Data analysis identifies three dominant discourses in the popular response to coding for all: the digital ubiquity, the disadvantage and the habits of mind discourses. But before data informing these discourses are identified, it is worthwhile noting the relative ‘strength’ of the discursive themes in terms of the number of individual expressions or arguments within broad groupings deployed by each case. In terms of themes shared by both cases, among the strongest are the economic, personal benefits, thinking skills/creativity, technologies/languages, social justice, literacy and engaging AI/computers/robots. Within these, the Yes case dominates in terms of the number of arguments articulated in all except for the social justice and AI/computers/robots themes. In terms of themes deployed by the Yes case alone, a number of these were relatively weak in terms of the number of individual arguments deployed. By contrast, several of the arguments within the No case alone themes were relatively strong. These included overstating the easiness of learning programming, and coding in schools being too esoteric and of little use. However, other than these perhaps ‘outlier’ themes, the majority of the arguments in the popular response are contained within broadly shared themes, though argued from oppositional positions. This coalescence and coherence of argument around strong themes informs the dominance of the discourses identified in this debate.
Digital ubiquity discourse
Both sides in the debate deploy a digital ubiquity discourse. The Yes case maintains the digital impacts all aspects of our lives (Ruseva & Rissola, 2016), and that computers are pervasive (Kafai & Burke, 2014, p. iv), while the No case acknowledges this is ‘an increasingly tech-driven world’ (Editorial Board, 2017, para 1). Argument diverges, however, in relation to providing evidence supporting the representation of digital ubiquity and in constructing what digital ubiquity implies for education.
Regarding evidence, the Yes case (like the No case) is grounded in sources of textual authority, citing industry reports, government data and authoritative voices. We are told that ‘six of the world’s ten most valuable companies are technology companies, with Apple leading the pack’ (Broadband Solutions, 2016, para 1). We are also told that according to the Australian Workforce Productivity Agency, ‘in 2025 there could be an undersupply of qualifications for key ICT occupations, with employment projected to grow between 64 and 72 per cent faster than overall employment growth’ (Carroll, 2016, para 3). This argument references Australian Chief Scientist Ian Chubb, stating that ‘if the digital economy is the arena, then the skills you need to play include computer programming and coding’ (Carroll, 2016, para 5).
Consequently, the Yes case constructs the economic value of learning to code in terms of the digital economy and its skills requirements. Cohen, for example, reproduces statistics from Gartner showing that globally in 2015, IT spending would not only reach $3.5 million, but use of digital technologies would ‘grow the size of the global GDP by $1.36 trillion in the next 5 years – all of those dollars equal opportunity for the people who are fluent in tech’ (Cohen, 2015, para 2).
But as the reference to Gartner suggests, the Yes case represents the economic need and employment opportunities as not only of national significance, but also international. Puiu notes that ‘according to the European Commission, there will be a shortage of over one million programmers in Europe by 2020’ (European Commission, 2016, para 1), while according to Mayo, ‘in the US alone there’ll be a million more computing jobs than computing science graduates by 2020’ (Mayo, 2013, para 1).
For some in the No case, the economics argument is not so much an opportunity for workers as for ‘capital’, as argued in the literature by Bresnihan et al. (2015). Silicon Valley, it is claimed, ‘has been unusually successful in persuading our political class and much of the general public that its interests coincide with the interests of humanity as a whole’ (Tarnoff, 2017, para 12). Tarnoff adds, ‘it will proletarianize the profession … flooding the market and forcing wages down – and that’s precisely the point’ (para 4). The relevance of this argument is somewhat validated by Yes case discourse, not in the sense of a capitalist conspiracy, but in recognition that an increased supply of developers will lead to lower wages. Yes-case proponent, McAllister (2008), suggests that ‘computer literacy should be seen as a baseline skill for the U.S. workforce, not a differentiator’ (para 20) leading to a ‘ticket to a golden future’ (para 20).
Referencing economics, the Yes case also highlights ethno-national divisions as justification for compulsory coding. This boundary constructing position references fear and danger, expressed through the notion of crisis and risk (Bagshaw, 2015; Mayo, 2013). In the Australian context, Carroll (2016) highlights the ‘global nature and ferocity of the competition’ in software development (para 5). Others maintain that without ‘compulsory’ coding, there is a risk that ‘Australia [will be] left behind in the digital age’ (Calixto, 2015, para 1). Regarding the USA, Bajarin (2014) notes that it is ‘far behind in having a robust technical workforce created within our own boarders’ (para 13). Reference to borders and ‘us’ and ‘them’ is significant in the Yes case. Regarding innovation, we are told that coding in schools will ensure that the Microsoft and Google continue to be grown ‘at home’, in the USA (Cellan-Jones, 2014). Elsewhere, ‘we’ are warned of reliance on ‘foreign’ workers, including those from ‘a variety of Asian companies’ (Debate.org, 2013, para 3), and specifically ‘coders in China, India and other parts of the world’ (Bajarin, 2014, para 12).
The Yes case’s deployment of the ubiquity discourse further represents compulsory coding as an individual and social good. Lexical items such as ‘empower’ (European Commission, 2016, para 4) and ‘enhance’ (Lucisano, 2017, para 2), construct the value coding skills add to the individual. Compulsory coding will enable individuals to understand ‘how their devices work’ (Bajarin, 2014, para 9), allowing them to fix them (Hinsliff, 2015). It will allow individuals to know ‘how [digital technologies] operate and how to customize them for better functionality’ (Lynch, 2017, para 3). Coding, Enobrev argues, will help students gain a ‘better understanding of the world and hopefully better interact within (sic)’ (Atwood, 2012, para 3). Agon likewise associates coding with how to better ‘live in this new economy and society’ (Rawlins, 2015, para 5).
By contrast, the No case asserts, ‘no matter how pervasive a technology is, we don’t need to understand how it works’ (Felker, 2013, para 9). Constructing less relevance between digital ubiquity and education, the No case highlights the division of labour and employment specializations. Argument maintains that as the employment sector is characterized by different people doing different jobs, the need to understand how code works in order to fix something is mitigated by the existence of those whose job it is to fix such code. As Felker notes, ‘society divides it labour so that everyone can use things without going to the trouble of making them’ (para 9). A further argument advises the reader not to overlook that society still requires a broad spectrum of employment fields.
We’re always going to need doctors and nurses, teachers and chefs, all sorts of jobs that don’t need to write computer code at all. To assume that everyone should be required to learn it is to be caught in the bubble of your own profession (Hall, 2017, para 8).
The ‘ubiquitous though less relevant’ argument proceeds by highlighting the extent to which ubiquity in other fields does not mandate subject specialization. Parallels are drawn with music (Rawlins, 2015), motor vehicles (Atwood, 2015; Gray, 2014), plumbing (Atwood, 2012), carpentry (Bethune, 2016), aircraft (Gray, 2014) and electrical work (Cohen, 2015). Stucky (2015) notes that he does not use any of his coding knowledge to shop or bank online, suggesting that:
For the majority it’s probably more like understanding a car engine. You don’t need to be able to strip an engine and rebuild it to drive a car, but if you basically understand how a car works you can drive and maintain it efficiently and effectively (para 9).
Data show that argument in disadvantage discourse is more explicitly grounded in liberal notions around resource distribution (Riberio, 2014). The Yes case is grounded in a Rawlsian liberal-distributive framework, the No case in a more libertarian position. Both, however, assert disadvantage. The Yes case maintains those without coding skills are, and will remain disadvantaged, without state intervention. The No case maintains that state intervention disadvantages those already equipped with skills. Both argue from a position grounded in pursuit of liberty.
At its broadest level, the Yes case challenges a (claimed) elitist image of software development, on grounds that it is uninviting to the underrepresented masses (Raja, 2014). Compulsory coding, it is argued, will demystify programming (Rawlins, 2015). A challenge to ‘elitism’ is also expressed through re-asserting the value of amateur coding as opposed to professional. Neil argues that ‘it skirts close to snobbery to discourage somebody merely on the grounds that it isn’t their day job’ (Atwood, 2012, para 73).
Specifically, the disadvantage discourse foregrounds equalizing access as the principal justification for compulsion. Compulsion, it is argued, will prevent disadvantage arising from some schools implementing coding and others not, and boys choosing coding and not girls. Compulsion will equalize access to a field often unavailable to or not accessed by underrepresented students (Della Cava, 2015; Gilbert, 2017), exposing their talent (Guzdial, 2014). As AnonymousCoward, 2012, para 53) notes, ‘it is a meritocracy, and it should be - but if you don’t look for talent, you won’t find it all’. In sum, as argued by Lyonnais, compulsory coding might ‘level the playing field’ (Lyonnais, 2015, para 8), particularly in terms of gender (Gilbert, 2017). Compulsion will allow the underrepresented to exercise choice about the relevance of coding to later school and post-school life (ECDL, 2015).
Like the Yes case, the No case deploys a disadvantage discourse across several fronts. One concerns existing inequalities, another concerns likely inequalities, and yet another responds to (in) equality arising from state intervention in the coding ‘market’. With regard to the former, Keneally (2015) asserts there are unresolved matters in school education more significant to improving student outcomes. Citing Krugman’s commentary, Keneally reaffirms that while ‘talking breathlessly about how technology changes everything might seem harmless … [its] a distraction from more mundane issues’ (para 11), adding:
that includes issues such as failing schools, poor teaching quality, lack of specialist support for students with a disability, the increasing shift away from comprehensive publicly funded schools to private education, poor international rankings in literacy and numeracy, and countless other real, and very difficult issues, that confront Australia’s education system (2015, para 12).
Keneally concludes that ‘sorting out those problems will take much more than a few lines of code’ (2015, para 13). Here, compulsory coding perpetuates existing disadvantage. Elsewhere teaching quality is highlighted. The demand for skilled coding teachers, which may not be met, will result in unqualified teachers delivering the curriculum, resulting in a watered-down curriculum (Guzdial, 2014, para 19). Guzdial (2014) adds, regarding inequality in terms of access, ‘let’s work first at making it accessible, before we try to require it’ (para 19).
The No case references the ‘crowded curriculum’, constructing a concern for teacher welfare (Debate.org, 2013; Dodge, 2017) and asserting the greater value of other subjects to the liberal-humanist project of dispelling ‘ignorance’. While this argument does not contest the role of the state in providing children with access to knowledge supporting liberty, it represents coding as deleterious to the capacity of ‘more valued’ subjects to achieve this goal. At Debate.org, a post in the ‘con’ argument asks: ‘what subject will you be robbing these children of, Science? Math? History, or the already underfunded art and music departments?’ (2013, para 6). Aside from representing compulsory coding as robbery, argument highlights the desire of some in the Yes case to have humanities subjects make way for coding, such as Geography (Lui, 2015), Religion (Debating Europe, 2015), or History (Lui, 2015). In Lui’s (2015) article, Dunford, for example, asserts, ‘give the boot to art/woodwork/cooking and lessen the time for sport’ (para 1), though he supports retaining Geography and History. Elsewhere, languages would be removed (Hait, 2017).
The No case equally acknowledges the significance of choice; however, its reference is elaborated in terms of the child’s right to make their own choice about subjects beyond the ‘basics’, in which coding is to be included. Capone, for example, asserts regarding mandatory coding, ‘no, this isn’t the right way’ (Debating Europe, 2015, para 1), arguing that elementary schools should provide the basics, and leave more specific areas to specific choices later on (para 1). This argument, highlighting student age, holds that younger students are disadvantaged through deprivation of the option to choose coding later in life, when more capable of making informed choices. Others, however, assert the right of parents to make choices about their child’s school subjects (Debate.org, 2013).
Argument deploying references to ‘choice’ also highlights likely outcome of compulsion—coding enjoyed by few and hated by many (Guzdial, 2014). Mainstreaming coding to generate equality may result in unequal distribution of the enjoyment of coding. Dkramer3 adds, ‘if you want kids to hate programming … make it mandatory’ (Haaramo, 2015, para 19). This argument represents not only the remainder of the ‘mainstream’ as potentially disenfranchised, but also students of lower academic ability. This too is a concern with disadvantage. Dillon asks: ‘do we know how to teach CS to students with learning or development disabilities? Can we confidently state that, without CS, those students should not earn a high school degree?’ (Guzdial, 2014, para 7). Consequently, the No case represents compulsory coding as yet another opportunity for students to fail, creating a two-tier subject performance cohort (Computing Education Research Blog, 2014). What is more, the No case argues that the level playing field will be distorted by the power of economically advantaged families to purchase support resources beyond those provided by the state (Editorial Board, 2017).
The No case maintains that a commitment to coding should arise naturally through individual passion (Chris-Granger.com, 2015) and intelligence (The Register 2012). While acknowledging one is not a Luddite if one cannot code (Chris-Granger.com), many in the No case assert that coding is neither easy nor quickly learned, such that to assert otherwise devalues the profession and intelligence of those who can code (Byrne, 2013). Hence, the No case holds that anti-elitism disadvantages existing coders. As Hartnell explains:
I hate the current railing against ‘elitism’ from the know-nots. They would be the first to complain if their football team didn’t field elite players, or their medical treatments weren’t by highly-skilled professionals, but they also feel threatened by the intelligentsia, so try to drag everyone down to their level (Young, 2012, para 100).
Habits of mind discourse
Whereas data show a ubiquity discourse constructs compulsory coding as enabling individuals to engage digital technologies at a generic level, the ‘habits of mind’ discourse deployed by the Yes case constructs engagement as both cognitive and cerebral. By contrast, the No case reasserts a need for engagement with the analogue. Indicative of the Yes case, Vikberg argues that programming needs to be integrated across the curriculum ‘to drive critical and creative thinking as computer programmes have an increasingly significant part to play in the society today’ (Haaramo, 2015, para 3). Others assert, ‘we need to remember that computer science is a creative field in which students are actively creating something’ (Team ISTE, 2015, para 9). Hai Hong from Google’s K-12 outreach programs maintains that ‘it offers problem-solving skills and promotes creativity’ (Della Cava, 2015, para 7). Elsewhere, a School Education Gateway news report informs that ‘learning to code also develops skills such as problem solving, [and] logical reasoning and creativity’ (2015, para 3). And beyond creativity (ECDL, 2015), a post at Familytech (2017) highlights coding’s capacity to improve ‘essential life skills like critical thinking, problem-solving, and creativity; which can lead to success in other areas of work’. Creativity, critical thinking and reasoning figure strongly in this (Yes case) discourse.
This representation constructs coding’s ‘spill-over effect’. Coding is good not only because it helps create programs, but also because its skillset can be used elsewhere. Though precise details of where else are often absent, the comment of Garun, reproduced by others such as Team ISTE (2015) and debate.org (2013), represents this argument.
Learning to code contains the same logic skills you apply in daily life: What is the problem? How can I solve the problem as efficiently as possible? Can my solution be helpful to others who are experiencing similar issues? If you can figure out the same steps from a programming perspective, it can help develop your logic and decision-making skills to streamline the best solution to your problems (2013, para 14).
Like other discourses, this too appeals to authority, in this instance, to well-known software developers, ‘research’ and education policy. Concerning the former, Bill Gates (Microsoft) and Mark Zukerberg (Facebook) are cited in terms of ‘code’ and creativity (Burks, 2017, para 5). Steve Job’s comments about programming and thinking are also reproduced by the Yes case (Puiu, 2017). Appealing to the ‘authority’ of research, we are told that ‘numerous formal studies have confirmed coding has measurable beneficial effects on cognitive abilities’ (Rawlins, 2015, para 18). Not only are state curriculum and policy documents referenced, so too are texts detailing educational principles and philosophy (Computing Education Research Blog, 2014).
The ‘habits of mind’ discourse is, however, also strongly grounded by the Yes case in the notion of computational thinking. As the president of the Australian Computer Society (ACS), Ansley argued that computational thinking should be taken ‘more seriously’ in schools, maintaining ‘it’s the fourth ‘r’ … three ‘r’s’ plus coding, or computational thinking’ (Calixto, 2015, para 19). Elsewhere, a Queensland Government (2015) paper identified coding as a twenty-first century literacy (p. 5).
This computational thinking framework is extended by the Yes case to include computational participation, emphasising networks, collaboration and the social (Kafai & Burke, 2014; Queensland Government, 2015). An EU text, for example, likewise asserts that, ‘each and every interaction between humans and computers is governed by code’, such that in our ‘hyper-connected world … coding is the literacy of today’ (European Commission, 2016, para 2). Others, however, reference the human brain, with Lopes arguing that, ‘it’s a matter of development of the human brain, to increase logical skills and prepare for the future of cyberworld’ (Sololearn, 2017), a (biological) framing reproduced in discussion at Broadband Solutions.
Like learning a language early in life, learning and practicing this type of [programming] thinking early in development actually influences a child’s brain as it is still developing (Broadband Solutions (2016), para 7).
By contrast, though deployment of the habits of mind discourse by the No case is not so focused on the cerebral, it engages similar discursive tactics. It appeals to authority, speaks (back) to the ‘spill-over’ effects and engages the impact of computational thinking in terms of creativity and critical thinking.
Appealing to the authority, the No case references Jeff Atwood, co-founder of Stack Overflow and Stack Exchange. Saines (2014) explains, for example, that ‘those defending the need to teach young children to program don’t have a solid counter-argument when luminaries like Jeff Atwood say that not everyone should learn to program’ (para 4). Indeed, Atwood (2012) himself maintains that ‘the whole everyone should learn programming meme has gotten so out of control that the mayor of New York City actually vowed to learn to code in 2012’ (para 1). Other references delegitimise ‘authority’, through defining the authority to speak about education. Commenting on the views of the Apple CEO, Pewen and Shotton represent Tim Cook (and Bill Gates) as unqualified to speak about education. Pewen asserts that beyond Cook ‘making [ing] the most idiotic statements on a regular enough basis’, he is gravelled by Cook and Gates ‘shooting off about education it’s not your field’ (Hall, 2017, para 16).
In response to claimed spill-over effects, Sandy replies, ‘try teaching a few dozen computer science classes and see if you still think CS is better prep for life skills than *any* (sic) other subject. It’s not’ (Rawlins, 2015, para 21). Sandy challenges the veracity of the ‘research’ cited as evidence in terms of creative thinking, asserting, ‘it is a myth. You can say the same for creative writing, but neither should [it] be compulsory’ (2015, para 19). This is an interesting response, not because all subjects might claim spill-over effects, but because it interrogates the positive link between computational thinking, creativity and critical thinking constructed by the Yes case.
By contrast, the No case foregrounds a more humanities-based understanding of creativity and critical thinking. At one level, this emerges in referencing the significance of humanities-based knowledge to the digital. Software designer, WiscoNative, explains that his ‘useless’ sociology degree taught him much about social interaction, meaning that he could ‘design better software that users can find more intuitive’ (Hall, 2017, para 16).This comment contests the primacy of quantitative thinking to creativity by constructing a link with humanities in this domain. Indeed, in claiming we have gone too far with the STEM agenda, Dumdum argues, ‘let’s not forget that many of our values and cultural developments, from democracy to the Enlightenment, actually came from times and places where the Liberal Arts ruled’ (Hall, 2017, para 14). In contrast to the networked minds of the Yes case, the No case maintains that children should first be taught to play outside and read books (Debating Europe, 2015). What is more, the No case argues that children should be ‘developing human skills … running around, throwing a ball, scratching out drawings, learning fine motor skills and developing normal interpersonal skills’ (Dvorak, 2014, para 2).