Post 10: Constructive Idea Discussion

By Basilia Dulawan

Reflection + Summary

Sharing my draft proposal with a few peers as well as Jacquie was a really beneficial exercise as it challenged me to think about how I would actually approach the design proposition, what parameters I needed to include, and shown what needed to be stronger in my proposition.

Initially the generative system I proposed was going to rely on data generated from Twitter, catching tweets that used the phrases “He is a…” or “She is a…”, but the problem that Jacquie highlighted was that it was too generic and the responses can be taken out of context. Jacquie suggested that I instead create a survey with a series of semi-specific questions that participants can answer. Additionally, to avoid skewing the results toward people who share similar mindsets to me, it was raised that I shouldn’t just post this on my Facebook page, but use Reddit and other survey forums to allow a greater variety of people to respond. This was a critical realisation for me as I wanted to be able to produce a variety of honest answers, but didn’t think beyond the way in which I know how to scrape the web for data, which is Twitter. The next challenge was writing semi-specific questions that didn’t probe at a certain response.

Another key moment that came out of this discussion was Jacquie suggesting that I speak to Chris Gaul and Thomas Ricciardiello about ways in which I could make the generative system update in real-time, and use processing to make sense of the data generated. Overall, through this discussion I was able to distill what it is I wanted to investigate further, and how my proposition would respond to this.


Project Title: RE-THINK. 

Practice Type: Generative System (with a side of Data Visualisation)

The Issue: Gender Inequality in Language

The Possible Change: Awareness about the gender inequalities that exist in society’s everyday language and the realisation that it is through language that we support the continuation of Gender Inequality. With this awareness, people can make a conscious change in the words they choose to use toward Women and Men. 

Design Action to support change:

A key moment in my research was listening to filmmaker and documentarian Lauren Greenfield emphasise the power of words, and how they shape the development of young girls. I was particularly inspired by her work for Always with the initial #LikeAGirl Campaign which she aimed to change the perception of the ‘Like A Girl’ phrase, from an insult to an empowering compliment that girl’s can own. Coupled with another key finding of the way in which we raise boys and girls that made me question –  Why is it that society raises boys to be brave, but girls to be cautious, and lady-like?

The common element throughout my research was, language. How women refer to other women, how women refer to men, how men refer to women, how men refer to other men and how we describe ourselves. For my design proposition I want to explore gendered language, and how it is used, if used, by 18-25yr olds on social media. I aim to create a generative system that aggregates data/user responses, then visually plots these words on screen. As the data grows over time, the visual produced will update in real time – increasing the size of words that are more commonly used, plotting new words and changing the colour of the text depending on what gender it was used to describe. What I am hoping to achieve is a visual that illustrates the gendered language we  as 18-25yr olds use, and with this, bring an awareness to the change that needs to occur in our everyday interactions simply by the language we choose to use. I hope that this generative system that can be visited online, makes everyone – specifically 18-25yr olds, more aware of the words that support the continuation of Gender Inequality and thus more conscious about the words they choose to use toward men and women.


What are we talking about? The Language of Homelessness

Post 10 By Alice Stollery

Reflection & Revision

My colleagues raised a number of issues with my initial draft proposal. Being my first attempt, it was quite rough and I had not given enough consideration to the requirements of the brief.

My colleague could not see the link between my proposal and the required 18-25 year age bracket. And she was right, I had got so carried away with my research that I had lost sight of this requirement. She questioned whether this issue of language had come out of this age bracket or whether I was attempting to target 18-25 year olds with my design proposal. Reviewing this point, I will use tweets that have come out of this age bracket while contrasting their misuse of language with facts and statistics that focus on homeless youth within Australia. I also aim to target the 18-24 year old age bracket through my design proposal, by basing the exhibition at the UTS campus or other university campuses. I would like to target this age group, as I believe it is important for them to be empathetic towards this issue as they are the next generation of leaders, teachers, politicians and by starting with them, I will be able to instigate change in the future. Their views on this issue are incredibly important.

Concerns were also raised with the location or geographical nature of my data. Am I able to tell where tweets are being tweeted from and whether this issue of language is an issue that occurs within Australia. Reviewing my data, I have found that terms such as tramp or hobo are geared more towards an American context while misusing the term homeless occurs within Australia. Therefore I have narrowed my focus to the misuse of this term. I have also experienced the misuse of language in my daily life long before this assignment, throughout school, work and university. It is not unusual to hear someone describe themselves or their friends as looking homeless. However, it was not until I saw all of these comments collated on a spreadsheet of tweets that I was able to recognise language as a key barrier in solving the issue.

Another piece of useful feedback included the form of my response. I was told not to limit it to a book so I have given further thought to how this data could be represented. I have decided to create a public installation or exhibition that could possibly include posters, flyers or brochures as well. I will elaborate further on this in my proposal below.

The Issue (From research)

The misuse of language is a significant barrier in tackling homelessness. Insensitive, and politically incorrect terms such as hobo, tramp and bum, and the casual misuse of the term homeless to describe ones appearance, have seeped into the common vernacular. Dehumanising those affected by homelessness through this passive misuse of language takes away from the real issue, meaning wider perceptions of homelessness are less empathetic and communities have become detached from sufferers.

Through research into homelessness in the mainstream media, journal articles, social media platforms, image libraries and brainstorming sessions, the misuse of language and terminology around the issue has emerged as a negative actor that is creating a barrier between those in need and those with the power to help. We talk about homelessness in reference to appearance, rather than experience. In short, we no longer seem to be talking about the actual issue.

Possible Change

A design response that tackles this wider problem of perception and language will create influence rather than direct action. This will be an attempt to create internal change in those that misuse these terms, in order to create empathy, and ultimately to generate positive outcomes, enabling more people to engage with the issue rather than offering an immediate solution.

Design Action to Support Change: Data Driven Design

An exhibition titled “What are we talking about?!”  that aims to juxtapose the the misuse of the term ‘homeless’ in everyday conversation with the real issue and experience of homeless youth. Ultimately highlighting the disconnect we are currently experiencing between the two. The exhibition will be a visualisation of data collated from twitter and online statistics on youth homelessness collated during the research process. It will be a contradiction of meanings within the same issue and will highlight how language is acting as a barrier in our ability to help the homeless.

Sketch of the exhibition space depicting alternative perspectives of the issue.


I will design the exhibition, mapping how the audience will move through the space as well as designing the look and feel for the exhibition, including collateral such as postcards and posters. The exhibition will be a series of hanging posters that enable you to see both sides of the issue. Looking in one direction you will be bombarded with the misuse of language as you see tweets that misuse the term homeless, for example “OMG I look so homeless today” or “That moment you look at a new pic of your ex and wonder how you could have dated him. #whatwasithinking #lookinghomeless” while the other side will contradict this with overwhelming statistics about youth homelessness such as “How can we still call Australia home when 32,000 young people don’t have one?” or personal experiences of sufferers such as “My friends don’t know I’m homeless”. The idea is that while you are looking in one direction at the language we use, you are unable to see the real issue  on the other side of the posters and as a result you are unable to empathise with sufferers. If you choose to talk about homelessness in this way, you are unable to be empathetic and to understand what sufferers are really going through. Visualising and organising data in this way will enable people to see both sides of the issue, one at a time and will hopefully generate internal change within the audience without publicly shaming those who have used this language in the past.


Post 8: It’s all in the language

By Basilia Dulawan

In this week’s session what I found most beneficial was being able to work individually on our own issue before brainstorming as a group for design possibilities. Going through each of the questions ‘Who does the problem affect?’, ‘What are the boundaries of the problem?’ and specifically ‘When + Where does the problem occur?’, made the design problem seem a lot clearer. Language was something that kept coming up in my answers to the questions, and that’s what I based the brainstorming of possibilities around.

The boundaries of the problem is representational – gender norms and stereotypes, and the language used.

Who does the problem affect?
Adolescent girls & boys particularly those going through puberty as this is their formative stage that influences their confidence, mindset and perspective.

What are the boundaries of the problem?
Boundaries are representations: Gender Norms and stereotypes as well as the particular language we use that can be skewed to one gender.

What if the problem was fixed?
Girls could feel as thought they really could do anything, additionally, once they transition into adult life, they would have less gender norms and stereotypes to influence their career choices. Additionally the perception of Boy and Men would also change and become more free and open.

Young girl’s lack of limitations in their mentality that they feel in their pre-pubescent stage, as well as their confidence, can therefore carry through to every stage of their life instead of dropping in puberty.

What if the problem wasn’t solved?
Society and the way girls loose confidence in their adolescent years, would just continue at the same rate as it is going now.

When does the problem occur?
During adolescence for girls, particularly through everyday interactions with adults, friends, parents and the language they use around and toward them.
i.e raising boys to be brave and girls to be proper, cautious and ladylike.

When does the problem need to be fixed?
Now ideally, but specifically in the early stages of primary school all the way through highschool. It requires education and awareness between young girls and boys – and even their parents.

Where is the problem occurring? 
At home, at school, in the media and entertainment, in the playground, in clothing stores, in books and in extra curricular activities.

Why is it important that the problem is fixed?
So that Women and Men have equal opportunity. So that Women feel just as empowered as men to do what they want. So that there is no or at least less pressure of how Men and Women should be.

Issue Statement

The root of Gender Inequality starts with the way we raise our boys and girls, particularly with the language we use when we address, encourage and treat them day-to-day. 

> Initially, I had forgotten about the target audience being 18-25, so this issue Statement was mainly targeting Parents, Teachers, Coaches – basically anyone who has an influence of the development of children. This issue is highlighted in this article and these case studies here and here.

[Amended] The root of Gender Inequality starts with language. The way in which we interact with men, women and children in our families, friendship circles and workplace – particularly with our choice of words we use to describe, compliment, encourage, or even put down, can be very different between each gender.

Five point summary of design possibilities:

  1. Create a generative system that looks at the language used when describing women and men on twitter. For example, every time the word ‘strong’ is used to describe a woman the size of the word is increased by 1pt, and more pink is added to it’s colour value, and if it were used to describe a male, more blue would be added to it’s colour value. In this way, we can identify the language which we have applied gender to, as well as language which isn’t separated or reserved for a specific gender. (Generative System/Data Visualisation)
  2. Inspired by the work of two Australian advertising executives Georgia Patch and Kiah Nicholas, who noticed that google’s definition for words like “housework” and “promiscuous” were only described as female “she was…” or “her mother…” and created a social movement on Instagram @redefineWomen #redefineWomen, I propose visualising the language used to describe women throughout history. I think it would draw out some insights in how the representation and perception of women has changed (hopefully), and possibly highlight how outdated these current definitions used by google are.  (Data Visualisation)redefinewomen
  3. Scraping social media such as Twitter on Instagram to generate data of three things:
    Men describing Women
    Women describing Women
    Women describing themselves
    Using this data I would create a visualisation of these words keeping into account how many times they were used and from which context (Men describe Women etc), and I’m not sure what the results would be, but based on this interview in which Taraji P. Henson says “When you [Women] embrace each other and you stick together, look at what these women [in ‘Hidden Figures’] were able to accomplish.”, I think what’s important to ask is ‘How are Women portraying themselves?’, ‘What language do Women choose to describe each other? Is it at all different to how Men describe Women?’. By visualising this, I think it would bring an awareness to Women that we need to support each other and to Men that it’s not ok to describe us in a sexist way – but in saying this, I could never be too sure that this is what the data would highlight. Nevertheless, it would be interesting to explore.
  4. Data visualisation of words associated with Gender Norms and Stereotypes as well as words that have been used outside of those ‘traditional’ terms. I think it would also be interesting to include the time/year in which these new words started to be used in relation to Women. The main thing here that it might show would be the progress we as a society is making in the perception of Women and their abilities.
  5. Researching influential Women throughout history and visualising the language used to describe them as well as the gender of the author who wrote the description. This would be another way in which we could see the progress society is making in the perception of Women and their abilities throughout history.


Draft Proposal 

Words are something we use everyday. When we speak to people physically or through the many possibilities of the internet, we use words the convey what we mean. But what if what we think we mean, isn’t actually our intention? Words are powerful. They carry meaning beyond their typical definition and through history, take on their own connotations. As history changes, how too have our words? The root of Gender Inequality began with the formation of traditions and gender roles, but what stems and supports it is language. The way in which we interact with men, women and children in our families, friendship circles and workplace – particularly with our choice of words we use to describe, compliment, encourage, or even put down, can shape the very people we interact with. That being said, I want to look at the language we use toward each gender and explore what words cross over and what words are more skewed toward one gender.

Therefore, considering the target audience of 18-25yr olds spend a lot of their time on social media – Facebook, Instagram and Twitter, I propose that I create a generative system that uses data from a social media channel such as Twitter to bring insight into the words we use to describe males and females, in an attempt to understand in what ways Gender Inequality exists in our language. The generative system would be based on Twitter data that is generated over a 24hr period from users based in Australia or more narrowly, New South Wales, every time a word is used to describe a female or a male using the phrases “She/He is…” or “She/He was…”, the words at the end of the phrase will be plotted onto the generative screen. As certain words or phrases are repeated in the data, the word plotted with grow in size and opacity. Colour will be used to visualise what gender the word was used to describe – instead of using the typical pink and blue, I propose using alternative colour combinations such as purple and orange or pink and green. This is an important aspect of this generative system as it can really emphasise the results and possibly make these gendered words more clear.

With this approach it would be interesting to see what words are more commonly used to describe males as opposed to females and vice versa. With this focus on language, it could open up the conversation about the subtle yet impactful ways Gender Inequality exists in our society. One thing that I want to focus on in my design response is to be able to understand the human aspects of the data – this is something Jer Thorpe emphasises in his 2011 TED talk.




Redefine Women, 2016, Instagram, viewed 6 September 2016, <>.

Redefine Women, 2016, Instagram, viewed 6 September 2016, <>.

Redefine Women, 2016, Instagram, viewed 6 September 2016, <>.

The Hollywood Reporter 2016, Taraji P. Henson, Octavia Spencer “Power of Women Sticking Together” in ‘Hidden Figures’ | TIFF 2016, video recording, Youtube, viewed 10th September 2016, <>.

BLOG POST 6: Scraping the web for data

‘Lexicons + The Internet Language’

The history and context of language are always changing and developing. As the emergence of technology and the Integration of the Internet changes the way we consume media. Our linguistics and vocabulary also expand. Social Media in its own platform is a major contributor in the ways we communicate visually and audibly. The format and structure of social media influences writing styles as well as content. Twitter is a new form of media that delivers its messages in a 140 character limit. This restriction creates a succinct, creative and empowering conversation that users are easily able to engage and scroll through.

Lexicons are a linguistic resource that we use to understand the vocabulary of a person in association to words of sentimental value (emotions). Whether they’re positive, negative or neutral. I.e. ‘NO!’ and ‘no’ conveys a different tone of voice and with the slight alterations in its composition, It delivers a different message. Twitter is a primary social media platform that deals with languages of informal expressions. Generally a collation of data and colloquial expressions. Such as acronyms, the use of incorrect spelling/ terms and abbreviations. Due to the vast majority of language expressions and variable factors, It is difficult to determine whether the responses are of sentimental value (positive, negative or neutral) therefore the use of emoticons are applied.

Emoticons are a highly recognised attribute to the Internet language. The use of visual expression displays a greater range of sentimental values and is a language technique globally practised. Emoticons are considered to be opinion lexicons and are stable for sentimental classification, unlike literal words.

The default Twitter search allows users to add emoticons to the search to find positive/ negative tweets. The majority of tweets does not contain emoticons which impact the search and statistics by DTA: 25th Australasian Database conference shows that only 9.40% of tweets in 2011 contain at least one emoticon. 7.37% of that is positive and 2.03% negative. (Mohammad, S, A. Wang, H. 2014). Due to these results, It shows a decline and insufficient use of lexicons and emoticon limitations.



Twitter features using # syntax as a mean of collating tweets into categories and as a new form of internet language. Hashtags are also a form of metadata by collecting words of the same topic giving context to the tweet. For example #idontwanttowritethisblogpostanymore groups tweets with similar concepts. Although topics that are not typical are often more difficult to evaluate and contribute to the global expansion of lexicons, providing better performance to searches and collation of material.


Sharaf, M.A., Wang, H. 2014, ‘Databases Theory and Application: 25th Australasian Database Conference, Springer, Brisbane, viewed 4 September 2016,
Bravo, F. 2016, Lexicon Expansion, viewed 4 September 2016,
Reed, J. 2014, How social media is changing language, blog post, viewed 4 September 2016,


Post 2: ‘Illegals’–A semantic framework

By Olivia Tseu-Tjoa

Last week, my interest in the language used to describe asylum seekers led to an investigation of the rhetoric surrounding refugees in an Australian context. Through my readings of scholarly articles, it became increasingly apparent that the government’s rhetoric is reinforced by the media, therefore shaping public opinion on refugees and asylum seekers. 

Call me illegal
Ben Doherty’s dissertation,‘Call me illegal: The semantic struggle over seeking asylum in Australia,’ offers a historical overview of the language and its relationship with media, government and public opinion. Interestingly, he notes that there has been a significant change and shift in attitudes towards asylum seekers. It covers the various events and ‘waves’ of asylum seekers from the aftermath of the Vietnam War in 1976, up to recent years. According to Doherty, phrases such as ‘boat people’, ‘queue-jumpers,’ ‘illegals’ and ‘coming in by the backdoor’ emerged in Australian vernacular in the late 1970s during the first ‘wave’ of refugees from the aftermath of the Vietnam war.  He examines the different narratives portrayed and the government’s response through primary records, such as government documents, to secondary sources like media reportage. An example of this change he notes is that 
‘in the 1970s it was argued by government that Australia had a humanitarian obligation to assist and accept boats trying to reach Australia. In 2013, it was posited that it was humanitarian to forcibly push boats back (Doherty 2015).’  He claims that the language used has become hostile and is a deliberate shift in conjunction with Australia’s tough, ‘hard-line’ policies, significantly influencing public opinion. Also, the author argues that Australian journalism perpetuates this rhetoric as it has easily embraced this shift in language without challenging it. This is an interesting position as Doherty himself is an Australian journalist and Immigration correspondent for the Guardian Australia who has frequently reported on the topic of asylum seekers and has recently worked on the release of the Nauru files. Although published in the University of Oxford, Institute of Journalism, the Guardian itself is a ‘left-wing’ newspaper, which may indicate the author’s bias.

Politics, media and the community’s attitudes
In the second scholarly article, ‘It Would be Okay If They Came through the Proper Channels’: Community Perceptions and Attitudes toward Asylum Seekers in Australia,’ the academics, Fiona H. McKay, Samantha L. Thomas and Susan Kneebone conducted a community survey looking at the perceptions of asylum seekers and the socio-cultural factors at play. Through qualitative and quantitative methods, participants responded to the survey and a series of questions relating to asylum seekers, while also recording their characteristics such as age, occupation and education. The researchers categorised their responses into three broad themes:

  1. ‘Asylum Seekers Exploit Australia’s Democratic System’
  2. ‘Asylum Seekers Threaten Australian values and identity’
  3. ‘Asylum Seekers Threaten Australia’s national security’

Negative opinions from respondents claimed that media reporting surrounding asylum seekers indicated that they were unwilling to assimilate into Australian culture. Certain participants who were highly educated and from high socio-economic backgrounds, pointed out that the imbalance in media reporting created ‘fear campaigns’ of the impact of asylum seekers in society. Evidently, there is some level of recognition of the media’s influence in forming public opinion among participants in the study.

“When writing about the method of arrival of asylum seekers, respondents often used the terms, ‘illegal’, ‘illegal asylum seeker’, ‘boat people’ or queue jumpers’. Some wrote that asylum seekers ‘cheated the system’ by not following refugee processing procedures in their own countries before travelling to Australia (McKay et al. 2012).”

McKay, Thomas and Kneebone came to the conclusion that there needs to be a shift in political and media rhetoric in order to change these negative views of asylum seekers. The authors are academics and sociologists (from Monash University at the time of this study) who appear to be sympathetic to asylum seekers, as they hope for more balanced reporting and to dispel the ‘threat rhetoric’ in public discourse. 

The pursuit of balanced rhetoric
Both scholarly articles highlight the way in which the media supports the government’s negative rhetoric, such as a perceived threat to Australia’s national security. Doherty’s paper claims that the government’s harsh rhetoric and change in language is echoed by the media, while McKay, Thomas and Kneebone’s qualitative data confirm the media’s impact on community attitudes towards asylum seekers. Both articles conclude that there should be a balance in media reporting, which is factual and objective. The dynamic interplay between media reporting and political rhetoric has proven to be influential in forming community perceptions of asylum seekers. As Doherty writes:

In a global order predicated upon nationality and bounded territoriality, people displaced from their homelands lose rights and lose agency. Often, they are voiceless in public discourse, defined by the language used by others to describe them. Their image – the public’s fundamental understanding of who they are – is created not by themselves, but by others (Doherty 2015).

Suffice it to say, words do indeed matter.

Doherty, B. 2015, ‘Call me illegal: The semantic struggle over seeking asylum in Australia’, Reuters Institute for the Study of Journalism, University of Oxford, viewed 6 August 2016, < >.

McKay, F, Thomas S and Kneebone S. 2012, ‘It Would be Okay If They Came through the Proper Channels’: Community Perceptions and Attitudes toward Asylum Seekers in Australia’, 2012, Vol 25, No. 1, Journal of Refugee Studies, Oxford University Press, pp. 113-133