portal project, with a welcoming minimalist interface (easy to translate), where a small community can do what they want to describe their knowledge. If they want an encyclopedic part, they can, if they want a library part, they can, if they want a dictionary part, etc…

The first contact with these kind of "all included" projects is to change our « can you contribute to our big project » position to a « what do you need to describe your knowledge? » position. And, if the description of one community becomes so big that they need a whole Wikimedia project, we can switch the concerned part to the wikimedia galaxy. These reflections can be seen like an overhaul of incubator, and as a matter fact, there are lots of small linguistic communities present inside the incubator. Thanks for reading me, and thanks for your recommendation. Lyokoï (talk) 15:44, 23 January 2020 (UTC)

@Lyokoï: It is a great pleasure to read this message. I personally believe that oral projects can be positive, not only to capture the knowledge of minority or oral languages but also to complement any other language and provide learning materials in other formats than the written text. This is what we wrote in the recommendation "Wikioral". The recommendation "Innovate in Free Knowledge" wants to be an open call to the whole Movement to start innovating and creating new ways to capture knowledge, either by changing the current projects or by creating new ones. I think your approach to understanding people and "what do they need to describe your knowledge" is very appropriate to create the environment in which new projects can flourish. Thanks your message again and for your project Lingua Libre. --Marcmiquel (talk) 23:46, 27 January 2020 (UTC)
I was going to write about a new project idea and then saw this. Thank you User:Lyokoï for writing this wonderful piece and User:Marcmiquel for furthering a project idea. If I can add,
  1. the project has to be centered around the oral language speakers in the first place i.e. multiple ways to capture oral knowledge with low use of text, use of keywords, hashtags and other identifiers that will make the content searchable. This also means that the current structure of Commons which don't have a simple and user-friendly interface, and most importantly, the hassle one has to go through to upload an image/audio/video file from a phone, HAS TO CHANGE. It's a nightmare for an experienced user and even they can't get around the complications while using a phone. It's a bottleneck right there that the privileged users tend to not discuss enough. Many native speakers of indigenous languages that I work with are from developing countries where bandwidth and lack of other infrastructure is a huge challenge. Mobile-friendly interfaces and easiness with multimedia files are a must. People get on Tik Tok and would never get on any Wikimedia platform for such entry-level barriers.
  2. Oftentimes, the decisions about platforms and processes are defined keeping more privileged users (for starters, tech-savvy, English/other major European/largely spoken language speaking, access to a computer and decent internet bandwidth, etc.) in mind. The proposed platform and its policies MUST include all the challenges the least privileged contributor would face and what can be done to address those issues.
  3. ...adding more here soon. --Psubhashish (talk) 07:56, 15 February 2020 (UTC)

Feedback from French-speaking Wikimedians[edit]

  • Strong support was expressed from co-founder of Lingua Libre and Wiktionary admin Lyokoï

See feedback links and context here.

DRanville (WMF) (talk) 16:41, 28 January 2020 (UTC) - French Strategy Liaison

Primary sources/Original research[edit]

There is a potential clash here with our longstanding aversion to primary sources/original research. I suspect that the only way that will be well addressed is through a new project or projects, distinct from Wikipedia, WikiSource, Commons, etc. For example, there is no way we bring oral cultures into our "ecosystem" without either using audio/video direct from participants in that oral culture or original research by others who put related content into a written form, which constitutes original research.

Parallel to this, and possibly using some or all of the same mechanisms, is a proposal I have jokingly called "WikiAffidavits": the ability of people about whom we have articles (they don't have to be the main subject of the article, just directly involved), and who feel that the formally citable sources get the story wrong, to be able to tell their own version, either orally or in writing, allowing them in turn to be cited for their view, since we would then be able to quote them reliably and with context. This actually came up in a relatively trivial context: rock musicians saying that the written record (really surprisingly sparse except for very major bands) got basic facts wrong (wrong dates; exaggerated versions of stories; bias from a situation where one person involved wrote a book, but no one else did; etc.) - Jmabel (talk) 17:56, 1 February 2020 (UTC)

Early idea of dedicated project support[edit]

I realise, of course, that the actual discussion on this will be held in the implementation stage, but when I read the latter half, I get a strong feeling of "this would be a nightmare for Wikipedia, but would work quite well as a dedicated project" - some other comments allude to sharing this viewpoint, but I was interested in gauging the general position. Do other viewers think that a new project is also the way to go? Nosebagbear (talk) 10:47, 3 February 2020 (UTC)

Better, not fewer sources[edit]

The recommendation seems to suggest that the sourcing requirements of (English?) Wikipedia are an obstacle to knowledge equity, and policies should be put in place to get around such notions as notability. I think that's a mistake. The only reason Wikipedia is even a little bit trustworthy is because we have sourcing requirements. We need better tools to support discovery of sources, and we tools to uncover content farms and other sources of disinformation. There is a lot we could do to help our readers understand which sources can be trusted, and innovating in those areas would be immensely helpful. Relaxing our standards would be detrimental to our relationship with out readers. Vexations (talk) 14:04, 3 February 2020 (UTC)

@Vexations: this somewhat begs the question of whether our relationship with our readers is what we think it is (Wikipedia has, famously, displaced traditional encyclopedias well before it had any claim to reliability, ie. readers have chosen it over far more trustworthy sources; many smaller Wikipedias are similarly popular in their own language community, despite not being particularly reliable; and we know that only a minuscule fraction of readers actually interact with the source footnotes), but even granting that readers care a lot about trustworthiness, there are obvious ways to manage it without discarding less-then-maximally trustworthy content, by clearly explaining to the reader that due to the lack of reliable sources the given article is based on less reliable ones and should be treated with some doubt. (That would be quite useful even with the current sourcing requirements, which, let's be honest, aren't really applied consistently. The WP 1.0 effort was supposed to produce such a "reliability rating", but it has fallen by the wayside.)

After all, consider what happens when a reader does not find the topic they have been searching for in Wikipedia, because it has been deemed non-notable - in most cases they will just look for some other place that does have information on it; and if that place has even lower standards than what Wikipedia could manage with the available sources (which seems likely), everyone will be worse off. --Tgr (talk) 19:49, 22 February 2020 (UTC)

Please differentiate between Wikipedia and other Wikimedia projects[edit]

The "why" part of this recommendation seems ambiguous in the use of Wikipedia and Wikimedia. Wikipedia is an encyclopedic project, limited by that form. An oral history project or one based on "folk lore" - in the original meaning of the word - seem interesting and valuable endeavors for many cultures and languages, that lack written, academic or other established material. But they are not an encyclopedia and should not happen under the Wikipedia label. Please feel encouraged to grow communities of volunteers that are interested to grow such knowledge in new projects under the Wikimedia brand and umbrella. --h-stt !? 17:08, 6 February 2020 (UTC)

Agreed. A distinct separation would help to get more clarity and to address worries raised by the community. Alice Wiegand (talk) 18:53, 6 February 2020 (UTC)

Focus on the pathway for new projects[edit]

"Community consultation" is a bad way to think about how to interact with enwiki on notability. Enwiki has already plenty on community consultations around notability. If there's a desire from the WMF to influence notability in enwiki it should provided engage with the existing RfCs and provide research based arguments of why it makes sense why one should decide for wider notability guidelines in particular cases. The idea that it's the role of the WMF to do community consultations about whether enwiki should widen their notability guidelines is just going to lead to unnecessary tension in our community. Providing a pathway to create new projects is a better focus. Langcom's way of blocking the creation of new Wikiprojects should end and be replaced by a more democratic procedure. ChristianKl❫ 08:51, 10 February 2020 (UTC)

Free Knowledge in Portuguese Wikipedia???[edit]

Hey everybody,

What kind of Free Knowledge exists if a robot https://pt.wikipedia.org/wiki/Usu%C3%A1rio:%C3%89rico blocks me because I told this kind of action in candidature pages??? This robot invaded my account and created 32 guarded pages????

The free knowledge??????????? I denounced this fact to user LTeles https://pt.wikipedia.org/wiki/Usu%C3%A1rio:LTeles_(WMF), he deleted my post and the same robot blocks me for three days? It is a criminous organization? https://pt.wikipedia.org/wiki/Usu%C3%A1rio:Douglasboavista

They deleted my comments on cadidature pages: https://pt.wikipedia.org/wiki/Wikip%C3%A9dia:CheckUser/Candidaturas/EVinente/4#Coment%C3%A1rios https://pt.wikipedia.org/wiki/Wikip%C3%A9dia:Administradores/Pedidos_de_aprova%C3%A7%C3%A3o/79a#Coment%C3%A1rios https://pt.wikipedia.org/wiki/Wikip%C3%A9dia:CheckUser/Candidaturas/Editor_D.S#Coment%C3%A1rios

This man got out of own candidature https://pt.wikipedia.org/wiki/Usu%C3%A1rio:Douglasboavista

Free knowledge???? I guess I prove the terrorism of the wiki-structure. We need KNOWLEDGE AS UNIQUE CRITERIA.

Please, does anyone help me????????????????--Camillo Cavalcanti (talk) 00:29, 14 February 2020 (UTC)

Feedback from Hindi Community members on Innovate in Free Knowledge recommendation[edit]

There was support on the development of new content formats for more inclusive cultures within the editor community and readers. There was a strong support for policy revision, especially around policy revision of notability criteria for better integration and inclusion of oral knowledge systems and indigenous knowledge systems to be included in Wikimedia projects. It was mentioned that this can improve efficiency in the community to learn and engage with projects. RSharma (WMF) (talk) 19:23, 16 February 2020 (UTC)

Feedback from Office Hours from 2 February and 9th February with members from Emerging communities from South Asia and Africa[edit]

Anass Sedrati (Strategy Liaison - Arabic) and Rupika Sharma (Strategy Liaison - Hindi) hosted Office hours on two Sundays (February 2nd and 9th) to discuss with members of Emerging communities from South Asia and Africa. Overall, around 50 people in total participated in either of the calls, representing various areas and regions.

Regarding recommendation 11, it was stronlgy supported by participants. There was an overall support for inclusive knowledge systems to address content and technology gaps for different projects. Inclusion of Oral Knowledge was in particular supported.

As suggestions for improvements, the participants mentioned:

  • There should be improved and improvised policies for projects and revision of existing notability criteria in Wiki projects and inclusion of non traditional system of knowledge.
  • There should be user friendly mobile application interfaces to make it easy for editors to contribute easily.

Regards -- ASedrati (WMF) (talk) 21:27, 16 February 2020 (UTC)

Feedback from discussion in the Netherlands[edit]

There has been a discussion about the Movement Strategy 2030 during a Wiki-Saturday organised by Wikimedia Nederland. A summary (in English) can be found here. This comment relates specifically to innovate in free knowledge:

Oral history

Currently, the use of oral history as a source does not fit the narrow definition of encyclopedicity. This excludes a lot of knowledge from Wikipedia and that is a pity. This mainly concerns knowledge of groups that are already traditionally excluded. Our job is to represent reality, not the truth. The question should not be whether we want to be able to use sources of this kind, but how we will make that possible while maintaining reliability and neutrality. To explore the possibilities of oral history within the Wikimedia projects, it is best to start a new sister project. The step to integrate this in one go in Wikipedia is too big. The aim is also to develop criteria for oral history. What do you use oral history for, where do you record it, what are the rules? The license used will also have to be considered.

During a discussion about the recommendations in De Kroeg (the Village Pump) on NL Wikipedia some concerns were expressed:

  • Some editors saw a desire to shift from objective and neutral writing to activism. There is some room for activism on Wikipedia.; it can be a way to create new content. However: Wikipedia is primarily an encyclopedia; neutrality, verifiability, etc. are core principles on which the identity, image and influence of Wikipedia is based. Too much activism undermines this.
  • There was concern about stretching criteria to be able to include previously excluded knowledge. It is important to fill gaps but essential to monitor encyclopedic value and reliability

Sandra Rientjes (talk) 11:05, 18 February 2020 (UTC)

Tabular data[edit]

Pretty words about experimentation with different types of knowledge, and different formats of content.

But will WMF actually back them up with engineering resources to make them work, and transparency about reasons for prioritisation and go/no-go decision making?

Case in point. In 2016 User:Yurik created the Data: namespace on Commons for tabular data (Announcement).

This has the potential to be incredibly valuable -- eg for storing the data behind graphs in an accessible, reusable way; plottable, and downloadable in a range of standard formats. All sorts of processes can generate such tabular data, which it is useful to be able to store -- for example, the coordinate matchings used to map a historical scanned map onto present-day latitudes and longitudes, to allow the old map to be 'warped' over a present-day map. (As I experimented with at Wikimania 2019). More generally, we might think of time-series of relevant economic and social data, a somewhat important part of the "sum of all human knowledge" that it would be good to make rather more accessible and reusable.

There are some quite interesting things that have been done with the Data: namespace -- see for example this recent Commons:VP post on data imported from the "Our World In Data" project. It is also used to store structured shapefiles of boundary coordinates for various places, administrative regions, countries, etc, which can then be picked up and shown in a Wikidata query.

But the Data: namespace is hamstrung because it has no equivalent of a Commons file description area to describe in wikitext what the file is, where it has come from, categories, maintenance templates, etc. There's also no slot to be able to give Commons Structured Data statements about the data. None of these are available, making the files essentially un-discoverable. (See Phab ticket T155290).

Yet WMF refuse to put in any resourcing to fix this. Indeed they don't even seem to engage. There's no discussion about Pros and Cons. The namespace is just left to be hard to use, different to other Commons objects, and un-navigable. (Never mind even think about what could be done -- eg improved editability, or OKFN data package compatibility for uploading/downloading). Before any of that just being able to find and describe the data would be a start. But from WMF nothing.

Informally, at WikidataCon 2019 I was told the attitude was: "We didn't want this. We told him not to do it. He did it. So we fired him". But any transparent on-the-record open-to-assessment discussion about why they think this, or why they don't want to lift a finger to fix it, by giving data objects some proper wiki description pages like any content-blob on Commons -- anything like that seems to be entirely lacking.

So: two cheers for the warm words about experimentation with different types of knowledge, and different formats of content. But until WMF shows itself more ready to be more open regarding discussion and prioritisation of resourcing and support for existing different types of knowledge, and different formats of content, that the platform is not supporting very well, forgive me for thinking these warm pronouncements are just self-serving abstraction, able to mean anything to anybody, when in this very area there has been a lack of concrete resourcing or action. Jheald (talk) 18:07, 19 February 2020 (UTC)

Highlights from the Spanish Speaking Conversations - Innovate in Free Knowledge[edit]

About 70-80 present in conversations at different degrees (not everyone said something, of course).

Due to the proximity to indigenous people's context, the possibility of adapting or creating projects suitable for oral knowledge is weel received.

There are doubts about how to write oral languages, they agree that we lack platforms suitable for them. Existing projects with indigenous communities already showed that there is a problem with oral knowledge.

The whole recomendation is well received and perceived as necessary. People need more clarification or debate about the concept of “Knowledge as a service”.

Writing is regarded as ambiguous when talking about "Create policies for continuous experimentation". Proposal: Design a FOSS (Open Source Software) communities-like organigram where everyone can create or develop ideas freely, with no permissions, intended to solve specific issues. Then, allow them to expose to the community to test and, if successful, implement in wiki-core. Improve sandboxes and create like wix and github, including on it functionality, design, UX, metrics, analytics etc for ideas and suggested projects. This would be good because it would be much easier to test new projects.--FFort (WMF) (talk) 14:38, 21 February 2020 (UTC)

Highlights from the Catalan Speaking (Valencian) Conversations - Innovate in Free Knowledge[edit]

About five people present in different ways, giving feedback via both formal and informal channels. Different backgrounds.

In València we have people very in favor of innovating in new formats (audio) because of the obvious reason of projects carried out from here, on our own initiative, and without anyone saying anything from any superior level (so, in favor).

Innovating in new projects goes against the line of closing projects and / or merging them in Wikipedia, which should make Amical to have a favorable position, as we are in favor of the visibility of every Wikimedia project.


Criticism: Unambitious. It proposes spaces to experiment. If we experiment without spending money, I understand that there should be no complaints.

Inclusion boundaries and the conflict between notability and verifiability are a real problem that needs to be addressed. If communities opt for “hard” verifiability we will find that certain forms of knowledge transmission (oral sources, e.g.) are completely out of the question. A possible proposal would be to collaborate with local institutions and universities, as classic knowledge-generating centers, to generate material that can then be leveraged by Wikimedia projects. What do we do if there are no such structures (indigenous languages, minority groups, various sectors generally illiterate even in the Global North)?

Knowledge as an equity and as a service are concepts that have raised doubts and have never been cleared up (also found in the Spanish-speaking community)

It is said that the current bureaucracy has failed to create new projects. Amical should be in favor of ending bureaucracies. The local community has historically criticized the admissibility limits on content, which have been judged to be too high (es.wiki case), so we support this part of the text too.

Genealogy is currently one of the few red lines regarding inclusion in Wikidata. We perceive as dangerous to open Wikidata-inclusivity to every human being whose existence can be proved. ; it also has problems with the treatment of personal data (there are family family trees where you cannot put the date of birth of certain people until they die, real and frequent example on family websites). We don't want to lower our inclusion limits that much.

More innovation to study in implementation: Commons servers in different countries so we can use different PD-laws in our favor (ex. Servers in Italy so we can take profit of their “simple photography” exception globally and not only locally on it.wiki).--FFort (WMF) (talk) 14:38, 21 February 2020 (UTC)