Grants talk:Project/slevinski/ASL Wikipedia 2-D Font Development for SignWriting/Archive 2

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

Round 1 2017 decision[edit]

IEG IdeaLab review.png

This project has not been selected for a Project Grant at this time.

We love that you took the chance to creatively improve the Wikimedia movement. The committee has reviewed this proposal and not recommended it for funding, but we hope you'll continue to engage in the program. Please drop by the IdeaLab to share and refine future ideas!


Next steps:

  1. Visit the IdeaLab to continue developing this idea and share any new ideas you may have.
  2. Applicants whose proposals are declined are welcome to consider resubmitting your application again in a future round. We ask that you first email projectgrants(_AT_)wikimedia · org to indicate your interest in resubmission so staff can review any concerns with your proposal that contributed to a decline decision, and help you determine whether resubmission makes sense for your proposal.
  3. Check back at the schedule for information about the next open call to submit proposals.

Questions? Contact us.


Aggregated feedback from the committee for ASL Wikipedia 2-D Font Development for SignWriting[edit]

Scoring rubric Score
(A) Impact potential
  • Does it have the potential to increase gender diversity in Wikimedia projects, either in terms of content, contributors, or both?
  • Does it have the potential for online impact?
  • Can it be sustained, scaled, or adapted elsewhere after the grant ends?
6.1
(B) Community engagement
  • Does it have a specific target community and plan to engage it often?
  • Does it have community support?
4.4
(C) Ability to execute
  • Can the scope be accomplished in the proposed timeframe?
  • Is the budget realistic/efficient ?
  • Do the participants have the necessary skills/experience?
5.9
(D) Measures of success
  • Are there both quantitative and qualitative measures of success?
  • Are they realistic?
  • Can they be measured?
6.0
Additional comments from the Committee:
  • I am not sure that it fits Wikimedia's strategic priorities. It will be probably more useful outside the Wikimedia environment. Its online impact is also doubtful as there is a controversy regarding the usefulness of Sutton's system in general.
  • in line with the strategic mission of providing every human with the access to knowledge; quite sustainable but I put just 7 due to the fact that this sign writing is just one of different existing variants; it is not adopted globally and it limits possible scale.
  • It does have a potential for online impact, and is especially good for scaling elsewhere outside Wikimedia projects.
  • The success can be measured but there are risks that few people will really use the developed font or that the software developers will adopt it. The approach is more iterative than innovative.
  • Risks are much greater than potential impact.
  • If proposed characters are not accepted into Unicode, this proposal will fail.
  • If this proposal requires significant changes to MediaWiki core and there are no WMF devs available, it will fail.
  • It is not the only font for sign language, and there is no proof community will prefer this font over any others.
  • It is not clear this will help build sustainable community/ies for sign language Wikipedia/s.
  • The project and its budget is essentially a sum of two unrelated projects: the font development and training. The budget is probably realistic except for the training part which heavily depends on success of the first part.
  • Realistic, the grantees have worked on this topic for years and have tried to create sign Wikipedia since 2013/2015.
  • Not clear. Major part of the budget is dedicated to development of Unicode characters and is poorly explained.
  • There is some community engagement and community support. And , yes, it does support diversity.
  • Plus for diversity and minus for the lack of endorsements: I hoped to see more involvement of other editors and readers. Yes, we have letters of intentions, but I don't see much editors/active users except from grantees who work on sign wikipedia.
  • There is no large public discussion or analysis.
  • On one hand, it is very good for diversity. On the other hand, it looks very much like selling the system to a community, not vice versa.
  • I don't see an impact for Wikimedia projects. Basically I agree with the comment of Blue Rasberry. The development of these fonts make sense if there will be a sign language Wikipedia, but at the moment the outcome seems to me to be only to advantage the Sutton Foundation.
  • Given the budgetary constraints we have, I'm recommending that we only fund the technology portion. A separate project should be funded for the training portion.
  • I'm concerned about the focus on VOLT and would like to see more focus on adapting Unicode with the needed characters. Note that this comment in phabricator would have helped, some.
  • After reading more about the schedule, I think this should be focused on making MW work with the font and then a second project should handle the education and training bit.
  • Education should focus on editing the new ASL or Brazillian Sign language wiki.
  • Unfortunately I cannot recommend it for funding. First, it does not seem to be strongly related to the Wikimedia priorities. Second, the usefulness of the Sutton's system is not without doubt.
  • In case the font development is funded it should be at least separated from the training. Only after the font is developed, tested and incorporated into the software should further funding be provided for any training sessions.
  • The impact is quite limited, the writing itself isn't globally adopted and I see lack of endorsement: though it is a real step to diversity and accessibility, I prefer not to fund this request.
  • Big costs of Font Developer: 50,000 USD.
  • No serious analysis showing that people are interested in editing or improving ASL Wikipedia.
  • 17 new characters have not yet been officially proposed to Unicode, etc.
  • I don't think we are a good funder for this. This is not THE system, this is not THE standard and risks seem too high.

Round 1 2017 response[edit]

Thanks for reading and commenting on the grant. I appreciate the insights and concerns. I will be updating the grant request for round 2 2017. On the grant page, I have already revised the budget. I plan on rewriting the rest of the text to address issues and clarify the intent.

Below I have organized the committee comments into groups and I have provide our responses indented. Committee comments are bold. The comments are sometimes contradictory and range from positive to negative. For a few comments I used ... to indicate that I cut off the beginning or ending of a comment.


Strategic Alignment[edit]

  • Given the budgetary constraints we have, I'm recommending that we only fund the technology portion.
The project has been updated and now only includes the font technology portion.
  • in line with the strategic mission of providing every human with the access to knowledge; quite sustainable but ...
We agree. The Wikimedia Vision statement is "Imagine a world in which every single human being can freely share in the sum of all knowledge. That's our commitment." We believe this overlaps with our vision of sign language users sharing in the sum of all knowledge.
The mission of the Wikimedia Foundation is to empower and engage people around the world to collect and develop educational content under a free license or in the public domain, and to disseminate it effectively and globally.
Again, we believe there is tremendous overlap with the Wikimedia Foundation's mission and our goals for SignWriting.
  • Realistic, the grantees have worked on this topic for years and have tried to create sign Wikipedia since 2013/2015.
Sign Language Wikipedias are possible on Wikimedia Incubator today. This is a long term project that has been a priority for the Center for Sutton Movement Writing. With our technology, we have made very significant progress. We have a stable ASCII version, used on Wikimedia Incubator today. This grant is to help develop and support the Unicode version for the future. Regardless of ASCII or Unicode, MediaWiki integration will require HTML and CSS. The Unicode version requires less boilerplate and a simplified design, but Unicode also includes a set of drawbacks and complications.
  • There is some community engagement and community support. And , yes, it does support diversity.
  • Plus for diversity ...
  • On one hand, it is very good for diversity.
  • ... though it is a real step to diversity and accessibility, I prefer not to fund this request.
  • I am not sure that it fits Wikimedia's strategic priorities.
  • ... it does not seem to be strongly related to the Wikimedia priorities.
We believe supporting technical SignWriting development fits with both the Wikimedia Vision and the Wikimedia Mission.
Regarding Wikimedia development priorities, the support of vertical writing mode in MediaWiki will probably happen in the next year or two.


Unicode Support[edit]

  • It does have a potential for online impact, and is especially good for scaling elsewhere outside Wikimedia projects.
We have very high hopes for the Unicode form of SignWriting. The 2-D font will simplify design and encourage a common format that everyone can share.
  • It will be probably more useful outside the Wikimedia environment.
Wikimedia Incubator currently uses the ASCII / SVG format of SignWriting, so the 2-D font and Unicode are not required to keep working on Incubator. Looking towards the future, the Unicode form of SignWriting will be vital when moving the first sign language Wikipedia into production. The 2-D font and Unicode combination will dramatically simplify the structure of the HTML, CSS, and associated gadgets.
Unicode for SignWriting will allow the Wikimedia projects to freely share information with sign language users around the world. The Unicode form of SignWriting is probably a better standard to build an international community than ASCII.
Outside of Wikimedia, the 2-D font will be very useful. For example, the 2-D font will be essential for the LibreOffice integration. Earlier this year, LibreOffice announced vertical writing mode support.
  • Not clear. Major part of the budget is dedicated to development of Unicode characters and is poorly explained.
  • 17 new characters have not yet been officially proposed to Unicode, etc.
We will refocus the grant and expand on the Unicode background and strategy.
  • I'm concerned about the focus on VOLT and would like to see more focus on adapting Unicode with the needed characters.
VOLT will be used to produce the 2-D font. The 2-D font will be the cornerstone of the Unicode proposal. We will expand and clarify the grant to focus more on Unicode.
  • If proposed characters are not accepted into Unicode, this proposal will fail.
I recently submitted an informational document to the UTC about character options for SignWriting: the optimal set and the compliant set. The optimal set overwrite the Sutton SignWriting block and adds plane 4. The compliant set adds 17 new characters. I have created 1-dimensional fonts for both sets. The font for the optimal set works in most software, even as characters for file names and folder names for Linux and Mac OS. The font for the compliant set does not work in most situations and will need to be rewritten as font technology improves and as font rendering support improves. Any font created to support the compliant set requires font support over 16-bit. There is not one standard way to support fonts above 16-bit.
A subgroup of UTC members and experts recommend the UTC rejecting both character design options. They recommend using the currently available characters for the symbols within an XML markup similar to Musical XML notation. If a SignWriting word can only be written using XML, how do we link to word definitions? There are several deal breakers with the Unicode 8 characters and with the UTC support of SignWriting. Making progress with the Unicode standard is a long term project. Without the support of a UTC voting member and without a large established user base, it will not be possible to get the Unicode standard in line with the character use of the community.
With Formal SignWriting, every sign is written as a word. Each word is written as a string. When the string is written as ASCII, it requires an active transformation into SVG for viewing. When the string is written as Unicode, we can properly use fonts to view the sign with a system transformation, rather than an application - the same as other scripts. When strings are written with Unicode characters, we can have two views of the characters. The first is the proper 2-dimensional view created from a system font. This 2D font is the topic of this grant. The second is a 1-dimensional representation of the parts of the word. I have created the 1D fonts for both the option 1 characters and the option 2 characters. With this grant, I will create 2D fonts for both the optimal set and the compliant set.
Unicode is a long term and complicated goal. SignWriting has had many data formats including binary, comma delimited, XML, ASCII string, and Unicode string. A Unicode string is the right solution for a SignWriting word. UTC members have supported the idea that SignWriting words should be encoded in Unicode. They have also expressed the idea that a conversion routine will be possible for our current data.
Slevinski (talk) 15:42, 23 August 2017 (UTC)
  • If this proposal requires significant changes to MediaWiki core and there are no WMF devs available, it will fail.
MediaWiki core will not be involved with supporting SignWriting regardless of ASCII format or Unicode format. Vertical Writing Mode will hopefully be supported in MediaWiki soon. Moving a sign language Wikipedia to production will be a real concern. While it is possible to continue with the ASCII form of SignWriting, development will be simplified by using the Unicode form of SignWriting.

Other Writing Systems[edit]

  • It is not the only font for sign language, and there is no proof community will prefer this font over any others.
There are several sets of fonts available for SignWriting. Only the SignWriting 2010 fonts and the Sutton SignWriting fonts support the SignWriting in Unicode design. Outside of SignWriting, there are no fonts available for any other 2-D sign language script. The community can only use something that's available.
  • quite sustainable but I put just 7 due to the fact that this sign writing is just one of different existing variants; it is not adopted globally and it limits possible scale.
There are other writing scripts for sign language. Some are 1-dimensional and others are 2-dimensional. SignWriting is definitely and unapologetically a 2-dimensional scripts. SignWriting is a mature script that is used internationally. SignWriting has been adopted by a few language communities around the world, especially in Brazil. SignWriting with a 2-D font and full Unicode can scale globally for any and all sign languages. Being the first sign language Wikipedia in production will require much more work that being the fifth. Our technology ark is towards simplification, unification, and standardization. Things are getting easier and better. All trends point to increased usage and adoption.
SignWriting is leading the way for 2-dimensional scripts. All of the tools used to create the SignWriting fonts have been released under open licenses. Any success in this project will enable future projects. Lurking on the Language Committee mailing list, I've found that there is some controversy regarding language communities and script choice. Sometimes there is an easy conversion between the two scripts and other times the script are mutually exclusive. These issues are for the future and will need to be handled based on the individual language, the scripts available, and the associated user communities.
There is some understanding and cooperation between the SignWriting users and the ASLwrite users. It may be possible to cross-index the two scripts to a certain degree. Any advancement or achievement for SignWriting will raise the state of the art for 2-dimensional scripts.

Brazilian Training[edit]

The training portion has been split off into a separate grant: Brazilian Sign Language Wikipedia Project.

  • The project and its budget is essentially a sum of two unrelated projects: the font development and training. The budget is probably realistic except for the training part which heavily depends on success of the first part.
  • In case the font development is funded it should be at least separated from the training. Only after the font is developed, tested and incorporated into the software should further funding be provided for any training sessions.
The grant now focuses on the font development and Unicode proposal. The training has been removed from the grant.
  • A separate project should be funded for the training portion.
  • Education should focus on editing the new ASL or Brazillian Sign language wiki.
The education portion will be split from this font development grant. For a possible future training grant, the education portion will focus on editing sign language wikis and not teaching SignWriting. The 20+ schools and universities in Brazil interested in a sign language Wikipedia already use SignWriting in the classroom.
  • After reading more about the schedule, I think this should be focused on making MW work with the font and then a second project should handle the education and training bit.
This grant will primarily focus on the font development and Unicode proposal. The MediaWiki work will be a secondary concern in this project. Depending on the work involved, the MediaWiki work will be carried out in tandem with the font development. If the MediaWiki development is too extensive, the customization will either need to be carried out by the community or a future grant.

Risk and Other Comments[edit]

  • There is no large public discussion or analysis.
  • It is not clear this will help build sustainable community/ies for sign language Wikipedia/s.
  • Plus for diversity and minus for the lack of endorsements: I hoped to see more involvement of other editors and readers. Yes, we have letters of intentions, but I don't see much editors/active users except from grantees who work on sign wikipedia.
  • No serious analysis showing that people are interested in editing or improving ASL Wikipedia.
  • I don't see an impact for Wikimedia projects. ... The development of these fonts make sense if there will be a sign language Wikipedia, but at the moment the outcome seems to me to be only to advantage the Sutton Foundation.
  • On the other hand, it looks very much like selling the system to a community, not vice versa.
  • The success can be measured but there are risks that few people will really use the developed font or that the software developers will adopt it. The approach is more iterative than innovative.
  • Risks are much greater than potential impact.
  • The impact is quite limited, the writing itself isn't globally adopted and I see lack of endorsement: though it is a real step to diversity and accessibility, I prefer not to fund this request.
  • the usefulness of the Sutton's system is not without doubt.
  • Its online impact is also doubtful as there is a controversy regarding the usefulness of Sutton's system in general.
  • I don't think we are a good funder for this. This is not THE system, this is not THE standard and risks seem too high.
  • Big costs of Font Developer: 50,000 USD.


Eligibility confirmed, round 2 2017[edit]

IEG review.png
This Project Grants proposal is under review!

We've confirmed your proposal is eligible for round 2 2017 review. Please feel free to ask questions and make changes to this proposal as discussions continue during the community comments period, through 17 October 2017.

The committee's formal review for round 2 2017 begins on 18 October 2017, and grants will be announced 1 December. See the schedule for more details.

Questions? Contact us.

--Marti (WMF) (talk) 19:12, 5 October 2017 (UTC)

Comments of Glrx[edit]

I would decline this proposal. The problems raised in the first round have not been addressed (see archives). There is no quantitative description of Sign Writing in comparison to other sign writing systems. There is still no evidence of authors doing much publishing in any sign writing system. The situation is not one where there is significant SW publishing. We are not told how many signers know SW. Even publishing videos of signings is rare. There is no quantitative discussion about experience with Wikipedia style publishing efforts. In the last round, material supplied by the proposers showed there was little growth in such SW publication when tools were made available. Independent references about SW are missing. Adrean Clark's comment is not a ringing endorsement. Reddit is an even less reliable source, but it has the comment "I've never really tried to learn [si5s]. With video recording so widely available, I've never found a need to write ASL down." WMF has little expertise in this area. Better targets for funding Sign Writing R&D would be NIH or the Brazilian government. The Unicode troubles make success even less likely. "Unicode is a long term project that will require years to complete" does not bode well for WMF. The idea of replacing existing Unicode assignments is both radical and frightening. The project is not a good fit for WMF, project details are lacking, and the project does not seem ripe. Glrx (talk) 16:40, 9 October 2017 (UTC)

Proposal cites Adrean Clark at Quora: https://www.quora.com/How-is-SignWriting-used-when-written-by-hand

On that webpage, Prof. Donald Grushkin comments on Clark's post:
I actually agree with you — I feel that signwriting is too heavily drawing-dependent, and even if you're not exactly “drawing” with it, there's still a whole lot of shading going on, which slows down the process of writing significantly. Plus not all of us are good at drawing — I certainly am not. That's why I advocate an alphabetical approach to writing signed language.
Michele Westfall also comments on Clark's post (and is upvoted by Grushkin):
Keep in mind that Valerie Sutton is a hearing person who came up with this system in 1974, and as far as I can see, it hasn’t been adopted by the Deaf community (despite her claims in Wikipedia and her website). And to me, that’s about as definitive as it gets. We Deaf people decide what works for us and what doesn’t.

So I look at Grushkin:

Donald Grushkin, Professor at California State University, Sacramento. Contact information at http://www.csus.edu/coe/profiles/grushkin-donald.html

He's written an article

  • Grushkin D
  • Writing Signed Languages: What For? What Form?
    https://www.ncbi.nlm.nih.gov/pubmed/28238971 Am Ann Deaf. 2017;161(5):509-527. doi: 10.1353/aad.2017.0001.
    Abstract
    Signed languages around the world have tended to maintain an "oral," unwritten status. Despite the advantages of possessing a written form of their language, signed language communities typically resist and reject attempts to create such written forms. The present article addresses many of the arguments against written forms of signed languages, and presents the potential advantages of writing signed languages. Following a history of the development of writing in spoken as well as signed language populations, the effects of orthographic types upon literacy and biliteracy are explored. Attempts at writing signed languages have followed two primary paths: "alphabetic" and "icono-graphic." It is argued that for greatest congruency and ease in developing biliteracy strategies in societies where an alphabetic script is used for the spoken language, signed language communities within these societies are best served by adoption of an alphabetic script for writing their signed language.

Grushkin's article provoked debate:

  • Rosen RS, Hartman MC, Wang Y.
  • "Thinking-for-Writing": A Prolegomenon on Writing Signed Languages.
    https://www.ncbi.nlm.nih.gov/pubmed/28238972 Ann Deaf. 2017;161(5):528-536. doi: 10.1353/aad.2017.0002.
    Abstract
    In his article in this American Annals of the Deaf special issue that also includes the present article, Grushkin argues that the writing difficulties of many deaf and hard of hearing children result primarily from the orthographic nature of the writing system; he proposes a new system based on features found in signed languages. In response, the present authors review the literature on D/HH children's writing difficulties, outline the main percepts of and assumptions about writing signed languages, discuss "thinking-for-writing" as a process in developing writing skills, offer research designs to test the effectiveness of writing signed language systems, and provide strategies for adopting "thinking-for-writing" in education. They conclude that until empirical studies show that writing signed languages effectively reflects writers' "thinking-for-writing," the alphabetic orthographic system of English should still be used, and ways should be found to teach D/HH children to use English writing to express their thoughts.
  • Moores DF
  • Writing Signed Languages: What For? What Form? A Response.
    https://www.ncbi.nlm.nih.gov/pubmed/28238973 Am Ann Deaf. 2017;161(5):537-539. doi: 10.1353/aad.2017.0003.
    Abstract
    In his article in an American Annals of the Deaf special issue that also includes the present article, Grushkin divides his discussion of a written sign system into three basic parts. The first presents arguments against the development of a written form of American Sign Language; the second provides a rationale for a written form of ASL; the third advances opinions of the form such a system might take. The arguments in the first part are weak and reflect the same bias that historically has been shown against ASL itself. The third section advances some ideas that should provide the basis for interesting discussions. Among these are the relationship, if any, of a written sign language to English print, the extent to which it should be alphabetic and horizontal, and its role in the current American educational system.
  • Supalla SJ, Cripps JH, Byrne AP.
  • Why American Sign Language Gloss Must Matter.
    https://www.ncbi.nlm.nih.gov/pubmed/28238974 Am Ann Deaf. 2017;161(5):540-551. doi: 10.1353/aad.2017.0004.
    Abstract
    Responding to an article by Grushkin on how deaf children best learn to read, published, along with the present article, in an American Annals of the Deaf special issue, the authors review American Sign Language gloss. Topics include how ASL gloss enables deaf children to learn to read in their own language and simultaneously experience a transition to written English, and what gloss looks like and how it underlines deaf children's learning and mastery of English literacy through ASL. Rebuttal of Grushkin's argument includes data describing a deaf child's engagement in reading aloud (entirely in ASL) with a gloss text, which occurred without the breakdown implied by Grushkin. The authors characterize Grushkin's argument that deaf children need to learn to read through a conventional ASL writing system as limiting, asserting that ASL gloss contributes more by providing a path for learning and mastering English literacy.
  • Mayer C
  • Written Forms of Signed languages: A Route to Literacy for Deaf Learners?
    https://www.ncbi.nlm.nih.gov/pubmed/28238975 Am Ann Deaf. 2017;161(5):552-559. doi: 10.1353/aad.2017.0005.
    Abstract
    While there have been attempts to develop written systems for signed languages, none have been widely used or adopted. In his article in an American Annals of the Deaf special issue that also includes the present article, Grushkin makes a case not only for why, but how efforts should be renewed to develop a written signed language, suggesting that increased written-English competence will be a consequence of increased competence in written and signed American Sign Language, with literacy-related skills transferring across languages. The present author responds in terms of what is known about linguistic interdependence in spoken-language contexts and in light of the evidence base from hearing bilinguals. She argues that, given the field's current context, no compelling rationale exists for pursuing this route to literacy for deaf learners, and that other routes are more workable from pragmatic, theoretical, and evidence-based perspectives.

Glrx (talk) 04:35, 11 October 2017 (UTC)


It is easy to find opinion and controversy regarding deaf communities. Your sources above seem to fall into two categories: people with their own sign script and others who promote literacy through oralism. While their arguments may be interesting, I do not have access to the articles linked above so I can not comment further. From the abstracts though, it appears that the oralism proponents start with the premise that writing sign language is irrelevant without backing up their claim.
Connie Mayer's statement that for sign language scripts, "none have been widely used or adopted" is demonstrably false. Over 20 schools and Universities in Brazil teach and require SignWriting. Consider this recent 12 minute video that aired in Brazil about the history of SignWriting in Brazil or this recent 8 minute video about Marianne Stumpf, one of the leading experts in SignWriting.
Consider this report from a deaf school in Brazil, auto-translated from the Portuguese.

At the request of dear Valerie Sutton, I share with all our experience with SignWriting at the Special School for Deaf Friars Pacific. Located in the city of Porto Alegre, state of Rio Grande do Sul, Brazil, the school was founded in 1956 and continues its activities until today, attending all Elementary School (equivalent to the Junior School of the United States).

Since my entry as a Mathematics teacher in 2013, I can say that SignWriting began to be used more forcefully in almost every year (some teachers did not feel prepared to use or had no affinity with the system). In the last year we developed a kind of institutional policy of ours, so that SW was in all the materials produced and in the classes of all the disciplines, especially in the first years, when the children go through the process of literacy and literacy. This year we held a seminar focused on SW and the pedagogical possibilities it provides. We were surprised to see the great adherence of the school community and students, who also chose to attend the event. We are also happy to see that our students, especially the younger ones, have been able to express themselves and write, for example, letters to teachers - just as hearing students do. Since last year, we have offered SW courses to the outside community. This year I am as a teacher of this course, which was developed in partnership with the Federal University of Rio Grande do Sul. Many of our activities are registered in three virtual media, in which photos and texts of what is happening at school are published: - Website (http://www.freipacifico.org.br); - Facebook Page (https://www.facebook.com/frei.pacifico/); - Blog (http://freipacifico.blogspot.com.br/). I place myself at the disposal of conversations and debates, so that we carry on the beautiful project that we are all part of. Big hug, Fernando


Please research #DeafHaveVote. For the #deafhavevote campaign, you can find several websites with detailed reports and several playlists with over 50 videos. Hundreds of deaf representatives from four European countries took part in an event of discussion groups, workshops, presentations, and opinion research. SignWriting was used throughout the event. A strong majority voted to adopt SignWriting for use with the community.
Sutton SignWriting is unique because of Valerie's holistic vision. Formal SignWriting is unique because of simple, mathematical ideas. We did not design by consensus or the opinion of experts. What we have created has undeniably spread around the world. We have a visible presence online and an even bigger presence that is invisible online but exists in the real world. In dictionaries alone, we have 75 different sign languages available and continue to get new requests. The top 21 dictionaries have over 1,400 sign each, with the German Sign Language dictionary having over 24,000 sign entries. I'm hoping to import these dictionaries into Wiktionary and/or Wikidata some day.
I realize it is unusual for a writing script to request funding from Wikimedia for font development, but I was personally pointed to this funding source during a Wikimedia event. I acknowledge that this project falls outside the norm for funding, but this project is fundamental to help spread the Wikimedia movement within the deaf communities with SignWriting. I have been an advocate for this idea and working to make it happen. Other scripts aside, SignWriting is International (script code "Sgnw"), SignWriting is available on Wikimedia Incubator, and SignWriting has an open project on the Wikimedia Cloud services. This integration will continue and some day a sign language Wikipedia will go live on the production server. Funding this project will help make this happen faster and stronger. -Slevinski (talk) 16:07, 11 October 2017 (UTC)

Comments of Ruslik0[edit]

Just a follow up to the previous comment. Can you briefly summarize the changes that you made to the proposal as compared to that from the spring? Ruslik (talk) 19:40, 9 October 2017 (UTC)

There are several changes.
  • removed Brazilian training so the grant focuses on font development only
  • reduced grant amount as some preliminary font development has been completed
  • expanded the Unicode details
  • added attendance of another Unicode Technical Committee meeting
-Slevinski (talk) 13:34, 10 October 2017 (UTC)
The scope of the proposal changed considerably.
The original proposal was about adding 17 characters to the Unicode standard to handle 2D layout issues. The current proposal terms that as the "Compliant Solution" or "Option 2".[1] The original proposal wanted to make the 17 characters produce the 2D layout with VOLT.
The current proposal now also wants to pursue an "Optimal Solution" or "Option 1". ("We are targeting two Unicode designs: an optimal encoding and a compliant encoding.") The Optimal Solution "overwrites the Sutton SignWriting block in Unicode and uses plane 4 for the symbols of the International SignWriting alphabet 2010". The Optimal Solution apparently wants a lot of characters in en:Plane (Unicode) 4: "S10000 to S38b07" (this looks like it takes 3 planes unless it's a typo).
The two options suggest the problems are not settled. That Option 1 overwrites Option 2's existing Unicode assignments raises serious issues.
Glrx (talk) 15:32, 12 October 2017 (UTC)
It is good to embrace standards. It is best to produce something that works. I will produce a 2D font that works with a standardized character set. I will attempt to use the Unicode 8 specification with the 17 additional characters, but this set has several major unresolved issues. The optimal character set overwrites the Sutton SignWriting block and adds plane 4. Neither character set has the support of a UTC voting member yet, but we have yet to hear back officially from the UTC.
Unicode is a long term process that will not resolve itself without significant work. There are only three options: give up, start over, or move forward. We will not give up. We have no interest in starting over and there are no other workable encoding ideas. Our only choice is to move forward and continue to write sign languages and offer our users better tools. -Slevinski (talk) 17:06, 12 October 2017 (UTC)
Regarding timelines, SignWriting takes the long view. For characters and Unicode, here is the history and current plan. -Slevinski (talk) 16:07, 30 October 2017 (UTC)
2008
  • International SignWriting Alphabet 2008
  • SignWriting in Unicode (SWU) prototype
2010
  • October - International SignWriting Alphabet 2010
2011
  • April - Unicode Proposal version 1
  • August - Unicode Proposal version 2
2012
  • January - Formal SignWriting in ASCII (FSW) version 1. Stable but requires SVG for viewing.
  • October - Unicode Proposal version 3
2015
  • June - Unicode proposal version 3 added to Unicode version 8
2017
  • July - SignWriting in Unicode (SWU) version 1. An experimental design that is the same as FSW, but better.
2018
  • August - Completion of Two-D font for SignWriting using Unicode design
  • August - Unicode proposal version 4
2021
  • June - Unicode proposal version 4 added to Unicode

Another questions: what is duration of the project? Can you add a timeline? Ruslik (talk) 18:47, 29 October 2017 (UTC)

The project is scheduled to occur during the year 2018. Here is a general outline. I've added this timeline to the main grant page. -Slevinski (talk) 16:07, 30 October 2017 (UTC)
January - February
  • Font development and research
March - April
  • Font development continues
  • Preliminary Incubator integration
May - June
  • Font development continues
  • Improve Incubator experience
July - August
  • Completion of Two-D font for SignWriting using Unicode design
  • Unicode proposal and UTC meeting
September - October
  • Incorporate user feedback into Incubator
November - December
  • Import sign language dictionaries into Incubator

Deployment[edit]

What plans do you have for actually putting this into use, e.g. in some sign-language Wikimedia project (we have translatewiki:portal:asl for instance) and/or in mw:Extension:UniversalLanguageSelector)? --Nemo 11:43, 31 October 2017 (UTC)

The ASL language code is ase. The translatewiki link is translatewiki:portal:ase. We have 47% of the most important MediaWiki messages translated. We are currently using the ASCII version of FSW. Once the 2D font is ready, we will update these entries to the Unicode strings.
Regarding the Universal Language Selector, that is a great idea. I've looked at that project over the years, but never jumped into the code. I will definitely incorporate the ULS into this project if possible. When's the next hackathon? -Slevinski (talk) 15:47, 31 October 2017 (UTC)
Regarding the Universal Language Selector, American Sign Language written in Sutton SignWriting is coded as "ase-Sgnw". My name sign would be written as the following:
  • <span lang="ase-Sgnw" style="font-family:SuttonSignWritingTwoD;">𝠀񆄱񈠣񍉡𝠃𝤛𝤵񍉡𝣴𝣵񆄱𝤌𝤆񈠣𝤉𝤚</span>
Alternatively, if I wanted to see the One-D representation with the existing font, I could update the font-family as the following:
  • <span lang="ase-Sgnw" style="font-family:SuttonSignWritingOneD;">𝠀񆄱񈠣񍉡𝠃𝤛𝤵񍉡𝣴𝣵񆄱𝤌𝤆񈠣𝤉𝤚</span>
-Slevinski (talk) 20:27, 17 November 2017 (UTC)

Round 2 2017 Aggregated feedback from the committee for ASL Wikipedia 2-D Font Development for SignWriting[edit]

Scoring rubric Score
(A) Impact potential
  • Does it have the potential to increase gender diversity in Wikimedia projects, either in terms of content, contributors, or both?
  • Does it have the potential for online impact?
  • Can it be sustained, scaled, or adapted elsewhere after the grant ends?
7.4
(B) Community engagement
  • Does it have a specific target community and plan to engage it often?
  • Does it have community support?
7.0
(C) Ability to execute
  • Can the scope be accomplished in the proposed timeframe?
  • Is the budget realistic/efficient ?
  • Do the participants have the necessary skills/experience?
6.4
(D) Measures of success
  • Are there both quantitative and qualitative measures of success?
  • Are they realistic?
  • Can they be measured?
6.8
Additional comments from the Committee:
  • The project goals fit with Wikimedia's strategic priorities and the project has a potential for online impact. However the sustainability and scalability of its results are unclear - there are many unresolved problems with SignWriting: lack of necessary unicode characters and lack of wide adoption by deaf communities.
  • I think it’s a great idea and will make Wikipedia more inclusive and open.
  • Although this project poses many risks regarding acceptance of the proposed system, and the required technical changes. It is a important to generate ways to improve accessibility to users of all conditions, if this project has some improvement in this field it needs to be valued.
  • The project seems innovative. The potential is significant but so the risks. However the measures of success should be made more specific.
  • Measures of success are not specific enough, specially for the "secondary measurement" indicated, therefore project's outcomes may become impossible to determine if they are successful.
  • The project can be probably accomplished in the requested 12 months. The budget looks realistic and participants probably have necessary skills.
  • I don't know if the user has the capacity to execute
  • An approximate figure of how many hours dedicated to every task is not set neither cost/hour relation for the different jobs. Project plan may face delays related to the ability of WMF developers & sysadmins to review, test and apply suggested changes.
  • The community engagement appears to be limited but the project supports diversity.
  • Apparently some communities feel represented by this proposed solution but not all of them. Maybe this proposal should start from a bigger analysis receiving input from other communities.
  • Echo most of what Glrx says.
  • I am conflicted on this proposal. It has been significantly descoped from one that we considered in the spring. On the other hand many of the same problems remain. In the end I want to give it a chance. Though before the grant is proved it should be reviewed by WMF developers.
  • I need a breakdown of the budget components
  • Giving the total of available funds and the absence of a WMF tech staff full review I would incline not to fund this proposal.
IEG IdeaLab review.png

This proposal has been recommended for due diligence review.

The Project Grants Committee has conducted a preliminary assessment of your proposal and recommended it for due diligence review. This means that a majority of the committee reviewers favorably assessed this proposal and have requested further investigation by Wikimedia Foundation staff.


Next steps:

  1. Aggregated committee comments from the committee are posted above. Note that these comments may vary, or even contradict each other, since they reflect the conclusions of multiple individual committee members who independently reviewed this proposal. We recommend that you review all the feedback and post any responses, clarifications or questions on this talk page.
  2. Following due diligence review, a final funding decision will be announced on March 1st, 2019.

Questions? Contact us.



Round 2 2017 Aggregated feedback response[edit]

Thanks for reading and commenting on the grant. I appreciate the support and feedback.

Below I have organized the committee comments into groups and I have provide our responses indented. Committee comments are bold. The comments are sometimes contradictory and range from positive to negative. I have split some comments by sentence and topic, using ... when part of a sentence has been removed. -Slevinski (talk) 22:20, 22 November 2017 (UTC)

Support[edit]

  • The project goals fit with Wikimedia's strategic priorities and the project has a potential for online impact.
  • I think it’s a great idea and will make Wikipedia more inclusive and open.
  • It is a important to generate ways to improve accessibility to users of all conditions, if this project has some improvement in this field it needs to be valued.
  • The project can be probably accomplished in the requested 12 months. The budget looks realistic and participants probably have necessary skills.
  • The project seems innovative.
  • ... the project supports diversity.
  • In the end I want to give it a chance.
I really appreciate your efforts to read and think about the grant. Sutton SignWriting starts with Valerie Sutton's vision. Deaf writers were involved from the start. SignWriting has spread around the world because Valerie freely gave her script to an international audience. Through the Center for Sutton Movement Writing non-profit, she has provided books, videos, websites, and software. Sutton SignWriting is free to use and the Center for Sutton Movement Writing is committed to helping writers create standardized dictionaries, educational content, and various literature.
The current efforts to standardize Sutton SignWriting with fonts and Unicode is vital to bringing all of the various SignWriting communities together and encourage new communities to form. With a successful 2-D font, the name of a sign will be fully established. The way we write today will be comprehensible in 100 years. Binary data and other improvised solutions will loose their meaning over time, but plain text ensures standard and stability.

Doubt[edit]

  • The potential is significant but so the risks.
  • The community engagement appears to be limited but the project supports diversity.
  • Apparently some communities feel represented by this proposed solution but not all of them. Maybe this proposal should start from a bigger analysis receiving input from other communities.
  • I am conflicted on this proposal. It has been significantly descoped from one that we considered in the spring. On the other hand many of the same problems remain.
  • However the sustainability and scalability of its results are unclear - there are many unresolved problems with SignWriting: lack of necessary unicode characters and lack of wide adoption by deaf communities.
  • ... this project poses many risks regarding acceptance of the proposed system, and the required technical changes.
  • I don't know if the user has the capacity to execute
  • Echo most of what Glrx says.
Thank you for the time you spent considering this grant. I understand the doubt and accept that we face challenges. I hope one day we can earn your support or respect.

Measures of Success[edit]

  • However the measures of success should be made more specific.
  • Measures of success are not specific enough, specially for the "secondary measurement" indicated, therefore project's outcomes may become impossible to determine if they are successful.
The first measure of success is really based on the completeness of the 2-D font. If the font is complete, the grant was a success. Regarding secondary measures of success, I tried to categorize different areas of concern on the main grant page, but I should probably rewrite it. I am open to suggestions for this section.

Budget[edit]

  • I need a breakdown of the budget components
  • An approximate figure of how many hours dedicated to every task is not set neither cost/hour relation for the different jobs.
This project will account for 50% of my time next year. The other 50% of my time will include working with Valerie on her projects, server administration, website development, and other various needs of the Center for Sutton Movement Writing.
For this project, I have a general list of tasks that I will tackle, but I can not say with certainty which avenues will be productive and how quickly they will bear fruit.
For simplicity, I divided the budget into three line items.
1) Project Administration: 4,000 USD
This expense covers general administration along with other incidental expenses such as software or consultations or attending a hackathon.
2) Font Developer: 36,000 USD
This covers 50% of my time for the year 2018. I will be creating the font using a variety of techniques and tools. Some will be fruitful and others will not. My secondary focus will be on Incubator and the Universal Language Selector. The 2-D font is the best route for integration, and I will document the specifics and help develop the solution.
3) UTC Meeting: 3,000 USD
Document preparation, meeting materials, and travel expenses.

WMF Developers[edit]

  • Project plan may face delays related to the ability of WMF developers & sysadmins to review, test and apply suggested changes.
  • ... before the grant is proved it should be reviewed by WMF developers.
  • Giving the total of available funds and the absence of a WMF tech staff full review I would incline not to fund this proposal.
The creation of the 2-D font will allow the WMF developers to focus on supporting vertical writing mode and other general issues. With the 2-D font, they will not need to worry about the specifics of SignWriting, although their involvement is welcome.
The 2-D font can be developed and tested on Incubator. We already have a SignWriting gadget running on Incubator. Code reviews are quick and deployment is simple. The Universal Language Selector is ready for integration. American Sign Language (ase) written in the Sutton SignWriting script (Sgnw) is known as "ase-Sgnw".
If/when vertical writing mode is added to the WMF developer roadmap, it will be easier to include SignWriting with the availability of the 2-D font.

Round 2 2017 decision[edit]

IEG IdeaLab review.png

This project has not been selected for a Project Grant at this time.

We love that you took the chance to creatively improve the Wikimedia movement. The committee has reviewed this proposal and not recommended it for funding, but we hope you'll continue to engage in the program. Please drop by the IdeaLab to share and refine future ideas!


Next steps:

  1. Visit the IdeaLab to continue developing this idea and share any new ideas you may have.
  2. Applicants whose proposals are declined are welcome to consider resubmitting your application again in a future round. We ask that you first email projectgrants(_AT_)wikimedia · org to indicate your interest in resubmission so staff can review any concerns with your proposal that contributed to a decline decision, and help you determine whether resubmission makes sense for your proposal.
  3. Check back at the schedule for information about the next open call to submit proposals.

Questions? Contact us.


Thanks for the consideration[edit]

Thank you to everyone who read and considered the grant. Unfortunately, this grant will not be funded. We will not be submitting this grant to the Wikimedia foundation again. The lack of activity on the ASL Wikipedia was the major downfall of the grant. Ironically, the best way to increase the activity of the ASL Wikipedia was through this grant. Regardless, the two-dimensional font development will continue, just a bit slower. We are still committed to helping people write SignWriting content within a wiki environment and believe it is the best way for each community to create, share, and grow.

Another issue with the grant was community involvement, both inside and outside of the Wikimedia movement. Within the movement, there is general support and approval of SignWriting. Some groups (Cloud Services, Incubator, and TranslateWiki) have embraced SignWriting and the 2012 standard of Formal SignWriting in ASCII (FSW). We expect further engagement with the 2017 standard of SignWriting in Unicode (SWU).

In the future, it will be natural to write SignWriting within Mediawiki software. It is not a question of if, but when it will happen. Similarly, SignWriting will have full script support within the Unicode Standard some day. Our journey is not yet complete, but it is well on its way.

Outside of the Wikimedia movement, SignWriting has a wide and deep international community of supporters who want to write their sign languages now. There is some frustration with the tools currently available. We are working to simplify the task of writing while embracing international standards. With our 2017 standards, we have a stable platform for growth.

Each of the various sign language communities has the task of starting or continuing their literary tradition. Some groups are entering their third and fourth generations of writers. With the proper tools and support, there is no limit to their accomplishments.

When given the chance, humans will use language. Sign languages are real languages. Sign languages can be written. Of course deaf writers want to write. Of course deaf groups will join the Wikimedia Movement; it is a vibrant and diverse community with a universal ideal.

We appreciate that this is a push project and far outside of the Wikimedia grant process. We also appreciate the struggles of deaf to write their language. For a real and universal linguistic justice, sign languages need to be used in the home, in the schools, on paper, and on electronic media. This grant was to improve the electronic media representation of sign language.

Funding support services and technology developments for SignWriting is difficult to problematic. We would rather help our neighbor to write than tell him he needs to pay us first. SignWriting and our fonts and our tools are free to use. Our target audience is often under-served and marginalized.

We are writers, inventors, and software developers. The process of grant writing is outside our area of expertise and we have yet to find a reliable funding strategy for the non-profit. Thanks for your consideration and we're always open to new ideas and collaboration. Regards, -Slevinski (talk) 17:08, 19 December 2017 (UTC)

Reply by Glrx[edit]

When I came across the first proposal, I thought it was a great idea to fund. Yes, provide the deaf community with writing tools. The initial enthusiasm dulled when I started looking at the details. The Sign Puddle experience suggested that similar Wikimedia ventures would not turn out well. There is little publishing volume. Yes, there are some schools that have adopted Sgnw, but that evidence is anecdotal; for proposals, you need to get away from such evidence. The script usage has primarily been dictionaries and teaching -- something that is not the province of a general xxx.WP. There's also the nagging thought about why hasn't the script been embraced in the US: there are clear statements that there is no definitive ASL script.

The story of SignPuddle is long and twisted with multiple updates to the symbol sets and data formats. As the sole developer, I am both proud and embarrassed. The last major update to the code base was in 2012, with a minor update in 2017 for the new characters. The SignPuddle experience is a testament to the utility of SignWriting, because people have used it despite it flaws and awkwardness. So far in December of 2017, there has been new material written in 13 different sign languages that use SignPuddle.
I have put my major efforts towards font creation and character encoding. I am very proud of these developments that can be used outside of SignPuddle. In 2018, I will be releasing SignPuddle 3; a total rewrite using the new fonts and characters, but with all of the data from the current SignPuddle Online dictionaries and literature collections.

There should be some effort at understanding and publishing outside of the SignWriting community. You should have a much better idea of how your work fits in with the work of others. In the refs above, I see serious debates about the goals of deaf education. Those debates seem to want deaf literacy in the national language rather than literacy in just written sign language. Proficiency in the national language attenuates the need for a signed Wikipedia in that language. You should have measures of utility and acceptance for various systems. Those measures could be revealing to you. My take SignWriting is pitched as a done deal, and CSMW is trying to get the world to accept it as is.

Deaf education is a controversial subject without many hard numbers. We lack the resources to develop these hard numbers ourselves.
In the US, the deaf read English at about a 4th grade level. This number is controversial too, but everyone admits that the deaf reading level is well below their hearing counterparts. This deficit is understandable because the deaf can't hear the spoken language and the deaf lack literacy skills in their primary language. When deaf learn literacy in their primary sign language, they are able to apply those skills to learning literacy in their second language. We have tons of anecdotal evidence, but no large studies.
With our symbol set of 2010 and our stable character sets of 2012 and 1017, we are pitching SignWriting as a done deal. While the foundational work is completed, it is now up to each individual sign language to build their literary tradition.

Sgnw seems appropriate as an IPA for signed languages. Using just that should be enough for an NIH or NSF grant. Sgnw can express details, so it has the same value as IPA. Such a proposal does not require acceptance by a broad, nonprofessional, community such as WM.

SignWriting can be written detailed (phonetic) or simplistic (phonemic). Research and linguistic efforts tend to use detailed writing. Daily writing has tended toward simplification.

I also wonder about unnecessary technical challenges. If I look at scripts throughout the world, ltr-tb is dominant, but there are rtl-tb and tb-rtl. Why does Sgnw choose the tb-ltr outlier? It just seems to guarantee problems with integration. Today, we can write Chinese rtl-tb; Chinese took an expedient route. The SVG library libsrvg fails on tb scripts; Microsoft's Edge browser does not position tb scripts correctly.

SignWriting is written top to bottom because deaf writers in the US requested this change. When reading, the left and right sides of a sign are very important. When SignWriting is written horizontally, each sign causes a stop to determine the left and right side of the sign. When SignWriting is written vertically, the left and right sides of the sign become obvious. Vertically reading is quicker and more natural.

You need to give clearer explanations of what you intend to do. There are 17 magic characters, but what are they and how are they supposed to work? What does the 2D font do? Why is it essential to do something a particular way? Give me explanations that show a clear understanding of the issues and how they will be solved.

The 17 characters are explained in the presentation SignWriting in Unicode Next v4. The 2D font uses standard font technology to display a sign as a cluster of symbols, without the need for SVG. We need the 2D font developed this particular way so that we can write any sign without the need to add new characters or update the fonts.

I also look at strengths, and my impression is mixed. I laud the accepted Sgnw Unicode characters; that was good. I cringe at overwriting those definitions, and I cringe at asking for an entire Unicode plane. Maybe it is the right thing to do; Chinese characters have radicals, but the character set does not express those radicals but rather entire ideographs. Practically speaking, Chinese can have most of Unicode plane because billions of Chinese writers can use them; there isn't the same demand for Sgnw yet.

Unicode discussion started years before my involvement in 2004, without progress. The standardization efforts of Valerie Sutton and myself led to the creation of the International SignWriting Alphabet 2010 and the initial Unicode proposal. The subsequent Unicode proposals modified the character encoding in unfortunate ways.
We are currently waiting to hear back from the Unicode Technical Committee. Steven Loomis has been charged with responding to our issues, but he is not ready yet and has not given us a time frame for his response.

I think you need to communicate with researchers such as Grushkin and others who are doing written sign language. Grushkin has a font. Maybe there is some synergy there; you can help each other out. Other researchers can also suggest funding sources.

I dug up this summary:

I like the summary; it shows strengths and weaknesses and similarities of several writing systems. Both SignWriting and si5s are pictorial and therefore have similar problems. It offers characterizations such as an intent for dictionary use rather than general use. The summary also suggests what can be done quickly today with existing linear fonts. Grushkin can provide a linear font to Wikimedia today. He could have some of his students jump start en.Wiktionary with 2000 transcriptions (20 students with 100 transcriptions each). Maybe proponents of other scripts will offer their own font and transcriptions. The transcriptions might become Wikidata properties. Somebody may compile the transcriptions into jerk animations. Maybe the lack of detail provides a push for Sgnw because it will compile to better animations. Maybe somebody else tries translating between different scripts. Somebody else makes a gadget that takes a script and searches for the word. Those are projects with small scope and good chances.

For more practical applications of Sgnw, I would think there are Deaf-specific funding agencies. I'd expect CSMW would have approached those sources. If CSMW had received funds from them, then I'd expect to be told about that success. It would show that there was substantial Deaf interest. The absence of such reports makes me wonder about acceptance. Most of your examples have been about teaching children, so there may be some public and private funders who would support that arena. The David and Lucile Packard Foundation might be a target. There might be difficulty there because the focus would results about helping children rather than developing new systems.

Glrx (talk) 01:01, 20 December 2017 (UTC)

Thanks for you consideration and input. I fear that you are correct that our evidence is largely anecdotal and insufficient for grant writing. Regardless, we will keep promoting SignWriting and trying to improve the tools for our users. We believe we are quickly approaching critical mass with widespread acceptance that is obvious and undeniable. We understand people who doubt.
We appreciate all of the support and infrastructure that Wikimedia has provided to SignWriting. We look forward to an inclusive future where the deaf can share in the sum of human knowledge with written sign language. That's our vision. -Slevinski (talk) 15:49, 20 December 2017 (UTC)
You should not feel embarrassed about anything. Thanks for your efforts. Glrx (talk) 18:42, 20 December 2017 (UTC)

Eligibility comments for 2018 Round 2[edit]

Hi Slevinski-- I wanted to notify you that the deadline for Projects Grant proposals this round was on November 30th, so the proposal will not be included for review in the current round. If you wanted to submit this proposal for the next round, please note that we have not yet scheduled the next open call, and that there will not be another open call this fiscal year (i.e. through June 1st, 2019) due to some temporary changes in our grant programs. I JethroBT (WMF) (talk) 19:57, 9 December 2018 (UTC)

Thank you for the note. I realized that I missed this current round. I am getting ready for 2019 and decided to resubmit again. Next year is fine. It goes along with my project idea on IdeaLab. -Slevinski (talk) 01:46, 10 December 2018 (UTC)