 |
| ABOUT | EVENTS | RESOURCES | PRESS |
BOTS, ROBOTS, CYBORGS
Just For The Record is happy to announce a new duo of events entitled Bots, robots, cyborgs, exploring the writing of non-human agents like the bots on Wikipedia.
If recent studies have shown that women represent a very small minority amongst the contributors of Wikipedia, other studies also show that a majority of its most active and founding contributors were and are non-human actors called bots, who seem to escape the gendered binary that we apply to human bodies.
Since the creation of the internet, some people have seen in the cyberspace a promise to go beyond gender, race and class. A space to construct an “empowering virtual reality”, as Essex Hemphill phrased it in his conference On the Shores of Cyberspace, Black Nations/Queer Nation? in 1995. In her Cyborg Manifesto, Donna Harraway presented the figure of the Cyborg as a post-gender, fragmented identity. She invited feminists to get involved with technology. Yet these same researchers noted already that off-line boundaries also exist in cyberspace. Since then researchers have noted the many ways in which on-line representations are structured around gender, race and class.
In our two upcoming events, we will take as object the Wikipedia bot: what is the gender of this denizen of cyberspace? Can the bot ever be, like Wikipedia wants itself to be, neutral?
- Venue: Muntpunt, Place de la Monnaie 6, 1000 Bruxelles, Belgium
- Date: Friday June 10 2016 and Saturday June 18 2016
- Time: Friday 13:00 – 20:00 and Saturday 13:00 – 18:00
- Language: Dutch, English, French
- Cost: Free
- Participants: Open to anyone interested in this experience: beginners welcome! débutant(e)s bienvenu(e)s! beginners welkom! Experienced Wikipedia editors will be present and will share their knowledge in editing Wikipedia.
- What to Bring: Attendees can bring their own laptops and power cords.
- A temporary library will be set up by the participants who are invited to bring their books and digital documents to share with the others!
During the event, we invite you to contribute to Wikipedia and to the discussions around the gender gap on Wikipedia. You can edit an existing article to improve it, create a new one about a subject that doesn’t exist, but we also highly value the sharing of editing experiences, and ideas about what could make Wikipedia a more welcoming and colorful place!
Here are interrogations we would like to share with you:
- if Wikipedia can be considered as a tool to re·write history, how would you like it to be written?
- what is the influence the way we write on the representation of history and its main figures?
- can writing ever be neutral?
- Add your name to the participant list of this event at the bottom of this page
- Write some informations on your own wiki page by clicking on your user name at the top of the page (so that your name doesn't appear in red when you start editing articles)
- Write some informations on your Sandbox, also at the top of the page
- Improve an existing article
- Create a new article
Here are lists identifying articles that could benefit from edits and expansion:
- The Art+Feminism list of article to be improved
- Women Artists, by the Women in Red
- A list by the project Women in Red
- Women Artists from all over the world, a list by the Project Women Artists
- Women scientists, a list by the Project Women Scientists
- Women writers, a list by the Project Women Writers
- These lists mostly link to English Wikipedia, don’t forget your own language’s Wikipedia!
- Man as false generic: Ban the use of the words man, men and mankind to refer to a person or persons of unspecified sex or to persons of both sexes.
The page Writing about women offers great insights. Look for the following problems in existing pages and try to fix them:
- Male is not the default: Avoid labelling a woman as a female (ex: author, politician etc.), unless her gender is explicitly relevant to the article. An opposite example is saying male nurse.
- Use surnames: Look for articles using surnames for men, while calling women by their first name. See example
- Infoboxes are an important source of metadata (see DBpedia) and a source of discrimination against women. For example, the word spouse is more likely to appear in a woman's infobox than in a man's.
- This is a good and terrifying exercise, try to find a page (that is not) defining women by their relationships in the first paragraph.
- One study found that women on Wikipedia are more linked to men than men are linked to women.
- Use gender-neutral nouns when describing professions and positions
- Try changing the order in which groups are introduced – man and woman, male and female, Mr. and Mrs., husband and wife, brother and sister
- The use of the generic he (masculine pronouns such as he, him, his) is increasingly avoided in sentences that might refer to men and women.
- Avoid problematic phrasing when talking about married people. See examples
- Do not refer to adult women as girls or ladies!
- Avoid images that objectify women!
- Art+Feminism resources
- Gender gap resources
- Writing about women
- Gender neutrality in English
- Avoiding Heterosexual Bias in Language
- the he-she package
- Every Wikipedia page has a Wikidata page attached (database of normalised information that can be read by humans and machines: datas): property P21 is “sex or gender”. Gender was one of the first properties (21st) defined! It has an item number. Sex or gender: link to a value, or many values (like woman, trans*, many other possibilities, …)
- Gender is a “Statement”, examples of others include "instance of" (“is a”), "date of birth", "first name", etc
- To get the Wikidata item, go on the left menu, under the “Tools” section.
- You can create a Wikidata page before the Wikipedia articles.
- Some categories implicitly identify gender: e.g., “women chemists”.
- “I don’t trust my bot” - it's Ok to make imperfect bots if you (a human) skim through the results and double check with your human eyes/understanding - still more efficient than do everything yourself/
- Which language to write: Python is the easiest programming language, but other languages can be used.
- Bots can be run on your computer, or uploaded in the tools server of Wikipedia
- So you could fake that you’re a bot! You could ask for being a bot, then being flagged as a bot, and then your edits do not show up in "recent changes"
- Probably this kind of behaviour is detected by "metabots"
.jpg)
- How can we imagine a bot that can identify gender-biased articles. In a “constructive” way (without focusing only on “vandalism”, without only deleting but proposing…)
- How would you find other biases?
- Will the bot be able to point at things that we cannot detect? (humans should be trained to detect)
- Guidelines for bots / for human?
- Can the code be able to work somewhere else, outside Wikipedia? (Yes)
- For instance: go and look on pages like “Writing about women”.
- Name? Gentlebot? Genderbot?
- Pattern: if an article is about a man, he’s addressed with the surname, and if it’s a female she’s addressed with her first name. This pattern can be expressed algorithmically:
- Limit the field: specifying the properties you want (ex: articles about women chemists)
- Identify the surname and the first name: occurrences of both in the article: if the first name is used more than *** times, then mark the article “possibly gender-biased”.
- Sometimes, there’s only a first name (Madonna) or Royal people are named in a specific ways.
- Pattern: articles that talk first about the brother/father/… in the first sentence, and even in the first paragraph.
- Family-related information in the first paragraph. (Basically: it is not a defining characteristics)
- Some very rare corner cases are counterexamples: "princess X is notable for being the only child of king Y"
- How to identify this?
- Find words like “mother”, “father”, …
- Links to an article with the same surname in the first paragraph.
- Collaboration human-bot…
- Can it take the sentence and put it under a family section? Or at least make a proposition that can be validated by a human editor. Sometimes it is already in a family section, but repeated on top. (so then it needs to be removed)
- Can it change the name? (add the surname when it misses)
- Is there space for proposition of edits?
- The bot could post it on its own page or on the talk page of the article: “Why don’t you move this sentence in a specific part?”
- The bot could add a template banner on a page.
- Link sex tape scandals with revenge porn. Sex tape scandals are glorifying but also destroying lives and careers. Things should stand out sometimes, but usually "hidden" and obscured, like being one item in a list of political achievements.
- Sexual harassment: considered as a morality issue, which is problematic…
- Search more about the concept of morality (in French)
- Further research is needed
- Article about refugees: quite gender-neutral, but needs a specific section related to women.
- Gender-specific and gender-neutral pronouns.
- Using standardised language is based on assumptions.
- Generic pronouns: he used as a generic use (“neutral”) and “grammatically correct”, generic she: “When most members of the group are usually thought of as women”.
- "Secretary and her temper", etc. as bad examples of legitimate language usage
- If it’s written like this, then it’s never going to change… Add more references that nuance the phrasing of this article.
- See Spivak: set of neutral pronouns that can be used.
- Saying on Wikipedia that “this is grammatically correct” is not Ok since it does not faithfully represent the reality.
- Saying on Wikipedia that “I do not think this should be correct” is not Ok since it does not faithfully represent the reality.
- If the issue is complicated and changing, Wikipedia should embrace it and describe it as such: "some older books see it as grammatically acceptable (source), some modern scientists see it as a problem (source), there are possible solutions discussed: (1), (2), (3)", etc
- Move female explorers on the List of female explorers and travelers to the general List of explorers
- Revenge porn research to improve the page
- /info/en/?search=Hero#Heroine_or_hero added a new quote to the section
- Craig Owens, american post-modernist art critic, gay activist and feminist french translation in progress
- Margaret M. Adams on it's way
Sign here to help us learn more about how users interact with Wikipedia! If you sign up, we can observe how your username uses the Wikimedia projects during and after this event. This will help us to better measure the effectiveness of this event and similar programs. This means that your publicly available activity and the information you share with us during this event may be processed by the Wikimedia Foundation, Art+Feminism, and the organizers of the local event, and may be transferred to or from the US and other countries that may not have the same level of privacy regulation that your country does. However, we will not share your information with third parties or publicly unless it's in aggregated or anonymized form.
Prior to the event:
- Do you have a Wikipedia User Name?
- No? Create a Wikipedia account
- Yes? Go to Step #2
- Sign up! Add your Wikipedia User Name to this section by clicking the blue button below (follow instructions). Your name will be added to the bottom of this page
Bots, Robots and Cyborgs - Part 1 10.06.2016: 17 people present
Bots, Robots and Cyborgs - Part 2 18.06.2016: 26 people present
|
| ABOUT | EVENTS | RESOURCES | PRESS |
BOTS, ROBOTS, CYBORGS
Just For The Record is happy to announce a new duo of events entitled Bots, robots, cyborgs, exploring the writing of non-human agents like the bots on Wikipedia.
If recent studies have shown that women represent a very small minority amongst the contributors of Wikipedia, other studies also show that a majority of its most active and founding contributors were and are non-human actors called bots, who seem to escape the gendered binary that we apply to human bodies.
Since the creation of the internet, some people have seen in the cyberspace a promise to go beyond gender, race and class. A space to construct an “empowering virtual reality”, as Essex Hemphill phrased it in his conference On the Shores of Cyberspace, Black Nations/Queer Nation? in 1995. In her Cyborg Manifesto, Donna Harraway presented the figure of the Cyborg as a post-gender, fragmented identity. She invited feminists to get involved with technology. Yet these same researchers noted already that off-line boundaries also exist in cyberspace. Since then researchers have noted the many ways in which on-line representations are structured around gender, race and class.
In our two upcoming events, we will take as object the Wikipedia bot: what is the gender of this denizen of cyberspace? Can the bot ever be, like Wikipedia wants itself to be, neutral?
- Venue: Muntpunt, Place de la Monnaie 6, 1000 Bruxelles, Belgium
- Date: Friday June 10 2016 and Saturday June 18 2016
- Time: Friday 13:00 – 20:00 and Saturday 13:00 – 18:00
- Language: Dutch, English, French
- Cost: Free
- Participants: Open to anyone interested in this experience: beginners welcome! débutant(e)s bienvenu(e)s! beginners welkom! Experienced Wikipedia editors will be present and will share their knowledge in editing Wikipedia.
- What to Bring: Attendees can bring their own laptops and power cords.
- A temporary library will be set up by the participants who are invited to bring their books and digital documents to share with the others!
During the event, we invite you to contribute to Wikipedia and to the discussions around the gender gap on Wikipedia. You can edit an existing article to improve it, create a new one about a subject that doesn’t exist, but we also highly value the sharing of editing experiences, and ideas about what could make Wikipedia a more welcoming and colorful place!
Here are interrogations we would like to share with you:
- if Wikipedia can be considered as a tool to re·write history, how would you like it to be written?
- what is the influence the way we write on the representation of history and its main figures?
- can writing ever be neutral?
- Add your name to the participant list of this event at the bottom of this page
- Write some informations on your own wiki page by clicking on your user name at the top of the page (so that your name doesn't appear in red when you start editing articles)
- Write some informations on your Sandbox, also at the top of the page
- Improve an existing article
- Create a new article
Here are lists identifying articles that could benefit from edits and expansion:
- The Art+Feminism list of article to be improved
- Women Artists, by the Women in Red
- A list by the project Women in Red
- Women Artists from all over the world, a list by the Project Women Artists
- Women scientists, a list by the Project Women Scientists
- Women writers, a list by the Project Women Writers
- These lists mostly link to English Wikipedia, don’t forget your own language’s Wikipedia!
- Man as false generic: Ban the use of the words man, men and mankind to refer to a person or persons of unspecified sex or to persons of both sexes.
The page Writing about women offers great insights. Look for the following problems in existing pages and try to fix them:
- Male is not the default: Avoid labelling a woman as a female (ex: author, politician etc.), unless her gender is explicitly relevant to the article. An opposite example is saying male nurse.
- Use surnames: Look for articles using surnames for men, while calling women by their first name. See example
- Infoboxes are an important source of metadata (see DBpedia) and a source of discrimination against women. For example, the word spouse is more likely to appear in a woman's infobox than in a man's.
- This is a good and terrifying exercise, try to find a page (that is not) defining women by their relationships in the first paragraph.
- One study found that women on Wikipedia are more linked to men than men are linked to women.
- Use gender-neutral nouns when describing professions and positions
- Try changing the order in which groups are introduced – man and woman, male and female, Mr. and Mrs., husband and wife, brother and sister
- The use of the generic he (masculine pronouns such as he, him, his) is increasingly avoided in sentences that might refer to men and women.
- Avoid problematic phrasing when talking about married people. See examples
- Do not refer to adult women as girls or ladies!
- Avoid images that objectify women!
- Art+Feminism resources
- Gender gap resources
- Writing about women
- Gender neutrality in English
- Avoiding Heterosexual Bias in Language
- the he-she package
- Every Wikipedia page has a Wikidata page attached (database of normalised information that can be read by humans and machines: datas): property P21 is “sex or gender”. Gender was one of the first properties (21st) defined! It has an item number. Sex or gender: link to a value, or many values (like woman, trans*, many other possibilities, …)
- Gender is a “Statement”, examples of others include "instance of" (“is a”), "date of birth", "first name", etc
- To get the Wikidata item, go on the left menu, under the “Tools” section.
- You can create a Wikidata page before the Wikipedia articles.
- Some categories implicitly identify gender: e.g., “women chemists”.
- “I don’t trust my bot” - it's Ok to make imperfect bots if you (a human) skim through the results and double check with your human eyes/understanding - still more efficient than do everything yourself/
- Which language to write: Python is the easiest programming language, but other languages can be used.
- Bots can be run on your computer, or uploaded in the tools server of Wikipedia
- So you could fake that you’re a bot! You could ask for being a bot, then being flagged as a bot, and then your edits do not show up in "recent changes"
- Probably this kind of behaviour is detected by "metabots"
- How can we imagine a bot that can identify gender-biased articles. In a “constructive” way (without focusing only on “vandalism”, without only deleting but proposing…)
- How would you find other biases?
- Will the bot be able to point at things that we cannot detect? (humans should be trained to detect)
- Guidelines for bots / for human?
- Can the code be able to work somewhere else, outside Wikipedia? (Yes)
- For instance: go and look on pages like “Writing about women”.
- Name? Gentlebot? Genderbot?
- Pattern: if an article is about a man, he’s addressed with the surname, and if it’s a female she’s addressed with her first name. This pattern can be expressed algorithmically:
- Limit the field: specifying the properties you want (ex: articles about women chemists)
- Identify the surname and the first name: occurrences of both in the article: if the first name is used more than *** times, then mark the article “possibly gender-biased”.
- Sometimes, there’s only a first name (Madonna) or Royal people are named in a specific ways.
- Pattern: articles that talk first about the brother/father/… in the first sentence, and even in the first paragraph.
- Family-related information in the first paragraph. (Basically: it is not a defining characteristics)
- Some very rare corner cases are counterexamples: "princess X is notable for being the only child of king Y"
- How to identify this?
- Find words like “mother”, “father”, …
- Links to an article with the same surname in the first paragraph.
- Collaboration human-bot…
- Can it take the sentence and put it under a family section? Or at least make a proposition that can be validated by a human editor. Sometimes it is already in a family section, but repeated on top. (so then it needs to be removed)
- Can it change the name? (add the surname when it misses)
- Is there space for proposition of edits?
- The bot could post it on its own page or on the talk page of the article: “Why don’t you move this sentence in a specific part?”
- The bot could add a template banner on a page.
- Link sex tape scandals with revenge porn. Sex tape scandals are glorifying but also destroying lives and careers. Things should stand out sometimes, but usually "hidden" and obscured, like being one item in a list of political achievements.
- Sexual harassment: considered as a morality issue, which is problematic…
- Search more about the concept of morality (in French)
- Further research is needed
- Article about refugees: quite gender-neutral, but needs a specific section related to women.
- Gender-specific and gender-neutral pronouns.
- Using standardised language is based on assumptions.
- Generic pronouns: he used as a generic use (“neutral”) and “grammatically correct”, generic she: “When most members of the group are usually thought of as women”.
- "Secretary and her temper", etc. as bad examples of legitimate language usage
- If it’s written like this, then it’s never going to change… Add more references that nuance the phrasing of this article.
- See Spivak: set of neutral pronouns that can be used.
- Saying on Wikipedia that “this is grammatically correct” is not Ok since it does not faithfully represent the reality.
- Saying on Wikipedia that “I do not think this should be correct” is not Ok since it does not faithfully represent the reality.
- If the issue is complicated and changing, Wikipedia should embrace it and describe it as such: "some older books see it as grammatically acceptable (source), some modern scientists see it as a problem (source), there are possible solutions discussed: (1), (2), (3)", etc
- Move female explorers on the List of female explorers and travelers to the general List of explorers
- Revenge porn research to improve the page
- /info/en/?search=Hero#Heroine_or_hero added a new quote to the section
- Craig Owens, american post-modernist art critic, gay activist and feminist french translation in progress
- Margaret M. Adams on it's way
Sign here to help us learn more about how users interact with Wikipedia! If you sign up, we can observe how your username uses the Wikimedia projects during and after this event. This will help us to better measure the effectiveness of this event and similar programs. This means that your publicly available activity and the information you share with us during this event may be processed by the Wikimedia Foundation, Art+Feminism, and the organizers of the local event, and may be transferred to or from the US and other countries that may not have the same level of privacy regulation that your country does. However, we will not share your information with third parties or publicly unless it's in aggregated or anonymized form.
Prior to the event:
- Do you have a Wikipedia User Name?
- No? Create a Wikipedia account
- Yes? Go to Step #2
- Sign up! Add your Wikipedia User Name to this section by clicking the blue button below (follow instructions). Your name will be added to the bottom of this page
Bots, Robots and Cyborgs - Part 1 10.06.2016: 17 people present
Bots, Robots and Cyborgs - Part 2 18.06.2016: 26 people present