Phenomenology of software. The materiality of digital platforms from the perspective of the visually impaired

Hele tekst

(1)University of Amsterdam Research Master Media Studies Academic year: 2015-2016 24 June 2016. Phenomenology of software The materiality of digital platforms from the perspective of the visually impaired. Anna Piccoli Student number: 10849025 Supervisor: Dr. Thomas Poell Second reader: Dr. Anne Helmond.

(2)

(3) Acknowledgements There are several people without whom the completion of this research would have not been possible. First and foremost, I am indebted to all the participants for the time and information they shared with me. I am deeply grateful because they opened my eyes to many issues that go well beyond this thesis and showed me their perseverance and positivity. Their help has been invaluable. Many thanks to Irene Matassoni and the staff of Irifor del Trentino, who supported me right at the beginning of this project, and to the Disability Office of the Opera Universitaria di Trento for spreading the word among students with visual impairment. I also want to show my appreciation for Carlo Piccoli, Patrizia Cordin, Lorenzo Piccoli, Matteo Carli, and Ettore Paris, who mobilised their networks to provide me with other useful contacts. Many thanks also to Paul de Nooij and Marten van Doorn from Bartiméus, Alice Schippers from Disability Studies in Nederland, Patricia Overheul, policy-supporting employee at the Ministry of the Interior and Kingdom Relations, Saskia Andriessen from Andriessen Arbeidsparticipatie (Aapar), and Emily Oosterom: without them I would not have gotten to know my respondents in the Netherlands nor would I have been able to carry out the comparative study as I wanted to. A special thanks to Thomas Poell, whose advice and steadfast guidance was fundamental all the time. Last but not least, I extend my gratitude to Jonathan Teoh, for helping me put things into perspective with his stories and a whistle as well as for taking the time to carefully proofread my work and comment on it; to Jeroen de Vos, for his honest feedback and for being my Dutch proof-reader when needed; to Felix Clasbrummel, for his attentive listening, acute comments, and constant support; and to my family for their encouragement throughout..

(4) Phenomenology of software The materiality of digital platforms from the perspective of the visually impaired Abstract How do visually impaired people access and use digital platforms? How is their relationship with platforms shaped by assistive technologies (i.e. screen readers, braille displays, magnifiers)? And ZKDW LV WKHWKHRUHWLFDO JDLQ RILQYHVWLJDWLQJWKH H[SHULHQFHRI³H[WUD-RUGLQDU\´XVHUV"7KHVHLVVXHV are socially relevant, as well as relevant for our understanding of the technological mediation of platforms, yet often neglected both in the academic and in the public debate. This thesis combines insights from new media studies, disability studies, and software engineering, with a qualitative methodology to provide answers to these questions. On the basis of data collected through eighteen semi-structured interviews with Italian and Dutch users with low or no residual vision, it is argued that the perspective of visually impaired users on digital platforms can be taken up to more clearly illuminate the materiality of software. It is claimed that assistive devices introduce an additional mediation that, on the one hand, is the sine qua non for worldmaking processes and consequent navigation; on the other hand, it presents structural information which is often invisible to sighted users. In other words, the passage from sight to audio has the potential to highlight software components that lurk behind friendly, highly visual interfaces. In particular, habit disruption and user frustrations are regarded as crucial occasions upon which the materiality of software opens up again. But how do visually impaired users deal with the conspicuousness of software? It is claimed that habit ± together with software development ± plays an important role in helping users restore seamlessness and take part in real-time data flows. The frequent passage from a condition of readiness-to-hand (Zuhandenheit) or familiarity with the technology to a situation of fragmentation and unreadiness-to-hand, and vice versa, is regarded as a valuable research object for those who aim to gain access to the materiality of software. Not only it enhances technical components, the inescapability of physical and temporal limits, and practical instantiations of software. It also problematises the way users experience the digital material and naturalise it. Ultimately, this thesis points to the relevance of engaging phenomenologically with perspectives other than the one of the ³RUGLQDU\´XVHU, for scholarly inquiry and for possible social and cultural advancement. Keywords digital platforms; visual impairment; materiality; accessibility; seamlessness; conspicuousness.

(5) Table of contents Introduction««««««««««««««««««««««««««««««««««. p. 4. Chapter 1. Problematising the literature: towards a phenomenology of software«««««. p. 10. 'LJLWDO³ZKLWHFDQHV´DQGSODWIRUPVGLVDELOLW\VWXGLHVDQGVRIWZDUHHQJLQHHULQJ««. p. 10. 1.2 Invisible materiality, corporeal interactions: new media and software studies................. p. 17. 7RZDUGVDSKHQRPHQRORJ\RIVRIWZDUHUHVHDUFKTXHVWLRQV««««««««««««. p. 18. Chapter 2. Taking a new perspective: methodological considerations and corpus«««««. p. 21. Chapter 3. Theorising audio and vision in digital platforms: on affordances and worldmaking. p. 31. Chapter 4. Encountering materiality through disruption: on structural inputs and frustrations. p. 43. Chapter 5. Balancing between seamlessness and conspicuousness: on habits and temporalities. p. 50. Conclusions«««««««««««««««««««««««««««««««««««. p. 59. List of references«««««««««««««««««««««««««««««««««. p. 62. Appendixes«««««««««««««««««««««««««««««««««««. p. 72. Appendix A.1 ± Interview guide in English««««««««««««««««««««.... p. 72. Appendix A.2 ± Interview guide in Italian«««««««««««««««««««««. p. 74. Appendix B ± Legend of transcription symbols««««««««««««««««««. p. 76. Appendix C ± Transcriptions««««««««««««««««««««««««««. p. 77.

(6) Introduction ³9RLFH2YHU8WLOLW\ZLQGRZ9RLFH2YHU8WLOLW\8WLOLW\&DWHJRULHVWDEOHURZVHOHFWHGFROXPQ RI

(7) General. You are currently on a table. To enter this table, press Control-Option-Shift-'RZQ$UURZ´ This is my first contact with VoiceOver, the screen reader embedded in all Apple devices, meant to support visually impaired people in their use of computers and tablets. I execute the instruction and access the VoiceOver tutorial. Everything is mechanically read aloud in quite a monotonous tone at a rapid pace. If I could not re-read what is written on the screen, I would have already forgotten which keys to press. And this is only the first slide of the tutorial. I try to concentrate and follow without looking at the screen. Although at the beginning it is fun to explore my personal computer in a different way, I realise how quickly I get overwhelmed and frustrated. I keep pressing the wrong combination of keys or I miss the right information and have to start anew. Besides the effort of having to learn how to use my keyboard to achieve what I want with the help of the screen reader, I need time to get used to several factors: the speed of the voice; the pronunciation; the way the commands are told; the fact that I am informed about structural aspects of the webpages or documents I am working on which I do not normally pay attention to and which go beyond the mere content; and last but not least, the language: if there are non-English words, they become almost unrecognisable ± let alone terms written in an alphabet other than the Latin one: these are not read at all. Given that most of my contacts on social platforms are not English and many queries I submit online are multilingual, making sense of what it is being read by the English-speaking screen reader is a challenge. Actually, it is not only hard for me to use the screen reader: at some point, it also starts being quite boring. Since, in principle, any change on the screen has to be signalled, the default voice, Alex, keeps talking, letting me know LI,DP³FXUUHQWO\RQDWH[WDUHDLQVLGHDJUXSSR>VLF@´RULI, KDYHMXVWRSHQHG³&KURPHZLQGRZ1HZ7DE1HZ7DEVHOHFWHG7DERI´DQG,DPQRZ³RQDWH[W ILHOGLQVLGHRIDQ+70/FRQWHQW´$V,W\SHLQVRPHWKLQJ$OH[ RUDQ\RWKHUVHOHFWHGYRLFH

(8) UHpeats it letter by letter, space by space. If I delete something, I am not told it. Instead, the erased letter is repeated. Understanding if I am misspelling words becomes quite complicated 1. Since I can actually see, I can quickly check a text or scan through web results on a search engine in order to access the page I am looking for. If this was not an option, I would lose plenty of time waiting for Alex to read aloud what is there. If I switch back and forth between documents, I have to listen to Alex repeating 1. A braille display can be the best solution for this kind of situations, provided the user can read braille, which is not the case for me.. 4.

(9) where I am each time, which can be very tiresome if the document titles are long. Indeed, the screen reader will first tell me what program or folder I have opened. It will then mention the name of the object that is now visible on the screen, say what action I can do and how I can carry it out. I could EHWROGIRULQVWDQFH³6\VWHP3UHIHUHQFHVZLQGRZ6\VWHPSUHIHUHQFHVWRROEDU<RXDUHFXUUHQWO\RQ a toolbar. To interact with the items on this toolbar, press Control-Option-Shift-'RZQ$UURZ´7KLV repetitive process becomes boring after a while, even if it is necessary when you have no other way to control what happens on your screen. Most inconveniences may derive from the fact that I am a beginner, but even advanced screen reader users have been found to be slower on average than their sighted peers (Andronico, Buzzi, Castillo & Leporini, 2006; Ivory, Yu & Gronemyer, 2004). Crucially, though, the frustration and boredom I face make me notice the absence of a key characteristic that normally defines my interactions on digital platforms: seamlessness. Claiming that platforms have become imperceptible is not groundbreaking in new media studies. Quite the FRQWUDU\ D ³P\WK RI WKH LPPDWHULDO´ van den Boomen, Lammes, Lehmann, Raessens, & Schäfer, 2009, p. 10) has gained DIRRWKROGERWKDPRQJHQGXVHUVDQGLQPHGLDUHVHDUFK<HW³WKHUHDOLVDWLRQ that we are merely dealing with a myth is crucial, since it also conceals the methodological challenge that the material aspects of digital media are not only increasingly invisible to users, but may also EHFRPHPRUHGLIILFXOWWRDFFHVVIRUUHVHDUFKHUV´ Reichert, Richterich, Pablo, Fuchs, & Wenz, 2015, p. 5). Starting from this observation, new materialism, software studies and forensic materialism have engaged with the hidden PDWHULDO RI WKH GLJLWDO EH LW DV KDUGZDUHPDWWHU ³SUDFWLFDO LQVWDQWLDWLRQ´ RU ³VLJQLILFDQFH´ /HRQDUGL Q.p.). In this study I present the opposite pole: a case of user-platform relationship where software is not transparent, but mostly conspicuous, and I suggest that a phenomenological approach which takes up the specific experiential dimension of visually impaired users can prove particularly effective in providing evidence for what has otherwise become invisible to the eye: the materiality of software and its limits. Imagine, for instance, opening a webpage where the time is updated every second. A nice effect, not too mesmerising for a sighted person. (S)he can always glance at the time and for the remainder forget about the script that updates it. This is not the case if you happen to use assistive WHFKQRORJLHV EHFDXVH RI SRRU VLJKW ,QGHHG D VFUHHQ UHDGHU ³ZRXOG RQO\ UHSHDWHGO\ VSHDN WKH XSGDWHGWLPHUDWKHUWKDQUHQGHULQJWKHSDJHFRQWHQW´ Sloan, Gregor, Rowan & Booth, 2000, p. 99). Every second you would know that a second has passed, but you would not be able to go any further than that in the page. What is there besides the clock? You cannot know. The chances are high that you will simply leave the page and opt for an alternative, provided you can find it. This is indeed what occurred to one of the evaluators engaged in the assessment of the accessibility and usability of 5.

(10) web resources in Sloan et al.'s (2000) research. Even a minor detail can make a page totally inaccessible for a visually impaired person, while at the same time letting materiality shine through. Nowadays accessing web resources and digital platforms ± in real time, I should add ± is an action that is often taken for granted. However, despite the substantial development of software, differently abled users2 still encounter several problems when using digital devices and, especially, when connecting online. Their main difficulty is not necessarily computer error, but user frustration due to the inability to carry out desired actions because of improper layouts (Lazar, Allen, Kleinman & Malarkey, 2007). Yet disability is not often accounted for within statistics about the digital divide. 7KH H[SUHVVLRQ UDWKHU VXJJHVWV WKDW ³LQGLYLGXDOV LQ FHUWDLQ GHPRJUDSKLF Jroups, such as racial minorities, rural communities, and individuals of lower socioeconomic status, are at a disadvantage GXHWRXQHTXDODFFHVVWRWKH,QWHUQHW´ (DVWLQ&LFFKLULOOR 0DEU\S

(11) 6WLOOXVHUVZLWK disabilities may also be unable to benefit from the same allegedly seamless access to computers, websites and digital platforms that so-FDOOHG ³RUGLQDU\´ XVHUV FDQ H[SHULHQFH 0any visually impaired individuals need assistive devices to use digital technologies, such as screen magnifiers, screen-UHDGLQJ SURJUDPV DQGRU EUDLOOH GLVSOD\V ³(YHQ ZLWK VXFK DFFRPPRGDWLRQV PXFK RI WKH Web is still not accessible to people with disabilities, forcing organizations for these individuals to FRPH XS ZLWK WKHLU RZQ DFFRPPRGDWLRQV « ,I WKHVH HIIRUWV are unsuccessful, individuals with GLVDELOLWLHVVLPSO\FDQQRWDFFHVVFHUWDLQUHVRXUFHV´ 'REUDQVN\ +DUJLWWDLS

(12) $OWKRXJK existing legislation should serve to limit the cases of inaccessible websites, access and usability remain problematic3. Nonetheless, as I have mentioned, there is another side to this coin, which is also the focal point of my research: cases like those described in the previous paragraphs suggest that, if we take WKHSHUVSHFWLYHRI ³H[WUD-RUGLQDU\´XVHUV ZHPD\ EHDEOH to powerfully expose the materiality4 of online platforms. In particular, my hypothesis is that, thanks to their different use of senses and due 2. 3. 4. ,QWKLVZRUN,ZLOOVRPHWLPHVXVHWKHFRPPRQH[SUHVVLRQ³GLIIHUHQWO\DEOHGXVHUVÍRUODFNRIDOWHUQDWLYHV+RZHYHU from the start I would like to distance myself from this terminology, which, in my opinion, often produces false assumptions regarding the skills of people with visual impairment. The European Commission's Study on Assessing and Promoting e-Accessibility showed that national policies on web accessibility had improved over time (Kubitschke et al., 2013

(13) EXW LW DOVR SURYHG WKDW ³WKHUH LV VWLOO D VLJQLILFDQW deficit on web accessibility implementation across Member SWDWHV´ (XURSHDQ'LVDELOLW\)RUXPS

(14) N.B. This thesis does not analyse any legal aspects. However, legal issues definitely deserve major attention. Indeed, it is important to be aware of the relevant role legislation plays in framing ± in the sense of both determining and limiting ± accessibility, technical applications, and media adaptations for differently abled users both on the national level and on the supra-national one. It should not be forgotten that web developments are actually questioning H[LVWLQJ OHJLVODWLYH IUDPHZRUNV HYHQ EH\RQG WKH UHDOP RI WKH ³H[WUD-RUGLQDU\´ XVHU

(15) SRVLQJ QHZ FKDOOHQJHV WKDW require constant adjustments and just as new solutions. In my text, I will alternatively use the terms materiality and physicality, althRXJKDV%URZQ

(16) QRWHV³materiality KDVDVSHFLILFLW\WKDWGLIIHUHQWLDWHVLWIURPLWVVXSHUILFLDOFRJQDWHV«:KHQ\RXDGPLUHWKHPDWHULDOLW\RIDVZHDWHU you're acknowledging something about its look and feel, not simply its existence as a physical REMHFW´

(17) . 6.

(18) to the additional mediation introduced by assistive technologies, visually impaired users are indeed in a unique position to let us see more clearly the hidden side of software and platforms we use on a daily basis. Through a phenomenological inquiry which examines the interactions between visually impaired users and digital platforms I show that, by experiencing bugs and by being forced to reconstruct the logics of webpages in ways that mostly differ from those of sighted users, the visually impaired are often confronted with the limits and the physicality of the technology even in an age of alleged seamlessness (cf. Blind Inspirationcast, 2013). As a consequence, they can highlight elements that are frequently (and paradoxically) invisible for sighted users (see Accessible Media Tube, 2014). 7KH ³RUGLQDU\´ XVHU IRU LQVWDQFH ZLOO QRW SD\ WRR PXFK DWWHQWLRQ WR WKH clock updating the time every second, whilst a blind user cannot help but be aware of the existence of such a script. There is more to platform accessibility and digital materiality than meets the eye. Becoming attentive to similar aspects could be a valuable method for scholars to pinpoint the elements that (il)logically structure webpages, platforms and documents as well as the human input behind those structures and its practical instantiations. Yet, despite the considerable number of visually impaired users5 DQGLQVSLWHRIWKHFHQWUDOLW\RI³WKHXVHUÍRUFXUUHQWVWUDQGVRIUHVHDUFKLQ new media studies, scholars in this field have thus far paid relatively little attention to the relationship between visually impaired users and digital platforms. Instead, most academic sources on the topic are from the field of disability studies or computer sciences. Many researchers devote their attention to test the actual usability of websites and evaluate the differences in navigation EHWZHHQ ³RUGLQDU\´ DQG ³H[WUD-RUGLQDU\´ XVHUV VHH DPRQJ RWKHUV 0F*RRNLQ %UHZVWHU -LDQJ 2008; Andronico et al., 2006; Buzzi, Buzzi, Leporini & Akhter, 2010; Buzzi, Buzzi & Leporini, 2011; Ashraf & Raza, 2013; Csapó, Wersényi, Nagy & Stockman, 2015). These studies mainly concentrate on equality in terms of website accessibility or analyse the functionality of assistive technologies, with the aim of developing alternative prototypes that might guarantee access and usability to the blind without subtracting the pleasure of online interaction for sighted users (Edwards, 1989; Façanha, Araújo, Viana & Pequeno, 2012; Loo, Lu & Bloor, 2003). In media studies, computer sciences and disability studies, though, the subtleties of the ³H[WUD-RUGLQDU\´ XVHU V H[SHULHQFH DUH UDUHO\ WKHRULVHG. Existing findings on materiality within VRIWZDUH DQG SODWIRUP VWXGLHV WHQG WR FRQFHQWUDWH RQ ³RUGLQDU\´ XVHUV DQG H[HPSODU\ FDVH VWXGLHV but largely fail to address the implications of the presence or absence of specific modalities 5. Despite the lack of reliable statistics concerning the number of non-sighted users (both globally and per country), the engagement of the visually impaired with digital platforms seems worthy of more research (see Wu & Adamic, 2014). WHO (2014) estimates 285 million visually impaired people worldwide, of whom 39 million are blind. 82% of them are currently older than 50 years.. 7.

(19) (visuality, audio, tactility) and the mediating effect of certain technologies (e.g. screen readers). This phenomenology of software aims to bring back the importance of sensory experiences in digital media looking at a specific minority of users. Rather than starting from the observation that software is transparent but nevertheless material, I examine a condition where platform corporeality is evident, but often, as it were, normalised. In this way I present another method to deal with greater clarity not only with questions of materiality that have been discussed before, but also with the perceptions of users concerning materiality. Although incomplete, previous discussions represent a good starting point for further analysis. It is indeed from them that we can draw the idea that VRIWZDUH LV QRW LPPDWHULDO EXW ³LQ-PDWHULDO´ van den Boomen et al., 2009). Building upon that, I look at how interactions meant to be mainly visual happen without vision, and what the implications of channelling visual elements into audio clues are. To do this, I explore how the relationship between visually impaired users and digital platforms is technologically mediated and how this affects web aesthetics and affordances. In my inquiry, I largely focus on the impact of screen readers and try to show how they can make software conspicuous again. My first chapter presents in further detail the literature regarding visually impaired users of platforms, referring both to disabilities studies and computer science researches and new media studies. In the second chapter, I present the methodology I used to collect data regarding how the access and use of digital platforms by the visually impaired is technologically mediated and how the relationship user-platform is experienced. In particular, eighteen semi-structured interviews provided insight into these aspects. Participants were men and women of ages ranging from 19 to 69 years, all of whom are blind or have low vision. They were recruited using convenience sampling and snowball sampling. The interviews were conducted in Italy and in the Netherlands in order to lay the foundation for a comparison regarding the impact of language on accessibility and usability. Considering that the majority of the Italian population tends to access content in Italian only, while the Dutch are more used to switch between their native language and English (see European Commission, 2012), the study inquired whether the interaction changes depending on the settings of the assistive device, the language spoken by the user and the language of the platform content. In the following three chapters I outline my findings. I do not discuss how platforms and accessibility software can be technically improved, as much of the research in disability studies does. Rather, I describe how accessibility software packages affect the use and perception of platforms, exploring the actual practices of people who use assistive devices to access digital platforms. Considering that the visual modality is translated to tactile and, more frequently, audio inputs, I describe what happens in the passage from one modality to the other by harking back to the fields of 8.

(20) PXOWLPRGDOLW\ DQG WUDQVODWRORJ\ %XLOGLQJ RQ WKH QRWLRQV RI ³K\SHUWH[W´ +D\OHV

(21) DQG ³ZRUOGPDNLQJ´ *RRGPDQ

(22) Chapter 3 analyses ways in which the visually impaired deploy web affordances and offers a short comparison with sighted users. The fourth chapter concentrates on habit disruption and user frustration as entry points to software materiality. This leads to a theoretical engagement with previous work on web art and web navigation, which also regards bugs, glitches, and errors as moments when the materiality of software opens up. The last chapter largely refers to Berry's (2011) Heideggerian analysis of software and shifts the focus to the visually impaired SHRSOH¶V perception of digital platforms. It does so in conversation with Garfinkel's (1986) ethnomethodology and Silverstone's (1994) domestication theory, arguing that visually impaired users are able, through habit and constant practice and thanks to software development, to conceive platforms as seamless and take part in real-time data streams ± at least until the next frustration. Once again, the precariousness of their condition helps me theorise the conspicuousness of software and the inescapability of both physical and temporal limits. My research thus helps not only to shed light on the accessibility of digital platforms for people with disabilities (i.e. visual impairment), but it also contributes to the critical discussion concerning the materiality of platforms and the way this is experienced from this specific perspective. But without much further ado, let us begin with the conceptual problematisation.. 9.

(23) Chapter 1. Problematising the literature: towards a phenomenology of software 1.1 DLJLWDO³ZKLWHFDQHV´DQGSODWIRUPVGLVDELOLW\VWXGLHVDQGVRIWZDUHHQJLQHHULQJ Research that specifically addresses visual impairment tends to neglect the material construction of platforms and its role in the interplay with non-sighted users. The majority of studies which investigate the ways people with low or no vision use computer resources and interact online are carried out either by disabilities scholars or within the fields of computer science, industrial research, application design, and software engineering. Whereas the latter are concerned with accessibility evaluation, usability testing and design improvement6, the former reflect on issues related to equality and identity. Both strands, though, highlight the fact that, despite the rhetoric of openness and LQFOXVLYHQHVVUHODWHGWR³SODWIRUPV´DQGWKHLQWHUQHWLQJHQHUDO FI*LOOHVSLH

(24) DODUJHPDMRULW\ of websites are still to some degree inaccessible or unusable, insofar as they present features which ³SUHYHQW WKH DVVLVWLYH WHFKQRORJLHV IURP FRUUHFWO\ LQWHUSUHWLQJ DOO RI WKH DYDLODEOH LQIRUPDWLRQ´ (Sloan et al., 2000, p. 96). Automatic accessibility evaluation together with user testing are frequently adopted to reveal the flaws of websites and platforms. In particular, the widespread use of detailed graphics and multimedia elements (e.g. animations) on the Web creates several problems. Possible alternative designs based on audio inputs/outputs or on tactile feedback (see Buzzi et al., 2011; Edwards, 1989) as well as on more logical page structuring (Andronico et al., 2006; Ashraf & Raza, 2013) are proposed, but seem to fail on the market. Part of the reason why much effort is devoted to evaluating accessibility and developing usable solutions is the fact that digital technologies have in principle an inclusionary potential for ³KDQGLFDSSHG´XVHUV³$YLVXDOO\LPSDLUHGSHUVRQLVDEOHWRDFFHVVGLJLWDOWH[WVZKHUHDVWUDGLWLRQDO analogue, books remain inaccessible. The Web also allows people with disability to interact with RWKHUV LQ D ZD\ WKDW LV QRW XVXDOO\ SRVVLEOH LQ JHQHUDO VRFLHW\ « SHRSOH ZLWK GLVDELOLW\ QHHG QRW LGHQWLI\ DV GLVDEOHG LQ RQOLQH VSDFHV DQG FDQ LQVWHDG EH MXGJHG RQ WKHLU SHUVRQDOLW\ ILUVW´ (OOLV Kent, 2008, n.p.). At the same time, however, the visibility of disability in society tends to decrease due to digital anonymity. Moreover, identity issues might come into play when interactions are clearly enabled by DVVLVWLYH WHFKQRORJLHV VLQFH ³GLVDELOLW\ LV FRPPRQO\ SHUFHLYHG WR V\PEROLVH. 6. Generally with the help of at least one blind co-author and/or with the involvement of visually impaired subjects in the assessment of technologies and websites.. 10.

(25) tragedy, invalidit\DQGIUDLOW\´ 6|GHUVWU|P <WWHUKXVS

(26) ± LWLVFRPSDUHGWR³DWRPEOLNHLPSULVRQPHQW´ .OHHJHS

(27) $VVLVWLYHGHYLFHVWHQGWRGUDZDWWHQWLRQWR ³UHVWULFWLRQ difference and dependency, characteristics that were associated with a medical understanding of GLVDELOLW\ DQG ZLWK DQ LGHQWLW\ DV GHYLDQW´ 6|GHUVWU|P <WWHUKXV S

(28) &RQVHTXHQWO\ 6|GHUVWU|PDQG<WWHUKXV

(29) IRXQGWKDW³ZKHQ \RXQJSHRSOHSHUFHLYHGDFHUWDLQ WHFKQRORJ\DV outside the range of the normal, or if the technology overtly spoke of impairment, using this WHFKQRORJ\KDGWREHHLWKHUGLVJXLVHGRUQHJDWHG´ S

(30) On the other hand, the costs of and limits to accessibility are exposed. In this respect, Dobransky and Hargittai (2006) argued that [f]or man\ SHRSOH ZLWK GLVDELOLWLHV VXFK DV « WKH EOLQG ZKLOH GHPRJUDSKLF DQG VRFLRHFRQRPLF factors play a role in lower levels of access to and use of computers and the Internet, these do not H[SODLQWKHGLVFUHSDQFLHVLQFRQQHFWLYLW\OHYHOVRQWKHLURZQ«WHFKQLFal accessibility barriers are SUREDEO\WKHFDXVHRIWKHVHJURXSV¶ODJJLQJEHKLQGWKRVHZLWKRXWGLVDELOLWLHV1RWRQO\LVDGDSWLYH technology difficult to learn and expensive, but it lags in development behind the technology to which it is supposed to enable access. (p. 329). McGookin et al.'s (2008) study observed that assistive devices are often costly because of their special design and due to the fact that the number of units sold is much lower than in the case of mainstream devices (p. 299). Nevertheless, this is not necessarily the case for screen readers, which are nowadays embedded in most so-called new media. Yet, as far as technology is concerned, it is PDGHFOHDUWKDWWKHUHLVVWLOOPXFKURRPIRULPSURYHPHQW³>Z@KLOHWKHLGHDWKDWDYLVXDOO\LPSDLred person can access books electronically, in conjunction with a screen reader, sounds like a disabilityIUHHXWRSLDWKLVLVQRWDOZD\VWKHFDVHDV GLJLWDO GRHVQRWDOZD\VPHDQ DFFHVVLEOH ´ (OOLV .HQW 2008, n.p.). This adds up to the fact that, as :KLWH)LW]SDWULFNDQG0F$OOLVWHU

(31) QRWHG³DVQHZ technological advances are made there is concern that standards to ensure accessibility are not NHHSLQJSDFH´ S

(32) 6|GHUVWU|PDQG<WWHUKXV

(33) DJUHHGWKDW³>P@RVWDVVLVWLYHWHFKQRORJ\LV reactive in design, and by the time accommodations are made mainstream technology has moved DQRWKHUVWHSIRUZDUG´ S

(34) 7. 7. Interestingly, the same was confirmed by Arianna (25, blind), one of the Italian interviewees in this study, who said: ³8QIRUWXQDWHO\DQGOXFNLO\FOHDUO\WHFKQRORJ\HYROYHVLQOHDSVDQGERXQGVEXWLWLVoften very difficult to keep up on WKH OHYHO RI DFFHVVLELOLW\ :HOO OHW V VD\ WKDW DFFHVVLELOLW\ DGDSWDWLRQ FRPHV DOZD\V D ELW ODWHU´ ³3XUWURSSR H SHU fortuna chiaramente la tecnologia evolve a passi da gigante, però spesso è molto difficile starci dietro a livello di. 11.

(35) Since webpages are normally designed by sighted people for sighted people, the right of the visually impaired to access the internet is often (unintentionally) disregarded (Loo et al., 2003). For instance, images that are not captioned or properly labelled cannot be read automatically, and need an external translation. The same goes for CAPTCHAs (Completely Automated Public Turing test to tell Computers and Humans Apart): translation software is by design unable to decipher the blurred sequence of numbers and/or letters and this represents a hurdle for blind users. Multimedia features and animations can also exclude the blind, unless additional information is available about those features and makes them describable (Sloan et al., 2000). According to Buzzi et al. (2010), social networks integrate content from different sources. $VDPDWWHURIIDFW³>W@he recent development in web technology has made it very interactive and easy for the normal users, and there are various web technologies which are used to make web pages more interactive and user friendly. Rich Internet Applications (RIA) is a term used for one of these technologies´ +XVQDin, Avdic & Langsten, 2011, p. 8). Consequently, digital platforms often have a complex layout that is only somewhat accessible through screen readers. On Facebook, for example, highly dynamic webpages created using JavaScript risk being technologically unreadable by screen readers (Wu & Adamic, 2014). Buzzi et al. (2011) noted that alternative, simplified web applications are sometimes available for differently abled users. Twitter, for example, could be used via Accessible Twitter ± now called Easy Chirp8, while in 2006 Google Labs launched Accessible Search9 to help users find the most accessible pages in the result list (Andronico et al., 2006). Some platforms, such as Facebook and Amazon, instead promote their mobile version as more accessible due to the more simplistic design (cf. Wentz & Lazar, 2011). Nonetheless, alternative applications are not necessarily widely known nor have an equalising effect: first of all, they might have limited applicability. Easy Chirp, for instance, is only available in English. Moreover, people who access a different version of the same platform can still feel, to a certain extent, discriminated: the user interface is different from the original, DQG WKH DSS FRXOG WKHUHIRUH EH UHJDUGHG DV ³D QRQ-LQFOXVLYH RSSRUWXQLW\´ Buzzi et al., 2011, p. 152; cf. also Wentz & Lazar, 2011). Furthermore, it may happen that the alternative interface is not updated as often as the original one (Wentz & Lazar, 2011). Hence, scholars do not consider multiple interfaces an optimal solution. Rather, they call for an adaptation of the main interface. In this respect, Sloan et al.

(36) DUJXHG WKDW GHVLJQ EHWWHUPHQW WDUJHWLQJ ³H[WUD-RUGLQDU\´. 8 9. DFFHVVLELOLWj,QVRPPDGLFLDPRFKHO DGDWWDPHQWRGLDFFHVVLELOLWjDUULYDVHPSUHXQSRFKLQRGRSR´

(37) See www.easychirp.com. N.B. Accessible Search is no longer supported.. 12.

(38) XVHUVZRXOGDOVREHQHILW³RUGLQDU\´XVHUVLQH[WUD-ordinary conditions (e.g. outdoors, on the move, using devices with limited capabilities)10. For instance, controlling the appearance of web pages by means of Cascading Style Sheets (CSS) would enable users with particular needs to avoid predetermined styles that can be impairing for them: they could change the colours of background and font to distinguish better the text in case of dyslexia or low vision (Sloan et al., 2000, p. 100). On the other hand, considering that blind users are more likely to avoid pages and platforms that they expect to have accessibility issues (Wentz & Lazar, 2011), complicated and heavily graphical pages could ultimately lead to a form of auto-exclusion. This is extremely problematic in terms of social inclusion (cf. White et al., 2008). In effect, GLJLWDOQHWZRUNV³KDYHEHFRPHa core element in how young people manage their identity, lifestyle DQG VRFLDO UHODWLRQV´11 (Söderström & Ytterhus, 2010, p. 310). Moreover, as Buzzi et al. (2011) noticed, an increasing number of services rely on social networks and online resources. Loo et al. (2003) highlighted for example internet banking, internet shopping, internet voting, internet telephone, internet television, online education and job searching. Façanha et al. (2012) pointed out that the impact of social networks goes well beyond entertainment: it encompasses also educational environments and political activism. Buzzi et al. (2010) claimed that social platforms such as Facebook and Twitter foster social interaction and collaboration. Therefore, they can be essential for people who are affected by mobility limitations, among them the blind. If prevented from accessing or normally using online pages and digital platforms, visually impaired users risk being cut off from relevant information sources, international communication channels, business spaces, e-learning environments and collaborative work tools (cf. White et al., 2008). Being unable to use social platforms could have repercussions also in terms of employment for blind people (Wentz & Lazar, 2011). They may be deemed less professionally competitive than their peers. Similarly, they may be discriminated against as citizens, in countries where government departments adopt digital platforms to strengthen contact with the public (Husnain et al., 2011; Wentz & Lazar, 2011). 10. 11. $ JRRG H[DPSOH RI WKH EHQHILWV ³RUGLQDU\´ XVHUV FRXOG JDLQ IURP using websites especially designed for visually impaired users was provided by Tim in 't Veld during a TED Talk: he described the case of an English retailer which decided to create a text-only version of the web-shop to enable blind people to order online³0RVWSHRSOHZKRXVHG that text-only website were frequent customers, so people who just needed to quickly click through to the product they wanted, place it in their basket and pay. Because with this text-only website ± no clutter, very good structure ± WKH\FRXOGYHU\TXLFNO\SODFHWKHLURUGHU´ 7('[7DONV

(39) $FFRUGLQJO\7LPLQ W9HOG VFRQFOXVLRQ LVWKDWWKH differently abled user is the best test user. According to a recent article by Quartz, the major social networks (Facebook, Twitter, and especially Instagram) can expect to keep seeing a user growth in the United States in the next years (Truong, 2016). Hence, having an account and keeping up with the trend seems to become increasingly important, especially for the younger generations.. 13.

(40) To access the interfaces ± no matter if they are the original or alternative ones ± visually impaired users commonly resort to assistive devices. This primarily means screen readers, and in some cases also braille displays. Several researches in computer science and software engineering concentrate on reviewing the different assistive technologies and assessing their functionality. According to Wentz and Lazar (2011), braille displays tend to be too expensive and difficult to use given the low rate of braille literacy among the blind (see also Edwards, 1989). Screen readers are therefore more common. They work by reading aloud the content visualised on a computer screen. Different typologies of screen readers exist depending on the browser used12. Examples include JAWS, a widespread commercially available software; Window-Eyes, which is diffused among Windows users; VoiceOver, the program embedded in all Apple devices; and NVDA (NonVisual Desktop Access), a free software developed in Australia but also popular in the Netherlands. Although these accessibility software packages are necessary navigate and use digital resources, it has been pointed out that for people using a screen reader, tasks tend to be more complicated since additional steps are needed. It was calculated that on average they took twice longer than sighted users (Sahib, Tombros & Stockman, 2012; Andronico et al., 2006; Ivory et al., 2004). In principle, indeed, content is explored sequentially, and it takes time to identify the structure of a page, find relevant information, skip ads, fill in possible edit fields, get rid of pop-ups and the like (cf. Buzzi et al., 2011). It is true that the ability to list headings on a web page, ordered according to alphabet, hierarchical or absolute position on the page, allows the user to quickly and easily find the information they require without KDYLQJWRVFDQWKURXJKFRQWHQWLUUHOHYDQWWRWKHLUFXUUHQWWDVN«EXWXQOHVVGDWDLVVWUXFWXUHGDQG exposed to the screen reader in an appropriate format the blind user is forced to fall back on FXPEHUVRPHDQGWHGLRXVOLQHDUVFDQQLQJ´ :KLWH et al., 2008, p. 138).. The process often causes frustration in the user who must listen to the same information each time (s)he follows a link or is confronted with a confusing layout (Lazar et al., 2007). Even a simple tweet can prove extremely time-consuming due to the presence of links and graphics. Indeed, screen readers do not simply read the text visible on screen, but also inform the listener about the type of. 12. The first models were developed in the Eighties (Loo et al., 2003). In the same years, other media were also looking into ways to render visual content through audio. In particular, audio descriptions for theatres, films and television were born at the time (Snyder, 2005).. 14.

(41) objects that (s)he is confronted with. Buzzi et al. (2011) proposed the following fragment as an example of the tiring length of the reading: ³Heading level 3 Google Accessibility Link graphic avatar for googleaccess Graphic verified 1. block quote We're watching Link @NASA Astronauts answer questions on Link @YouTube (with captions)! Link http://youtu.be/bJBuxcuftfU 2.block quote end from Link googleaccess Link graphic user timeline Link graphic direct message Link graphic tweet message Link graphic Make Favorite Link graphic Reply Link graphic Quote Tweet (old-school retweet) Link graphic Re-Tweet Link May 19, 10:24 pm | via web Heading level 3 *RRJOH$FFHVVLELOLW\´ (Fragment read by JAWS about a published tweet captured on the Accessible Twitter app in Buzzi et al., 2011, p. 154).. This also depends on the differences in the functioning of perceptual systems. Whereas the sighted can take advantage of the possibility of simultaneously observing elements that are distanced from one another, but structurally connected to one another, the blind can only proceed sequentially (Stöger, 2001). 2 &DOODJKDQ

(42) H[SODLQHGWKDW³while vision's objects are extended in space, and 15.

(43) are individuated and recognized primarily in virtue of spatial characteristics, audible individuals are extended in time, and are perceptually individuated and recognized primarily by virtue of pitch and temporal characteristics´ QS

(44) 2IFRXUVHDSURORQJHGGXUDWLRQLQ WLPHKDVDQLPSDFW RQPHPRU\ and cognition too: for the visually impaired it can be difficult to navigate also because they have to remember several pieces of information and exert additional cognitive effort before performing any actions (Buzzi et al., 2010). Notably, Ivory et al. (2004) found out that blind users would be interested in kQRZLQJWKHILOHVL]HEHIRUHH[SORULQJDFHUWDLQSDJHRQOLQHLQGHHG³>WKH\@could use the file size to estimate the effort required to explore a page, whereas sighted users could use the download time to make this estimate´ S

(45) Since time needed to perform tasks plays an important role in an age of real-time communications (cf. Berry, 2011), enabling the user to speed up navigation is a central aspect for software engineering. This is also why screen reader users can set up faster speech rates so that the reading process becomes quicker. Their higher auditory efficiency derives from attention, practice and adaptation (Edwards, 1989)13, even though it is often assumed that visually impaired people have more acute hearing due to a natural balancing of the senses. Notably, temporality is not only important for disability studies, where it is regarded as a variable that could potentially impact upon the quality of life of visually impaired users. It is also a core element for media scholars devoted to researching digital technologies and their material dimensions. As suggested by Munster (2006), together with other elements time is part of the inescapable lattice of materiality of software. It becomes obvious when taking up the experience of visually impaired users. Especially for beginners or with unknown pages, interacting online through assistive technologies requires a lot of time. This can lead to a disruption of the apparent seamlessness of the interaction, which makes software evident again. However, researches on visual impairment and digital platforms do not extend their findings to address the question of software materiality. Rather, they focus on functionality and efficacy of technologies (both the media and the assistive devices used to access them), discussing possible improvements in design, often in relation to the cultural construction of disability and problems of discrimination and exclusion. Turning to platform materiality, thus, means turning to a different branch of studies, i.e. new media.. 13. Daston (2008) clearly explains the impact of exercise and repetition on our observing capacity, claiming that only the ³WUDLQHGH\H´LVFDSDEOHRIVHHLQJFHUWDLQREMHFWVWKDWDUHLQVWHDGLQYLVLEOHWRWKHXQWUDLQHGH\H³The novice sees only blurs and blobs under the microscope; experience and training are required to make sense of this visual chaos, in order to be able to see things´ S

(46) +HUDUJXPHQWFDQEHH[WHQGHGWRRWKHUVHQVHVWRRDV(GZDUGV REVHUYDWLRQRQ hearing proves.. 16.

(47) 1.2 Invisible materiality, corporeal interactions: new media and software studies A phenomenological inquiry of digital platforms and the way visually impaired users relate to them can draw vital inspiration from theoretical traditions in new media and software studies. Debates FRQFHUQLQJVRIWZDUHPDWHULDOLW\KDYHEHFRPHFHQWUDOWRWKHOLWHUDWXUHEHFDXVH³WKHSK\VLFDOLW\RIWKH equipment is no longer familiar to us, it no longer shines. Rather, it acts as a carrier within which software is located and which as a plastic and black-boxed technology is both radically mutable and IUXVWUDWLQJO\ IL[HG LQ IRUP DQG IXQFWLRQ´ %HUU\ S

(48) 7KH UKHWRULF DURXQG SODWIRUPV suggests clarity, but effectively obfuscates: as the user, one is confronted with friendly interfaces, but not with what lurks behind them. What one sees, clicks, and manipulates makes one forget what is invisible (Chun, 2010). Although software does not have a straightforward physicality, an increasing number of scholars have started pointing to the importance of considering new media and related SUDFWLFHVDV³LQ-PDWHULDO´ Yan den Boomen et al., 2009). It has been claimed that one cannot ignore the dynamic enmeshment of technical features, social uses and discourses. The invisibility and intangibility of software and the wireless network infrastructures that enable people to use new media and participate online do not mean that such infrastructures are not thoroughly physical (Dourish & Bell, 2011). Studying platform materiality means more than investigating the physical infrastructure. It also concerns issues of embodiment and physical interactions between affective bodies and machines. As Brown (2010) explained and as the essays collected in Fuller's (2008) Software Studies collection demonstrate, software can be considered physical at many levels: its functioning relies on tangible components (hardware); physical are also its effects (in terms of programmability, operational constraints, use, compatibility, sharing, distribution, possible bugs), which very concretely extend to the societal realm, its technical infrastructures, and its patterns of work and FRPPXQLFDWLRQ &KXQ

(49) HYHQ VSRNH RI ³SURJUDPPHG YLVLRQV´ WKDW DWWULEXWH D GHWHUPLQLQJ agency to software, which shapes our future actions: invisible components (software, algorithms) generate visible effects with implications in terms of ideology, economics, gender, race, culture. According to Dourish and Bell (2011), electronic services and computational elements can also impact on our perception and understanding of the physical environment that surrounds us and they can affect our encounters with and movements through space too; at the same time, they depend on spatial situatedness, regulations and time, and rely on other tangible infrastructures (e.g. architecture) 17.

(50) as well as coincidental conditions (e.g. weather). These are all aspects that point at the way the ³YLUWXDO´LVPDWHULDORUPDWHULDOLVHG Along a similar line of thought, the ethnographic study carried out by Miller and Slater (2000) showed that online practices and digital platforms cannot be considered apart from off-line UHDOLW\ WKH DXWKRUV DUJXHG DJDLQVW WKH QRWLRQV RI ³YLUWXDOLW\´ DQG ³F\EHUVSDFH´ ZKLFK VXJJHVW WKH H[LVWHQFH RI D ³SODFHOHVV SODFH´ 0LOOHU 6ODWer, 2000, p.

(51) ,QVWHDG ³we need to treat Internet PHGLD DV FRQWLQXRXV ZLWK DQG HPEHGGHG LQ RWKHU VRFLDO VSDFHV « WKH\ KDSSHQ ZLWKLQ PXQGDQH social structures and relations that they may transform but that they cannot escape into a selfenclosed cyberian DSDUWQHVV´ LYL, p. 5). Indeed, digital media are an integral part of daily material practices that are configured in technology. Together with hardware and social practices, the body constitutes a part of inescapable physicality of digital experiences. The act of esperire, of learning by practical trial or proof, the manipulation of interfaces through the body is central to the encounter with (new) media. According to Munster (2006), the body ± with its sensations, movements, habits, its being situated in space and time ± plays a fundamental role in digital engagements with the machine, even when one is not aware of one's own proprioception. In particular, Munster (2006) envisioned a similarity between Baroque and digital spaces that cannot exclude material HQWDQJOHPHQWV ³Both baroque and digital spaces engage the viewer visually, seductively and affectively. They operate by creating clusters of objects, images, sounds and concepts that belong WRJHWKHULQYDULDWLRQDQGLQGLVVRQDQFH´ 0XQVWHU, 2006, p. 6). ,QKHUYLHZGLJLWDO³HPERGLPHQWLV produced through the relations between the participants' bodily capacities and the operations and OLPLWDWLRQVRIWKHSDUWLFXODULQIRUPDWLRQWHFKQRORJLHV´ LYL, p. 4). 1.3 Towards a phenomenology of software: research questions %XLOGLQJ RQ WKH QRWLRQ RI ³GLJLWDO HPERGLPHQW´ DQG RQ WKH FRQFHSWXDOLVDWLRQ RI VRIWZDUH DV LQmaterial rather than immaterial, this research follows two main lines. First, it presents strategies and perceptions of a specific category of users that has been neglected thus far. Secondly, it shows how the experience of users with low or no residual visions could help to expose materiality in an age that is more and more connoted by transparency and seamlessness. It is commonly assumed that blind people have no sense of vision at all and live in complete darkness. As a consequence, they would have no understanding of visual aspects and no right to intervene in debates about visual media and 18.

(52) visual culture, being unable to grasp the powerful scope of images (see Kleege, 2005). Kleege (2005) showed the limits of this perspective by noting that it would be like stating that sighted people cannot intervene in discussions about blindness. Besides, in her opinion, it would be beneficial to scale down the value the sighted place on vision, keeping in mind that one can collect knowledge about the world by activating other sensory functions. More fundamentally, it should be observed that, when the visual component (of online platforms) is taken away, other features remain and new ones come to supplement the lack of sight. For instance, whereas the sighted follow the pointer to move across a page and execute commands, the blind still use touch (the mouse/trackpad or the keyboard for shortcuts), but need a screen reader or braille display to know what the pointer is pointing at. Assistive technologies metaphorically act as the white cane blind people use to feel the ground and understand its physical constitution. Although the prosthetic element14 may invite the reader to think of a post-human condition, it can also help to highlight the materiality of software and digital platforms in a way substantially different from current approaches within new media studies. This leads us back to consider the central questions that motivate my research, namely: how is the relationship between visually impaired users and digital platforms technologically mediated? Can a phenomenological approach that takes into account the perspective of the visually impaired provide more immediate access to platform materiality, often imperceptible to the eye? In this respect, I argue that the idea of software as immaterial can indeed be effectively challenged by examining the experiences of the visually impaired. Not only is the interaction doubly corporeal (due to the touchable hardware and the reading voice), but it is also an entry point for discovering the physical bricks that serve to build a page or a platform. In the passage from visuality to sound, the illusion of seamlessness disappears and the ³LPPDWHULDO´ VXUIDFH LV EURNHQ WR UHYHDO WKH HOHPHQWV RQH SDJH FRQVLVWV RI JUDSKLFV EODQN OLQHV slashes; headings; levels). As if playing Lego, one can (and has to) deconstruct the final product and reconstruct it to make sense of it. This can also remind users of the otherwise easily forgotten human input that is involved in the creation of the website or platform (and, we should add, in the source code) as well as of the inescapable temporal dimension of allegedly real-time navigation. The missing link (vision in this case) creates a disconnection that highlights the embodied relationship we have with new media and brings forward the materiality of the platforms explored and its practical LQVWDQWLDWLRQV,Q%URZQ V

(53) ZRUGV³WKHREIXVFDWLRQof an object can be the requisite result of 14. The prosthesis is regarded as a metaphor useful to describe the interaction between body and technology.. 19.

(54) gaining greater access to its material components ± dramatizing its materiality, let us say ± especially ZKHQWKDWDFFHVVKDVEHHQWHFKQRORJLFDOO\PHGLDWHG´ S

(55) A phenomenological approach thus usefully brings together observations regarding both the technical components of software and its socio-political implications. Its strength also lies in the resort to a qualitative methodology that facilitates a switch in the focus of analysis: whereas new media scholars normally take ordinary situations of seamlessness as their main object of study and SURFHHG WR H[SRVH PDWHULDOLW\ , LQYHVWLJDWH ³H[WUD-RUGLQDU\´ VLWXDWLRQV LQ ZKLFK SHRSOH DUH confronted with a conspicuous software and deal with fragmentation, disruption, and frustration. This, in turn, brings me to take up a third and last question, namely: how is the conspicuousness of software experienced by the visually impaired? Notably, Kleege (2005) stressed, including people who can talk about their experience DQG IRUJHWWLQJ DERXW WKH ³+\SRWKHWLFDO %OLQG 0DQ´ ± who guided thinkers in theorising about consciousness ± would be the best option for answering similar questions. Accordingly, the next section explains in more detail the methodology I used to gain insights into the phenomenon and presents the men and women who helped me avoid the pitfalls of WKHRULVLQJWKURXJKWKH³+\SRWKHWLFDO%OLQG0DQÉ\VKDULQJWKHLUSHUVRQDOH[SHULHQFHV with me.. 20.

(56) Chapter 2. Taking a new perspective: methodological considerations and corpus Following Kleege's (2005) LQYLWDWLRQWRJHWULGRIWKH³+\SRWKHWLFDO%OLQG0DQ´,FRQGXFWHGVHPLstructured interviews with flesh-and-blood visually impaired men and women to understand the ways they use and relate to digital platforms. The data collected through the interviews have fed back into the theoretical analysis. Partially sighted and blind people were considered experts in the researched field of activity, insofar as they are in the best position to talk about daily practices of access and use of digital platforms through assistive devices. Normally, visually impaired users are asked to take part in usability tests. They are required to execute a certain set of tasks and their performance is closely observed and measured to determine the main obstacles and suggest possible improvements. Through a qualitative approach, I could go beyond the technical perspective and instead explore more in detail their relationship with digital platforms. The dialogic format served to make participants' poVLWLRQVPRUHH[SOLFLW$FFRUGLQJWR)OLFN

(57) ³>D@ goal of interviews in general is to reveal existing knowledge in a way that can be expressed in the form of answers and so become DFFHVVLEOH WR LQWHUSUHWDWLRQ´ S

(58) )XUWKHUPRUH XVLQJ GLIIHUHQW TXHstions I could deal with the presuppositions that I, as a researcher, brought to the interview (cf. Flick, 2009). Convenience sampling was adopted as a strategy to find participants for the study. Two organisations for blind people and one university body acted as the main vehicles to recruit interviewees. In particular, Irene Matassoni from Irifor Trento was the primary reference point in Italy, along with the personnel of the Disability Office of the University of Trento, while Paul de Nooij from Bartiméus helped me recruit contacts in the Netherlands. Additionally, snowball sampling was carried out for the purpose of reaching the desired sample size. The interviews were conducted between March and June 2016, and lasted between thirty minutes and one hour. They were recorded and transcribed using some of the symbols listed by Palermo (2009) on the basis of the SAMPA conventions (Speech Assessment Methods Phonetic Alphabet15; see Appendix B)16. Before starting the recording, all participants were informed about the goal of the study, its duration, and the background of the researcher, and were asked permission for recording. All interviewees consented. They were informed that their data would be anonymised and that they could discontinue their participation at any time. Accordingly, throughout the next chapters fictitious names are used to. 15 16. See www.phon.ucl.ac.uk/home/sampa. The transcriptions yield 262 pages in total and can be found as an appendix to this work.. 21.

(59) refer to the participants, with the exception of one participant who asked to be cited by his full name. The transcriptions (see Appendix C) were done in the same language used in the interview (either Italian or English, plus one predominantly in Dutch). However, all excerpts are presented in English with the original quote in the footnote if in a different language. My final, non-hypothetical corpus consisted of nine men and nine women of various ages, for a total of eighteen people (see Table 1): ten of them were Italian (with two cases of mixed nationality, Italian-Albanian and Italian-American), while the other eight were Dutch. Having samples almost equal in size, but different in terms of national belonging, enabled a preliminary comparison between two countries. The rationale behind the decision to conduct a comparative study is related to language factors. It was assumed that Italians tend to be exposed mostly to monolingual content, whereas Dutch people live almost in a condition of bilingualism with frequent switching between Dutch and English17. For this reason, it was also assumed that conducting the interviews with Dutch respondents in English was coherent with the aims of the research. In effect, it was decided to explore whether language settings combined with the language of the user and the idiom of the accessed content would affect the relation the user had with a given platform. As a matter of fact, in 2012 Façanha et al. highlighted the existence of language problems when the screen reader encountered content in an idiom different from its default language. If that happened, the screen reader read out the text as if it had been in the default language. That could make what was being said almost unintelligible. Ultimately, however, in my study only minor differences were found. I suggest that the absence of striking discrepancies can be interpreted as a signal of the high-quality performance level reached by assistive technology, at least within Western countries that use the Latin alphabet. Given that a different sight condition requires different strategies of use and access, it was also decided to include in the sample both blind people and people with low vision to evaluate the impact of different assistive technologies on platform access and use. In effect, Kleege (2005) DFWXDOO\QRWHGWKDWWKHEOLQGFDQUHWDLQWKHFDSDFLW\WR³SHUFHLYHOLJKWFRORUIRUPDQGPRYHPHQWWR VRPH GHJUHH´ S

(60) 6KH DOVR SRLQWHd out that, ³regardless of the degree or quality of residual vision, blind people differ widely in the ways that they attend to, use or value these perceptions´ (ibid.). Therefore, she argued, it would be desirable to stop thinking in terms of a blind/sighted binary and start considering a varied spectrum in visual acuity and visual awareness. Within my. 17. For detailed data see the survey on languages spoken in Europe per country requested by the European Commission (2012).. 22.

(61) sample, four persons (males) have been legally blind since birth or a very young age, five people (three males, two females) became blind due to degenerative illnesses, one (female) because of accident-induced brain damage, and eight people (six females, two males) have very low vision. Six persons are between 19-30 years old18, one person is between 31-40, five persons are between 41-50, one between 51-60, and five between 61-70. Although the population sector most likely to use highly visual online platforms comprises those aged between 18 and 50 years (cf. Duggan & Brenner, 2013; see also statistics by Statista Inc.), older interviewees turned out to also be quite active in terms of platforms they were registered for and applications they had downloaded. Therefore, their contributions were considered relevant to the research. In general, participants in this study showed a high degree of ICT literacy. Either they learnt to use a computer at a young age or have had years of experience using it first at work and then at home19 (see Table 2). Consequently, most are acquainted with assistive technologies: except for two Dutch people with low vision who can still manage simply by zooming in and regulating the brightness through an iPad, all other respondents possess screen readers (mainly JAWS or VoiceOver or both; Dutch participants also used NVDA), and ten people also own a braille display (see Table 3). Strikingly, with one single exception20, they all have a touchscreen device. Sixteen participants currently use a smartphone, i.e. an Apple iPhone, and carry out most digital activities with it. This might sound counterintuitive, given that touchscreens have often been considered inaccessible to blind people (see McGookin et al., 2008). However, thanks to software development and the introduction of screen-reading programs, the combination of audio inputs with continual practice on a spatially limited surface enables partially sighted or blind men and women to integrate touchscreen devices in their daily life. All interviewees expressed their satisfaction with the technology21 and only one of them raised questions regarding the monopoly Apple has managed to create by being more attentive to accessibility issues than other companies22. 18 19 20. 21 22. N.B. Some participants have their birthdays later in the year, which means that statistically they would be one year older. However, my report takes into account the age at the moment of the interview. This second point refers to older participants who had the chance of testing the first computers in the Nineties, mostly in the workplace, but did not grew up with a digital technologies, as it may be the case for younger interviewees. Giacomo (61, blind) indeed refuses to adopt a smartphone because he considers useless to invest time and effort to learn to use a touchscreen device when he can do everything through his old mobile phone and his computer deploying a normal keypad and keyboard plus braille assistive technologies. They all mentioned that Apple has developed particularly accessible systems and applications, which makes the L3KRQHPRUHVXLWDEOHIRU³EOLQG´XVHWKDQRWKHUVPDUWSKRQHV It should be pointed out that this person is specialising in telecommunication engineering, which makes him aware of risks and possibilities involved in the use of different kinds of technologies. He is also one of the few participants using Linux and the most opinionated about open access and the importance of offering full usability to all people.. 23.

(62) Table 1: Main demographics of the participants in the study. 24.

(63) Table 2: Professional background and/or current occupation of the participants in the study. 25.

(64) Table 3: Typologies of assistive technologies used by the participants in the study. 26.

(65) At this point, the reader might rightly wonder which platforms were specifically discussed during the conversations. The first interviews were used to better understand which platforms visually impaired users utilise the most. Subsequent conversations were meant to concentrate on a narrower selection of platforms. However, substantial discrepancies in the number and typology of platforms used emerged. Some people are extremely active on new media and have experimented with plenty of applications. Some others are more cautious in the use of digital platforms and access them to a limited extent. Some others expressed scepticism about social media, but showed an extremely high degree of digital literacy23. No pattern was found in terms of age, gender, nationality24, or educational/professional background. As a consequence, it was decided to keep the discussion open, without focussing too much on a specific platform25. Moreover, some questions were added starting from the very first interview, regarding the relevance and use of speech software such as Siri and instant messaging applications like WhatsApp. In this way, it was possible to discuss platform interaction in real time by looking at systems that are less visual, but potentially involve orality. In some cases, specific applications (e.g. TapTapSee and BeMyEyes26) were also mentioned by the participants, who considered them a positive outcome of technical developments of mobile phones with important implications for their daily lives. Nonetheless, the interviews were conducted with specific exemplary cases in mind, namely Facebook, Instagram, LinkedIn, Twitter, YouTube, and Google Search. Indeed, these are all highly visual platforms, which is intriguing because they raise the double-fold question of how visually impaired people use them and why they decided to make the effort to join. Curiosity was often mentioned as a drive to log in to certain platforms and download given applications that had become particularly popular among sighted people. As a result, my informants could either find out they were not in the position of using the platform (e.g. Instagram) or application (e.g. Telegram) or, if they were, they could keep using them to a greater or smaller extent. Other common reasons to register for Facebook and to use software such as Zello, WhatsApp, or Skype are communicating with 23 24 25 26. They also prefer the braille display over the screen reader (or often take advantage of a combination of the two devices at the same time) and use a larger variety of technologies and browsers than other users. Except that, on average, Dutch people tended to be more familiar with and active on LinkedIn than Italian users. LinkedIn is indeed less widespread in Italy, while it seems to be professionally important in the Netherlands. Moreover, some platform-specific studies already exist in the field of software engineering and computer sciences. See Buzzi et al. (2011) for Twitter and Buzzi et al. (2010) for Facebook. TapTapSee allows the user to take a picture of an object or person and get a short automatic description of the photographed subject. The photo is automatically eliminated after this. Instead, BeMyEyes enables a blind person to call a volunteer sighted user who gains temporary access to the blind (wo)man's camera and can read aloud what is needed (the expiration date of a medicine is the classical example).. 27.

(66) faraway friends, maintaining old contacts and creating new ones 27. Twitter is also used to follow interesting contacts, but above all it is regarded as a resource to keep oneself updated about current events. Entertainment was sometimes mentioned as a motivation too, especially by younger users. Whereas Facebook, Twitter, YouTube and Google Search turned out to also be quite popular among visually impaired people, LinkedIn and Instagram were barely used. The former was known (most interviewees had an account), but often not considered very useful: some interviewees claimed they were too young to have a good profile, some others did not need one being retired or unable to work. Moreover, most Dutch interviewees noted that LinkedIn is actually quite inaccessible. Instagram, meanwhile, is too visual and therefore useless for people who cannot see the images unless described and cannot know exactly what they are posting unless helped by a sighted person. &RQVLGHULQJWKDWWKHRSHQQHVVRIWKHFKRVHQPHWKRGRORJ\UHTXLUHVWKHUHVHDUFKHUWR³PHGLDWH between freedom to unfold subjective viewpoints and the thematic direction and limitation of what is PHQWLRQHG´ )OLFNS

(67) DQ interview guide was created in English and translated to Italian (see Appendix A). Its consistent use was meant to facilitate the collection of concrete statements about core issues and to increase the comparability of the data. The questionnaire prepared for the interviews was structured in five parts. The first one served to gather basic personal information (name, gender, age, profession, nationality and vision condition). Subsequently, the interviewees were asked about their use of online platforms according to four main leading questions, namely: a) how are social platforms perceived by users who need assistive devices to access them? In other words, how is the relationship between user and platform affected by the technological mediation of assistive devices? b) to what extent does the language of the user (combined with the language settings of the assistive device and the platform) impact on the user-platform relationship? c) what insights can we gain into the materiality of software from the disruption of apparent seamlessness due to the mediation operated by assistive technologies? d) how does a slower interaction affect the user of platforms that are supposed to work in real time? Each research question was translated into several open questions that were posed to the participants. Follow-up questions were employed on a case-by-case basis to further develop unclear or interesting statements. 27. Façanha et al. (2012) KDG DOUHDG\ QRWHG WKDW ³>D@tualmente, uma das formas de criar e manter vínculos sociais é através da utilização dessas redes virtuais seja para a publicação de vídeos e de fotos (e.g., Youtube) ou simplesmente para compartilhar opiniões, conteúdos e status (e.g., Facebook e Twitter

(68) ´ S

(69) - ³&XUUHQWO\RQHRIWKH ZD\VWR create and maintain social bonds is through the use of these virtual networks be it for the publication of videos and photos (e.g., Youtube) or simply for sharing opinions, contents and statuses (e.g., Facebook and Twitter

(70) ´ DXWKRU V translation).. 28.

No results found