http://pad.constantvzw.org/p/modifying.talk

MAKE CHANGES IN ODT FROM NOW ON!

=====  conclusion ==== 












What drew us into this

In 2015, The Unicode Consortium added five "Skin tone modifiers" to the ISO 10646 standard that encodes more than 800 emoji characters. It is this event that started a series of reflections and collective actions through which specific entanglements of technology, representation and normativity (re)-appeared.

While to most people consider emojis little more than a curiosity — a way to inject some humor, emotions or flirtation, into otherwise dry text messages – their popularity has coincided with a rising awareness of issues associated with identity politics. With the surge of instant messaging on both mobile and desktop-based applications, emojis have moved way beyond the smiley faces we type in e-mails by combining semicolons and brackets.

Technological systems increasingly represent, form and interact with diverse physical bodies and we need to radically rethink what it means to say 'everyone'. In an era that is apparently "post racial" and "post gender" we can see an actual backlash of racism and sexism, in terms of discrimination on one side, and strategies for affirmation on the other side. In post-Obama times, in times of Black Life Matters and with Gamergate still raging, we see how the actual companies that provide the infrastructures for on-line expression (Facebook, Twitter, Google ...) use an egalitarian discourse of diversity with increasing ease and success. 
The emoji modifier discussion turned out to be a fascinating case for analysing the way identity politics have been taken in account by global capitalism and how these issues have been turned into viable company assets that supplement and strenghten their commercial strategies.

The process of implementing emoji modifiers stages race, gender and technologies in a way that seems iconic of the commodification of identity politics into cultural and technical items. It shows how "identity washing" operates not only in city marketing or official international politics, but also through our inter-personal communication and devices. Throughout, we see anti-racist politics being emptied out of their deeper sense and meanings for the sake of a commodified version of equality.

The implementation of the Skin tone modifiers and following proposals for Emoji Mechanisms offered an opportunity to test out the possibility of intervening into the formation of technologised representation. By using the channels provided for feedback, and a close reading of the decision making process, we were able to experience the process of standard making as well as entering into a dialog with technologies.

===> Massification of communication tools in parallel (iphone, ipad, android, smartphones...)

Modifying the Universal

In 2015, The Unicode Consortium added five "skin tone modifiers" to the ISO 10646 standard that encodes more than 800 emoji characters. It is this event that started a series of reflections and collective actions through which we tried to address how specific entanglements of technology, representation and normativity (re)-appear.

While at the one hand you could consider emojis a pop curiosity — a way to inject some humor, emotions or flirtation, into otherwise dry text messages – their popularity has on the other hand coincided with a rising awareness of issues associated with identity politics. With the surge of instant messaging on both mobile and desktop-based applications, emojis have moved way beyond the smiley faces we type in e-mails by combining semicolons and brackets.

It is one of many examples where technological systems intensely interact with diverse physical bodies and we need to radically rethink what it means to say 'everyone'. In an era that is apparently "post racial" and "post gender" we can see an actual backlash of racism and sexism, in terms of discrimination on one side, and strategies for affirmation on the other side. In post-Obama times, in times of Black Life Matters and with Gamergate still raging, it is no surprise that the very companies that provide the infrastructures for on-line expression (Facebook, Twitter, Google ...) employ their  a-politizised and egalitarian discourse of diversity with increasing ease and success. The emoji modifier discussion turned out to be a fascinating case for analysing the way identity politics have been taken in account by global capitalism and how these issues have been turned into viable company assets that supplement and strenghten their commercial strategies.

The process of implementing emoji modifiers stages race, gender and technologies in a way that seems exemplary of how identity politics is transformed from a cultural, into a technical issue and eventually a commercial asset. It shows how "identity washing" operates not only in city marketing or official international politics, but also through inter-personal electronic communication. Throughout, we see how anti-racist politics is being emptied out of its deeper sense and meanings for the sake of a commodified version of equality.

The implementation of the skin tone modifiers and following proposals for “emoji mechanisms” was an opportunity to test out the possibility of intervening into the formation of technologised representation. By using the channels provided for public feedback, and a close reading of the decision making process, we were able to experience the process of standard making as it was happening before our eyes.

What is Unicode

[ROEL] + [FEMKE]

[image: wrong encoding]

Unicode is a non-profit organisation that concerns itself with universal character encoding standards, and thus represents a key underlying infrastructure that impacts all use of text on computers, mobile devices and the web.  These standards are designed to normalize the encoding of characters, the way they are stored, refered to and displayed in order with the goal of facilitating cross platform, multilingual and international text exchange. The Unicode Standard is mammoth in size, covering well over 110,000 characters.  Although only 2000 of these characters are considered emoji which, despite their relative marignality in comparison to the rest of the set,  they generate most of the public attention. The process of standardization is presented as open to discussion and proposals, the process for example relies on open public reviewing of issues and feedback and welcomes contributions for new additions. However, to become a voting member (i.e. have  power over accepting or rejecting proposals) one has to pay $18,000 per year (all prices are listed in USD only).  Most of the members work for one of the nine organizations that hold full membership in the Consortium. Seven of these nine are US-based technology companies: Adobe, Apple, Google, IBM, Microsoft, Oracle, and Yahoo. Obviously this has repercussions for the aim of universality of the project, as not all writing systems fit [footnote met congolese taal, conversationsboek interview met denis etc, bengali voorbeeld, femke kun jij dat oppikken? YES] the paradigm of assuming it necessary and possible to encode all languages of the world into a single template. More generally though, as we will see later in this text the problem of universality in computing begins by thinking that any problem can and should be encoded in symbolic logic [ Zach Blas, what source are we referencing? can someone add this? YES].


    Before Unicode

To understand the universal ambition of Unicode it is helpful to know about the context of its inception. Unicode was conceived in the late eighties as a response to the proliferation and incompatibility of many of the text encoding standards around at the time. During this period text exchange increasingly occured online and outside of a single language area, thus issues emerged as texts encoded in one language where shared and read on systems assuming an encoding in another language, resulting in illegible documents. This happened because different encodings assigned the same binary numbers to different characters. The response, partly made possible by increased computing capacity, was to strive to create a single universal encoding which encompasses all writing systems in the world. This encoding can be thought of as a single gigantic table which indexes all available characters to unique binary numbers, thus circumventing the issue of different encodings with overlapping character assignments. 

Maintaining this table and deciding what should be in it and where is the core activity of the Unicode Consortium, and only that. It is crucial to this story to understand that the consortium only deals with the assignment of numbers to characters. In other words, what Unicode maps is the 'idea' of for example latin capital (A) to a specific binary number. How that A itself is represented (italic, gothic, big, small etc) is the work of glyph [check: glyph = specific rendering] and font designers and not of the Unicode consortium. Furthermore, the consortium makes recommendations (the standard is non-binding), thus it depends on the manufacturers of soft- and hardware to implement these recommendations.  

The Unicode Standard

[image: wrong encoding]

Unicode is a non-profit organisation concerned with universal character encoding standards and  responsible for a key infrastructure that impacts all use of text on computers, mobile devices and the web. The Unicode standards are designed to normalize the encoding of characters, to efficiently manage the way they are stored, referred to and displayed in order to facilitate cross platform, multilingual and international text exchange. The Unicode Standard is mammoth in size and covers well over 110,000 characters of which only 2000 characters are technically considered emoji. Despite their relative marginal presence in the set, they currently generate most of the public attention for the Unicode standard and the activities of the Consortium.1

The process of standardization within Unicode is presented as open to discussion and proposals. The procedure for adding new characters for example, relies on open public reviewing of issues and feedback and the consortium welcomes proposals for new additions. However, voting members that have the power to decide if a proposal is accepted or rejected pay $18,000 per year2. Most of the current individual members work for one of the nine organizations that hold full membership in the Consortium, seven of these are US-based technology companies: Adobe, Apple, Google, IBM, Microsoft, Oracle, and Yahoo. Obviously it produces a bias in the universal project, if only because not all writing systems can be translated into a character based paradigm as well as the English language3. More generally though, the problem of universality begins with the assumption within computing that any problem can and should be encoded in symbolic logic4.

The universal ambition of Unicode can be traced back to it's inception in the late eighties. Unicode was conceived as a response to the proliferation of incompatible text encoding standards that were sinulaneously used at the time. Because electronic text was increasingly being exchanged online and between language areas, issues emerged when text encoded in one language was shared and read on a system assuming an encoding in another language. When different encodings are each assigned the same binary numbers to different characters, this results in illegible documents. The response, partly made possible by increased computing capacity, was to strive for a single universal encoding which would encompass all writing systems in the world. This encoding can be thought of as a single gigantic table which indexes all available characters to unique binary numbers, thus circumventing the issue of different encodings with overlapping character assignments.

Maintaining this table and deciding what should be in it and where, is still the core activity of the Unicode Consortium. It is crucial to understand that the consortium only deals with the assignment of numbers to characters, and not with the way they are rendered. In other words, what Unicode maps is the 'idea' of for example latin capital (A) to a specific binary number. How that A itself is represented (italic, gothic, big, small etc.) is the responsibility of glyph- and font designers, and not of the Unicode consortium5. Furthermore, the standard is non-binding and the actualisation of its universality depends on the willingness of soft- and hardware manufacturers to implement the recommendations of the consortium.
        
1 Mike Davis
2 All prices are listed in USD only: LINK
3 footnote met congolese taal, conversationsboek interview met denis
4 Zach Blas, xxxxx
5 http://www.unicode.org/reports/tr17/#CharactersVsGlyphs

First: All about rendering; room for interpretation. Specific implementation (Apple, pink faces) made the issue appear

[ROEL] + [PEGGY]

[image: person with ball]
[image: scattered emoji]

If one looks at the palette of emojis available on one's devices one can't help but wonder where they come from. Some symbols are relatively easy to understand [smiley exampe], yet others are to most people very obscure [japanse script symbol]. At first glance there also seems to be little logic to the all the emojis as a set. Additionally if one looks how emoji are distributed accross the unicode table, one will notice there is no 'Emoji' block, like there are 'Cyrillic', 'Syriac' or ' Mathematical Operators' blocks. Emojis are spread out across the unicode table.

http://stuff2233.club/~arra/unicode_poster/emoji_not_emoji.png [scattered emoji]
The above image shows all characters from the Miscellaneous Technical and Miscellaneous Symbols  blocks. The ones in black are considered emoji 

Emoji themselves date back to late 1990s Japan. When Japanese mobile carriers, introduced sets of pictographic images to use alongside text. The characters became an increasingly popular way for Japanese users to text and communicate with one another. Like the 90's browser wars, carriers would implement sets of images in order to seduce users to choose and stay with them over a competitor. The problem of course being that due to using proprietary encodings, emoji sent from one carrier wouldn’t display on phones from the competition, let alone to anyone outside of Japan.

In 2006 an issue was raised by engineers working on Google's Gmail about the need to be able to represent these kinds of characters in their Unicode systems. That was because ath the time Gmail wanted to take in all the information that was going on the email systems of japanese carriers and be able to handle it properly [https://www.youtube.com/watch?v=-n9ONNeACyw  talk by Mark Davis]. This eventually lead to the inclusion of the first emoji in Unicode 6.0 in 2010. However, rather than making new characters for all emoji it was decided that some (144) characters which had been in Unicode before and bore a likness to the new emoji, would now be redesignated as such. This move is interesting for two reasons, first this redesignation of certain characters as emoji started a process of reinterpretation and of sliding towards representation. Secondly it shows how standards bodies such as Unicode are thoroughly normalizing agents, focused on reinforcing existing standards. 

Both these movements can be seen at work if one attempts a genealogy of a single emoji. An interesting example is U+26F9 PERSON WITH BALL. This character was included in Unicode in 2009 in version 5.2. The character was included as part of a larger set of broadcast and map icons designed and standardized by the Association of Radio Industries and Businesses (ARIB), the japanese telecommunications standards body. The inclusion of ARIB symbols was first proposed in 2007, the reason being that ARIB was already a standard and it was expected to see more use in the future [http://www.unicode.org/L2/L2007/07391-n3341.pdf ]. The symbols proposed for inclusion are based on ARIB standard B24  first enacted in 1999 [http://www.arib.or.jp/english/html/overview/doc/6-STD-B24v5_1-1p3-E1.pdf  ]. In both the ARIB standard and the proposal U+26F9 PERSON WITH BALL is already present, yet with a different meaning. In the ARIB set the symbol, which shows a highly abstracted figure with a ball, it is described as 'track and field / gymnasium' [arib document p.52] and is part of a larger set of icons used for map legends.

how it changes from map symbol to person with ball, that eventually leads to inclusion in emoji where the description becomes guiding element for the design.

from pictogram to representation

Sliding towards representation

Since the early bitmapped images, or even the crude smileys in early e-mail and sms conversations, emojis have been sliding towards representation slowly but surely; their history within Unicode is interesting in that respect.



from broadcast (NTT), pictograms -> to personal communication (history of emojis), represenation

Importance of document exchange. 

Unicode developed from the perspective of character encoding of even dead languages; a linguist not visual tradition.
But coinciding with explosion of mobile use, popularity of emoji, user (vendors?) demand. Pressure on Unicode: public, vendors. After some hesitation, Unicode decides to implement emojis. The inclusion of emoji modifier became the xxx through which the unicode consortium got attention and worldwide press coverage, getting from an overspecialised technical group to a subject discussed in both techies spaces, newspapers and blogs as well as on The New York Time Op Ed or other mainstream medias. [ref: There’s been considerable media attention to emoji since they appeared in the Unicode Standard, with increased attention starting in late 2013. For example, there were  some 6,000 articles on the emoji appearing in Unicode 7.0, according to Google News. Extract from unicode emoji report 51]

But in the first versions of Unicode there were already dingbats. Images vs. characters; “realism”; ability (and tendency) of phones to display high-res images.
There’s actually even no Unicode block called “emoji”. The set of smiley faces is in a block called Emoticons, and most of the rest are in Miscellaneous Symbols and Pictographs and Transport and Map Symbols. Added for different reasons and usage.

Different visual traditions (maps vs. emotions: institutional vs individual communication). Emojis introduce color. Dingbat style.
Traditionally, many emoji fonts are cartoon-like and far from realistic, and have used a bright yellow for skin-like elements.  
Over time, humanoid emojis added and high-res representations.

When Apple decided to design new renderings, high res, pink and “realistic”, the trouble started: the abstract yellow started to look white.

Because one face does not fit all

[image: pink apple emoji with turban]
[image: emoji + modifier = ethnified emoji]

The massification of smartphones and fierce competition between vendors excelarated the investment in emoji's. They became a surprisingly important argument for buying the latest iphone, ipad or android. In this context, Apple launched their Iphone 6 with a completely redesigned emoji set, featuring gay and lesbian couples for the first time. While emoji had already started to slip from expression of emotions to expression of the self, the Apple designs where moving away from flat, cartoonlike images to an expression of volume and realism. It not only cemented the suggestion that emoji could stand in as usable digitalisations of the body, these “humans” all looked very pale.

Once Apple had launched its high-resolution, pink-hued emoji set, discussions flared up all over the web. The supposed realism of these renderings made people feel 'not represented' and subsequently users started to question the yellow base-color of emoji as well. Several petitions that asked Apple to increase the diversity in its emoji set, attracted thousands of signatures.1
The protest happened at a moment when companies such as Dropbox, Pinterest, Airbnb, Twitter publishing statistics on the lack of women and people of color in their workforces, publicly acknowledging their issues with diversity2. Each of the companies had hired specific employees, so-called “diversity managers” that were tasked with correcting these problems. The demands to technology giants to fix emoji diversity were therefore not without impact, nor urgency. The Unicode consortium that is made up of many of these companies, was now made responsible to respond to the pressure3. What was in essence caused by an awkward design-decision by Apple, conveniently became a problem to be solved on the level of the Unicode standard. In this meta-context it was clear that the issue could only be responded to with technological means.4

The solution the Unicode consortium decided to implement was to add six new characters that could modify only a designated set of emoji that were considered representing humans. Using essentially the same mechanism that is used to create ligatures, these skin tone modifiers allow users to specify any of six different shades of brown for emoji faces. If the device has the right icon is available, the emoji appears with this modification; otherwise the “default” face will be shown, next to the selected color swatch.
The consortium based the shades on the Fitzpatrick scale, an existing standard developed for measuring the sensitivity of skin to sun-exposure. From the little documentation of this surprising choice, we understand that it was believed that this standard could pass without triggering a complicated debate on the representation of ethnicity.5 In doing so, the Consortium conflated and misunderstood a medical standard for the way human skin responds to UV exposure, with a scale that represents skin color.6 The Fitzpatrick scale has not only a lineage to colonialism via the Von Luschan's chromatic scale, but by conflating the two lightest skin tones into one modifier, the Consortium carelessly suggests that light skin is less important to differentiate.

Apart from the colonial gaze already implied by introducing a skin color scale as such, it meant that the yellow emoji started to function as a white base with darker skin colors as an add-on. After Apple had started this confusion between yellow and white, it hardly comes as a surprise that people understand the modifiers a “blackface” move and a bastardized version of white superiority.7
Unlike the rigor that the Consortium usually applies to changes in the standard, the skin tone mechanism was implemented in a relatively short time and the argumentation for it comes across as hastily put together. We find little traces of a thinking through the possible consequences and in lieu of a discussion with the user-community, the sub-committee brings in enterpreneur Katrina Parott as an expert. She developed the succesful iDiversicon project in response to the on-line protests8. Parrott recently added a section “people with disabilities” to her commercially available emoji set.

While on-line representations are a way to make tangible what can not be changed in the physical world, should we see the addition of the modifiers as an example of successful user-agency, of powerful citizen action?9 Do the modifiers actually bring diversity to the emoji project?

1 Two petitions ran simaltaneously, http://web.archive.org/web/20140730201055/https://www.dosomething.org/petition/emojis         and https://www.change.org/p/groupme-and-emoji-developers-add-more-diversity-to-the-emojis?recruiter=7740596&utm_campaign=twitter_link&utm_medium=twitter&utm_source=share_petitionb The campaign was championed by pop singer Miley Ray Cyrus on Twitter https://twitter.com/hashtag/emojiethnicityupdate
2 Twitter: “We’re committing to a more diverse Twitter”  https://blog.twitter.com/2015/we-re-committing-to-a-more-diverse-twitter         Apple: “Inclusion inspires innovation” https://www.apple.com/diversity/  Google: “A diverse mix of voices leads to better discussions,  decisions, and outcomes for everyone.”https://www.google.com/diversity/
3 “When we originally designed emoji, the goal was to be as neutral as possible. The emoji charts that Unicode supports are black and white and other people will interpret them in color for realism ... we  struggled with how to deal with [diversity] for a bit because what  we didn’t want to do is multiply the emoji tremendously.” Meet the 63-Year-Old in Charge of Approving New Emojis. Victor Luckerson  interviews Mike Daves, Times, March 2016 http://time.com/4244795/emoji-consortium-mark-davis
4 “Tim forwarded your email to me. We agree with you. Our emoji characters are based on the Unicode standard, which is necessary for  them to be displayed properly across many platforms. There needs to be more diversity in the emoji character set, and we have been working closely with the Unicode Consortium in an effort to update the standard.” Katie Cotton, vice president of worldwide corporate communications for Apple, March 2014 https://www.yahoo.com/news/the-emoji-diversity-lobby-emoji-design-kevin-119455434306.html
5 “The Fitzpatrick scale was developed for use in dermatology, it is also used in cosmetology and fashion design (and) it has the advantage of  being recognized as an external standard without negative associations” xxxx        
6 http://www.beauty-review.nl/wp-content/uploads/2014/04/The-validity-and-practicality-of-sun-reactive-skin-types-I-through-VI.pdf
7 “These  new figures aren’t emoji of color; they’re just white emoji wearing masks” https://www.washingtonpost.com/posteverything/wp/2015/04/10/how-apples-new-multicultural-emojis-are-more-racist-than-before/                 
8 “Because One Face Does Not Fit All” http://www.idiversicons.com/   
9“UPDATE:  WE WON! You signed the petition. Now Apple is diversifying its Emojis! http://web.archive.org/web/20140730201055/https://www.dosomething.org/petition/emojis

Second: Criticism: Unicode assumes responsibility (Apple offloads) = modifiers

[PEGGY] + [FEMKE]

[image: pink apple emoji with turban]
[image: emoji + modifier = ethnified emoji]

The possibility for Universal representation that is encoded in the Unicode emoji implementation can only work if we first of all assume that emoji's are representing humans to begin with. As long as the action or emotion expressed by the cute yellow, a-sexual characters is emphasized, thoughts about gender, race and ability might go away. But who or what is the template for this 'universal' character?
 ? date of apple lauching and explanation => not sure about  wich step you are talking about => we should add images 

Once Apple had launched its high-resolution, pinkish emoji set, discussions flared up all over the web. Their supposed realism made people feel 'not represented' and subsequently they started to question the yellow base-color as well. A petition that asked Apple to increase the diversity in its emoji set, attracted thousands of signatures [ref: https://twitter.com/hashtag/emojiethnicityupdate xxxx].

The protest happened simultaneously to companies such as Dropbox, Pinterest, Airbnb, Twitter publishing statistics on the lack of women and people of color in their workforces, publicly acknowledging their 'issues with diversity'. Each of the companies hired specific employees, so-called 'diversity managers', tasked with correcting these problems. The demands to technology giants were therefore not without impact, nor urgency. The Unicode consortium that is made up of many of these companies, felt the need to respond to the pressure. What was in essence caused by a bad design-decision by Apple, became a problem to be solved on the level of the Unicode standard, an issue that in this context could only be responded to with technological means [ref: ““Tim forwarded your email to me. We agree with you. Our emoji  characters are based on the Unicode standard, which is necessary for  them to be displayed properly across many platforms.  There needs to be  more diversity in the emoji character set, and we have been working  closely with the Unicode Consortium in an effort to update the  standard.” Katie Cotton, vice president of worldwide corporate communications for Apple, March 2014 https://www.yahoo.com/news/the-emoji-diversity-lobby-emoji-design-kevin-119455434306.html.

When we originally designed emoji, the  goal was to be as neutral as possible. The emoji charts that Unicode  supports are black and white and other people will interpret them in  color for realism . . . we struggled with how to deal with [diversity]  for a bit because what we didn’t want to do is multiply the emoji  tremendously. [ref: Meet the 63-Year-Old in Charge of Approving New Emojis. Victor Luckerson interviews Mike Daves, Times, March 2016 http://time.com/4244795/emoji-consortium-mark-davis]

The solution the Unicode consortium chose to implement was to add six new characters that could modify a designated set of emoji. Using essentially the same mechanism that is used to create ligatures, these  “skin tone modifiers” allow users to specify any of six different shades of brown for emoji faces. If the device has the right icon is available, the emoji appears with this modification; otherwise the “default” face will be shown, next to the selected color swatch.

The mechanism was implemented in a relatively short time and the argumentation for it comes across as rather hastily put together. We find hardly any trace of a thinking through the possible consequences that seems in line with the rigor that the Consortium usually applies to changes in the standard. In lieu of a discussion with the user-community, enterpreneur Katrina Parott is brought in as an expert. She developed the succesful iDiversicon project (“Because One Face Does Not Fit All” http://www.idiversicons.com/) in response to the on-line protests. Parrott recently added a section “people with disabilities” to her commercially available emoji set.

Modifiers based on the Fitzpatrick scale, because:   

The Fitzpatrick scale was developed for use in dermatology, it is also used in cosmetology and fashion design (and) it has the advantage of being recognized as an external standard without negative associations [ref: Why Fitzpatrick]

What is The Fitzpatrick scale.

The problem of merging scale 1 and 2.

While on-line representations are a way to make tangible what can not be changed in the physical world, should we see the addition of the modifiers as an example of successful user-agency, of powerful citizen action? Do the modifiers actually bring diversity to the emoji project?

Third: Modifiers normalize rendering (exit Google blob; normalized humanoids)

[Roel] + [FEMKE]

[image: genealogy of google from blob to humanoids]

Apple immediately integrated skin tone options in April 2015

Interestingly, until the next Android update, Google/Android did not implement the modifiers. A Google spokesperson indicated that this was a deliberate choice, with all "human" emojis an unrealistic yellow:

"[Google's] emoji faces are playful and are all about conveying the emotion you're feeling. They aren't designed to look human or reflect human characteristics."
But recently:

"While cross-platform consistency was one reason for getting rid of the blob-people, another was to pave way for support of skin tone modifiers. It stands to reason that the blobs look great in yellow, but would look a bit weird if they had skin tones applied." [ref: Mike Davis]

Normalizing forces. Skin tone modifiers implemented for diversity start to call for reduction.

Fourth: our comment. How it was done and what happened

[FEMKE] + [PEGGY]

[sketches emojipedia ??]

In February 2016, following the perceived success of the modifier mechanism, the Unicode consortium made a proposal to allow further customizations of Unicode emoji characters. If accepted, it would ensure that gender variants (such as female runners or males raising a hand), hair color variants (a red-haired police woman), and directional variants (pointing a gun or a crocodile to the right, rather than only to the left) could be encoded.

The mechanism would use the same principle as the skin-tone modifiers, allowing only certain emojis to be modified by certain modifiers. But even if you could no type a female police officer or construction worker, why does no female dancer wear a sari, or is U+1F473, MAN WITH TURBAN the only man able to wear a turban? What about hairstyles, and different traditions of gesturing, let alone representation [ref: "In addition to gender bias, the clothing emoji are biased towards western and Japanese culture, so clothing items from other cultures might also need to be considered for inclusion. I think this is only the beginning of a discussion to make clothing items more gender & culturally inclusive, or to decide to what extent that is a goal." www.unicode.org/review/pri321/ ]?

By naively treating these images 'just like any other character', the Unicode consortium had opened a box of Pandora that would prove hard to close again [ref: When discussing the issue with Hin-Tak Lueng, developer of a font-validator aiming for full Unicode coverage, responded: “It was like they scratched an itch and then their whole skin fell off” (LGM 2016)]. 

It was with this observation that we arrived at the Execution event in Malmoe. At the workshop, a mix of students and academic researchers has gathered. Each from their own expertise is finding a perspective on the emoji project within Unicode. We propose to use the space of a workshop to write a collective response to the proposal, using the channel for public feedback provided by the Unicode consortium. After some initial reservations about the way critique would be possible or impossible within the confined space proposed by the consortium, we start to write. 

We finally agree on arguing against implementing the proposal based on four points, leaving out a fifth comment on the commercial drive of the Unicode consortium that we feel is at the root of the problem. We decided that we cannot reproach them for being what they are [ref: http://possiblebodies.constantvzw.org/feedback.html ].

1. By positing a "normal" baseline against which difference is to be  measured, the mechanism sets up problematic relations between the  categories that act as modifiers and the pictographs that they modify. 

*If we, for example, imagine what the consequences would be of adding  "disability" as a modifier to future Unicode specifications, it is easy  to understand this tension. Disability should never be conceived of as a  condition of modification to a base-line standard. In practice however,  it would have to be implemented exactly in this way, not unlike the way  the Skin Tone Modifiers are now implemented and more importantly  perceived as a "blackface" modifier to a "white" base. [ref: “These new figures aren’t emoji of color; they’re just white emoji wearing masks”  https://www.washingtonpost.com/posteverything/wp/2015/04/10/how-apples-new-multicultural-emojis-are-more-racist-than-before/ ]
*
*Apart from the colonial gaze implied of introducing a skin color scale, it means that the yellow emoji functions as a base and diversity are variants of it. After Apple had started to conflate yellow with white, it hardly comes as a surprise that people understandthe modifiers a 'blackface' move and a 'bastardized' version of white superiority. Intensifying the skin tone to make  darker, more 'coloured' emoji is apperantly their interpretation of “diversity”

2. To express diversity as a "variant" is a reductive response to the  complexity of identities and their representational needs. 

*If we  consider the implementation of gender variants (male, female, neutral)  for example, we can foresee issues with expressing more complex gendered  formations such as transgender or transsexuality. This issue would not  be solved by augmenting the resolution of the variants, as the mechanism  of varying between binary opposites itself is fundamentally flawed. [ref xxxxx]
   
3. The consortium should take into account how, once implemented, the modifiers will function in todays media environment. Should  Unicode-compliant search engines differentiate results according to modifier categories? 
*
*There is a documented case of Instagram searches  that return different results depending on emoji with the Skin Tone  Modifier applied. [3] We think that the responsibility for instituting  such potential for segregation lies not (only) with the one who  implements, but rather with the one who proposes and defines a standard.  Unicode can not neglect to consider such consequences. Aside from  impacting the equal access to information, the mechanism can be expected  to be used in reverse, as a method to identify authors of content on  the basis of their supposed race, gender etc.
*
*Russia is investigating if it can sue Apple for their representation of sexual diversity, since homoexuality is against the law. In the mean time, sex-positive emojis are being refused by app-stores because they do not permit 'sexual content'. Showing the lack of commitment to actual, complex needs for human communication.
*
*Segregation
*Racial marketing, racial filtering is what is at stake behind those supposedly funny way of rendering your emotions through textual communications. 
*In March 2016 was revealed that for the promotion of Universal's motion picture Straight outta Comptom, Facebook offered an "ethnic affinities" profiling that ressembled racial marketing segregation. Different ads were targetting different portion of it's users based on "their cultural taste" profiled by big data search on their posts. This work, providing a view on the movie completely different for "general population (non-African American, non-Hispanic)" and "African-american" audiences was the result of the work of jointed diversity teams in both companies. The argument behind those separated advertising was to offer a view on the movie to audience that were supposed to be attracted to the hiphop stars biopic for completely different reasons. That means that despite users could have refused to provide their ethnic background, Facebook used everything else available to guess it. Some could find this attention from this companies genuinely "marketing as usual". But we have seen recently activists and youth from Turkey arrested because of their social network accounts. We saw  how Libya used big data to target it's opponent, emprison, torture and murder them (see  Grandes oreilles et bras cassés). We also have seen how easily those companies are opening their stored information to police or intelligence services in the context of the global war on terror.
*http://arstechnica.com/information-technology/2016/03/facebooks-ad-platform-now-guesses-at-your-race-based-on-your-behavior/
*
*And of course, adding to this, soon as they were published emojis have been used for racist comments.
*http://www.independent.co.uk/life-style/gadgets-and-tech/features/apples-ethnic-emojis-are-being-used-to-make-racist-comments-on-social-media-10182993.html
*https://www.washingtonpost.com/posteverything/wp/2015/04/10/how-apples-new-multicultural-emojis-are-more-racist-than-before/
*
*From Glyphs to people  (gender chapter ?????)
*
*Emojis when slipping from expression of emotions to expression of the self, create a usable digitalisation of the body. 

4. The proposed modifiers for skin tone and haircolor are both based  upon questionable external standards. In the case of the Skin Tone  Modifiers, the Consortium has chosen to use the Fitzpatrick scale in an  attempt to find a "neutral" gauge for skin tone.
*
*Languages exists, no invention.

*The argument was made  that it 'has the advantage of being recognized as an external standard  without negative associations'. [4] In doing so, the Consortium has conflated and misunderstood a medical standard for the way human skin  responds to UV exposure, with a scale that represents skin color.[5] Furthermore, the Fitzpatrick scale has a lineage to colonialism via the Von Luschan's chromatic scale. To ignore this lineage is emblematic of  implementing a standard without careful examination of its scientific,  political, cultural and social context of production. In TR52, when discussing the options for haircolor, the consortium insists on a  limited palette by referring to the "cartoon style" nature of emoji.[6]  At the same time the proposal refers to the US Online Passport  application form as the "standard" to follow when choosing this limited  palette. The way the U.S. State Department chooses to view and  categorize people is a particular expression of how the border control  agency sees a person, it should not have to make its way into daily  communications. Rather than suggesting a less "loaded" standard to  follow, we argue that this is yet another example of the unavoidable and  unsolvable problems that the Unicode consortium runs into with the  logic of the modifier mechanism.

We send the comment a day after the workshop has ended, only a few hours before the request for comments is closed. A few weeks later we find out that the work on the new mechanism has been suspended. "Work on UTS 52 will be suspended for now in favor of an alternative (ZWJ) approach, focusing on female emoji, that allows for shorter development time and better fallback behavior on older systems." [ref: http://www.unicode.org/review/pri321/ ]

Fifth: All about implementation: switching to ZWJ means vendors have more agency in mixing and matching imagery; Unicode offloads (Google takes back?) politics back to vendor. Do-acracy

[ROEL] + [PEGGY]

[image: gender google]

add: uniformity of genders as a footnote, google http://www.unicode.org/L2/L2016/16181-gender-zwj-sequences.pdf

Why was the work really suspendend? Two months later, we gather to write up the story and attempt to understand 

Between the 9th and 13th of May 2016, the Unicode Technical Committee met in San Jose, California in a meeting hosted by Adobe.  Amongst the things up for discussion was the Technical Report #52 on the Emoji Modifiers. In the weeks leading up to the meeting, the members of the Unicode consortium had asked for and received public input for TR#52 and the proposed meetings. On the 10th of May the Emoji Subcommittee and the voting members of the consortium went through the agenda, reviewing the proposals and comments. This happened during the lunch in a so called 'ad-hoc' session (does this mean it is off the record?). During this session Google presented a document titled "Expanding Emoji Professions: Reducing Gender Inequality" (http://www.unicode.org/L2/L2016/16160-emoji-professions.pdf ) and simultaneously released this same proposal (which reads as a press release rather than a technical document) to the public. The Guardian (https://www.theguardian.com/technology/2016/may/10/female-emojis-google-equality ) and several other major news outlets ran a story  on Googles proposal the same day. After a short break a consensus was reached to suspend any work on UTR#52 and to pursue "[..] an alternative approach using ZWJ for representing female emoji [..]"(http://www.unicode.org/reports/tr52/ ), referring to Google's proposal.

These events represent an interesting shift in where the responsibility and agency shift from the Unicode Consortium to the vendors themselves. The proposed change was in favor of using the ZWJ (Zero-Width-Joiner) mechanism rather than a 'modifier' /modifier tag mechanism. The ZWJ is an invisible character already used in unicode to denote the combination between two separate characters. This is being used for example in the family emoji, where the unicode characters for man woman and child are written in combination with ZWJ and it is then up to the vendor to implement this as a family emoji . The important shift here is that new emoji can thus be made making combinations of existing symbols, rather than having to propose new modifiable emojis. In effect this means that any new emoji can be invented (and implemented) by the vendors, without having to go through the Unicode Consortium (example, rainbow flag, microsoft ninja cats). In effect this is a de-politization of Unicode, since any move towards representing 'diversity' via emoji can now happen through the vendors themselves (google already claims gender with gendered professions and rainbow flag).

The event also represents a typical case of do-ocracy, in which a (nominally) open and discoursive process of negotiation is sidelined by presenting faits accomplis. Doocracy is a popular mode of decision making in technical circles for its speed, decisivness and 'meritocracy'. Having done the task then also becomes the justification and validation for it ('why do we use ZWJ emojis? because google just did it!'). If you want to oppose, you just 'do' something else. It assumes everyone is able to 'act'. Whereas unicode nominally left a space for individuals and small organisations to participate in the discourse and creation of standards, these individuals and small organisations can never compete with google do-ocracy.  It turns the Unicode Consortium into what so many open standards bodies have become, a rubberstamping entity to validate unilateral decision by large players.

+a discourse of open discussion when in fact it's companies brute force in every step of the process. 
sidelining of the open process, the comments etc
http://www.unicode.org/L2/L2016/16169-gendered-prof-cmt.pdf no additional emojis
http://www.unicode.org/L2/L2016/16193-gendered-emoji-2.pdf oh, google is going ahead, heres how to limit the damage

Next steps will be out of the hands of Unicode; now how to comment, where to respond? Guess. It is ongoing.  (what's in unicode stays in unicode)

[FEMKE] + [ROEL] + [PEGGY]

oscillation, mirror-game: emojis get into unicode, apple creates a problem, unicode offers a solution, google abstains, extrapolates the solution, google steals the show, Remember: Unicode = google + apple, apple != google
sidelining of the open process, the comments etc 

The proposals by the Unicode Consortium are merely techno-centric patches, engineering solutions in response to the increasing complexity of cross-device and cross-cultural computing that demands a rethinking of compatibility/translation in terms of difference.

This superficial colorblindness abounds while a wide wave of reactionary movements gets structured, from anti-gay mariage rallies to tea parties, and national fronts of all kind with a new polished face while, in reaction of radical islamic bombings all over the world we have to face civil liberties restrictions that include the body and especially the facial recognition as a mean to evaluate who to survey, control or arrest.

Rainbow flag example [ROEL CAN YOU ADD THIS/A REFERENCE]

In this context, it is difficult to think about the way emoji standards are being handled without a sense of paranoïa. When social networks can target ads, based on previous post and supposed tastes, on what they present as an ethnic profile, where will led us the use of modified emojis?

By tightly keeping the "modifiers" in line with the constraint universalist belief-system of Unicode, at best a limited set of variations on the same could be made available.
The possible is aligned with the available, and not at any moment are the colonial assumptions underlying the system of encoding being questioned: the assumption that everything can and should be encoded in same system [back to Zach Blas]

now that diversity gender and ethnics -- real things being addressed, momentum but falls flat, superficial
BLM etc provide an opportunity to think more profoundly about issues, wheras emoji+modifier turn this moment into modes of thinking more superficially

We should demand more from software/communication standards than appeasement (geruststellende oplossingen) variation. Develop possibillities for multiplicity rather than solutions for diversity. Non solution way
This is not easy, because it means involvement/engagement, not demanding solutions. Claiming a voice in the way technological systems are being drawn up.
stop being a client
rigorous discussions beyond hyped look what i've got 
potential in open process to find a way to understand process of technology making and standard forming and for a larger public to engage with it. ref: our decision to do the comment

P.S.:
After a few hours of browsing Unicode documents, Roel finds the preparations by the Emoji subcommittee for the meeting. Our comment is dismissed with a cryptic "Snelting et al: Too late for ESC response" [ref: http://www.unicode.org/L2/L2016/16130-emoji-subcom.pdf ]

//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////

[maybe] missing in action: 
gender:/
women with bunny ears
us corporate lange and concerns translated in emojis :> cf women jobs reflect google interedt in having women in their teams

an history of slipping and commercial disruptions
from broadcast (NTT) to personal communication (history of emojis)

cross devices / cross cultural spaces/ creating spaces

//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////

Dear Mark,

On the 2nd of May 2016 we've submitted feedback to PRI32(?) on the propsed UTR#52. After doing so we've promptly received confirmation that our submission arrived in good order. Since then however, we have not heard anything from the Unicode consortium. We've spent quite some searching the website to see if a reply was published somewhere on unicode.org. Yet much later and only by accident we found out about L2016-16130 ( http://www.unicode.org/L2/L2016/16130-emoji-subcom.pdf ) that the ESC decided not to respond to our comment. It is for that reason that we contact you, because we submitted our comment in time and according to Unicode there would be a response within three weeks of the UTC meeting. 

In the mean time things have moved quite quickly, UTR52 has been suspended in favor of considering this proposal: http://www.unicode.org/L2/L2016/16160-emoji-professions.pdf . Our reading of the events is that the Emoji Professions proposal was already on the table before UTC149, which is why it was 

trying to understand the decisionmaking process
how long ago was http://www.unicode.org/L2/L2016/16160-emoji-professions.pdf prepared?
did we not get a reply because we talk of the modifier mechanism and the move is now to use ZJW?
Can we expect a response after UTC149?


/////////////////////////////////////////////////////////////////////////////////

Cute studies is here! https://docs.google.com/document/d/1qO8qiY8FaFlsMGrBFvH1GRVQqPMFlxp7pEGkIdosW1Q/edit?pref=2&pli=1

Often people think of emoji as a universal language, and they’re really not. Many different cultures will give different meaning to  different emojis. The emoji will start to         be used in different ways.

I’m sure you’re aware of the use of the eggplant emoji, for example.  (Ed. note: Eggplants are sometimes used in a suggestive fashion) ...  I  don’t know if anyone has done a study on it, but I think it’s clear that a lot of people are picking different skin tones according to whatever they’re trying to say." [ref: http://time.com/4244795/emoji-consortium-mark-davis/ ]

-------- Forwarded Message -------- 
Subject: Re: Issues with modifier mechanism, UTS #52 
Date: Mon, 02 May 2016 16:07:13 -0700 
From: Rick McGowan <rick@unicode.org>
To: femke@constantvzw.org

Thank you for your input. It will be added to the public feedback for 
PRI #321. 
Regards, 


Differences in rendering, communication 'errors'. Reasonable trust in what gets communicated … which was exactly the point.
Skype has removed some of its emojis which have the "potential to offend", according to its community managers. Some of the icons include a smiley face showing its middle finger and a woman's legs wearing high heels.


Computing bias towards English.

Much of the architecture and design as well as the everyday work is about permitting software to translate smoothly between institutional, political, linguistic and economic contexts. [ref: Adrian Mackenzie]

a recap of events after the submission of the comment:
    
    there was a Unicode meeting during which UTR52 was discussed. 
    There it was decided to suspend work on UTR52. That can be read in the meeting minutes: http://www.unicode.org/L2/L2016/16121.htm#147-C10

On the page of UTR52 the reason for suspension is written: http://www.unicode.org/reports/tr52/
"Work on Proposed Draft UTS #52, Unicode Emoji Mechanisms,  has been temporarily suspended. An alternative approach using ZWJ for representing female emoji will pursued. Using a ZWJ mechanism would allow for shorter development time and better fallback behavior on older systems. The Unicode Technical Committee will also pursue alternative approaches for expanding the representation of flags and addressing hair color."

ZWJ: 
    http://emojipedia.org/zero-width-joiner/
Modifier:
    http://emojipedia.org/modifiers/
Full emoji list:
    http://unicode.org/emoji/charts/full-emoji-list.html
    
The rainbow flag do-ocracy: 
    http://blog.emojipedia.org/rainbow-flag-emoji-details-published/
    http://www.unicode.org/L2/L2016/16183-rainbow-flag.pdf <- the actual proposal
    
Blog post on ZWJ and their possibilities
    http://blog.emojipedia.org/emoji-zwj-sequences-three-letters-many-possibilities/
    
Technical implementation of new ZWJ genders:
    http://www.unicode.org/L2/L2016/16181-gender-zwj-sequences.pdf


https://www.theguardian.com/technology/2016/may/10/female-emojis-google-equality
https://i.guim.co.uk/img/media/a0a5d38581477f4b6e6dbbe42b421b2666a6eae2/3_47_1044_626/master/1044.jpg?w=620&q=55&auto=format&usm=12&fit=max&s=5395c6ab2b1caf43d1baa84aad409fef

New jobs and gender options won't have to wait for a new version of Unicode.
http://arstechnica.com/gadgets/2016/07/now-all-of-your-emoji-can-be-either-male-or-female/



Unicode is a set of standards, an underlying infrastructure that impacts all use of text on computers, mobile devices and the web.  Unicode is a standard designed to normalized the encoding of characters, the way they are shaped and scripted in order to able cross platform, multilingual and international text exchange. ILarge! The Unicode Standard is mammoth in size, covering over 110,000 characters. Only 1000 of these characters are considered emoji but their relatively marginality does not prevent them from getting a lot of internal and public attention.

Unicode was first conceived in the late 80s by Lee Collins, Joe Becker, and Mark Davis, engineers at Apple and Xerox, as an attempt to create a universal character set for all text, not just English text. Together, they founded the Unicode Consortium, which determines the list of Unicode characters. The standard simply provides a number for each character used in a text. Every character you type – whether letters like ‘a’ and ‘b’, punctuation marks like ‘?’ has been given an ID number, a 'character encoding'. 
Characters vs. glyphs: A recommendation, not an implementation (nor representation). Difference between codepoint and glyph, between glyph and font/implementation.
Obviously differences between languages, complicated politics. Example congolese – type many characters at the same time. The Unicode project aligns with that tradition of assuming it necessary and possible to encode all languages of the world into a single template [ref: The problem of Universality in computing begins by thinking that any problem can be encoded into symbolic logic and thus computed. Zach Blas]

The Unicode Consortium is a non-profit organisation, but membership of the Consortium is not cheap. Full membership and voting rights cost $18,000 (all prices are listed in USD only) per year. But the Unicode process is presented as opened to discussion and proposals (open public reviewing issues and feedback, volunteering and public conferences) . Adding new langages or new set of characters, as well as commenting actual official proposals are constraint by set of rules and protocol;  For example, you have to follow guideliines that proves that the character set and the langage you propose  is in use in an actual community, have a written litterature.  For example Basa Vah from Liberia or the nearly extincted Ahom langage. are not present in Unicode, despite having a reserved block and considered UND (Unknown langage or invalid). Some langages, mainly ancient are considered unsupported scripts while actual supported langages could miss glyphs preventing proper writing.  http://pad.constantvzw.org/p/modifying.writingn it (so Star Trek's klingon was rejected www.unicode.org/L2/L2001/01212-RejectKlingon.html)
Bias can be easily seen in how the Unicode Consortium that maintains the standard, is made up of public and private members and aims to include all scripts used in the world. Most of the members work for one of the nine organizations that hold full membership in the Consortium. Seven of these nine are US-based technology companies: Adobe, Apple, Google, IBM, Microsoft, Oracle, and Yahoo. 

The very first version of the Unicode standard did include Bengali. However, it left out a number of important characters. Until 2005, Unicode did not have one of the characters in the Bengali word for “suddenly”. Instead, people who wanted to write this everyday word had to combine three separate, unrelated characters. (…) Even today, I am forced to do this when writing my own name. My name is not only a common Indian name, but one of the top 1,000 names in the United States as well. But the final letter has still not been given its own Unicode character, so I have to use a substitute. [ref: https://modelviewculture.com/pieces/i-can-text-you-a-pile-of-poo-but-i-cant-write-my-name ]

(not sure we shoud talk about NSWSTA) ? => the inclusion of  more vendors in the interpretation of emojis in the 9.0 chart 

Internally, the Unicode consortium solved the problem by zoning off the emoji discussions to a special mailinglist, and ad-hoc meetings, so the disproportionate attention would not overflow the other interest of the members. Other proposals have been, to remove them from the standard or at least limit.

Multiplicity vs. variation