Selected Papers of AoIR 2016: The 17th Annual Conference of the Association of Internet Researchers Berlin, Germany / 5-8 October 2016 Suggested  Citation  (APA):  Carmi,  E.,  Gehl,  R.,  Sullivan,  J.,  Burk,  D.  (2016,  October  5-­8).  Network   Standards  and  Culture.  Paper  presented  at  AoIR  2016:  The  17th  Annual  Conference  of  the  Association  of   Internet  Researchers.  Berlin,  Germany:  AoIR.  Retrieved  from  http://spir.aoir.org.   NETWORK  STANDARDS  AND  CULTURE       Elinor  Carmi,  Goldsmiths,  University  of  London.   Robert  W.  Gehl,  The  University  of  Utah   John  L.  Sullivan,  Muhlenberg  College   Dan  L.  Burk,  University  of  California,  Irvine     Media  networks  are  complex  assemblages  that  connect  different  kinds  of  humans  and   objects  together,  and  involve  multiple  actors,  materials,  intentions  and  products.   Standards  are  usually  meant  to  make  technical  artefacts,  protocols,  features  and   infrastructures  operate  in  an  efficient  and  uniform  way.  In  this  way,  standards  become   taken  for  granted,  and  the  rules  they  structure  are  seen  as  the  natural  and  only  way  to   engage  with  media.  But  behind  such  norms  lay  political,  ideological  and  cultural   struggles  of  power  between  different  interests  group  that  range  between  individuals,   private  companies,  governments,  NGOs  and  legal  entities.  These  multiple  actors   compete  over  a  dominant  position  in  the  media  industry  and  their  ability  to  influence,   control  and  shape  the  way  that  people  use  and  understand  media  and  communications.   This  panel  seeks  to  unveil  these  invisible  codes  and  the  conflicts,  incentives  and  politics   that  lead  to  their  formation.       Specifically,  Elinor  Carmi's  paper  examines  the  European  Union's  cookie  regulations,   focusing  on  the  construction  of  standards  that  allow  for  the  distinctions  between  cookies   and  spam.  Robert  W.  Gehl's  paper  takes  on  a  similar  object,  but  in  the  context  of  the   United  States,  looking  at  the  Digital  Advertising  Alliance's  self-­regulatory  standards,   which  purportedly  limit  the  online  advertising  industry's  capacity  to  monitor  Internet   users.  John  Sullivan  tackles  the  proposed  HTML5  standard  for  "Encrypted  Media   Extensions."  Finally,  Dan  Burk  considers  the  legal  and  technical  questions  of   interoperability  in  smartphone  platform  and  networking  standards.         2 1.  EAT  ME!  SPAM  VERSUS  COOKIES  IN  THE  POLITICS  OF  THE   EUROPEAN  UNION  INTERNET.       Elinor  Carmi,   Goldsmiths,  University  of  London.       Introduction       This  paper  examines  the  struggles  between  the  European  Union  and  private  companies   when  it  comes  to  Internet  definitions  and  standards,  while  looking  at  spam  and  web-­ cookies  as  a  case  study.  Telecommunications  in  Europe  were  controlled  by  national   monopolies  until  the  end  of  the  1980s.  According  to  Savin  (2014),  this  was  a  slow   transition  of  territorial  powers,  characterized  by  a  struggle  over  authority  on   telecommunications  policy  between  the  European  Commission  and  national  Postal,   Telegraph  and  Telecommunications  Administrations  (PTTs)  who  wanted  to  keep  the   status  quo  of  their  position  within  the  media  sector.  As  Savin  argues,  “[m]ember  states   were  not  readily  willing  to  abandon  their  monopolies  and  were  by  and  large  not   supportive  of  the  Commission’s  initiative.  The  late  1980s  and  early  1990s  saw  an   increasingly  complex  web  of  proposed  and  adopted  directives,  a  tug  of  war  between  the   Commission  and  the  member  states,  of  governmental  and  non-­governmental  actors”   (Savin,  2014,  p.  3).     The  mid  1980s,  as  Linda  Senden  (2005)  observes,  were  also  characterized  by  a   change  in  the  European  Community  (changed  to  the  European  Union  in  1993,  following   the  Maastricht  Treaty)  approach  to  legislation  towards  ‘soft-­law’.  This  approach  means   that  the  main  instruments  of  governance  in  the  European  Union  are  co-­regulation  and   self-­regulation.  These  instruments  have  no  legally  binding  force,  and  yet  still  hold  a  legal   effect.  This  approach  is  usually  contrasted  with  ‘hard  law’,  a  ‘legally  binding  obligations   that  are  precise  (or  can  be  made  precise  through  adjudication  or  the  issuance  of   detailed  regulations)’  (Abbott  and  Snidal,  2000,  p.  421).  Among  other  reasons,  the   transition  to  soft-­law  is  the  acknowledgment  that  legislation  is  not  sufficient  to  govern   specific  sectors,  especially  when  it  comes  to  the  Internet  which  crosses  states   territoriality.       In  the  European  Union,  this  power  conflict  becomes  complicated  as  actors  who  are   involved  in  creating  Internet  standards,  protocols  and  enforcement  negotiate  between   Member  States,  zooming  out  to  the  European  Commission,  and  to  global  actors  such  as   the  International  Telecommunication  Union  (ITU),  the  World  Wide  Web  Consortium   (W3C),  the  Internet  Corporation  for  Assigned  Names  and  Numbers  (ICANN)  and  the   Internet  Engineering  Task  Force  (IETF).  Most  of  these  organizations  were  founded  and   based  in  the  U.S.A,  and  receive  criticism  on  the  centrality  of  the  values,  language  and   standards  that  stems  from  that  region,  which  consequently  project  on  the  global  Internet   (DeNardis,  2009,  p.  37).  These  struggles  are  often  termed  ‘Internet  governance’,  which   “designates  the  technoeconomic  and  legal  issues  arising  from  any  decisions,  de  facto  or   3 by  law,  that  affect  the  design,  access,  and  use  of  the  Internet  as  a  specific  sort  of   communication  network  architecture”  (Franklin,  2013,  p.  138).  This  paper  argues  that   under  the  European  Union’s  soft-­law  approach,  other  organizations  have  a  say  in  the   way  that  the  Internet  operates,  its  standards,  and  architecture.       As  media  companies  struggle  with  states,  when  it  comes  to  creating  and  enforcing   definitions,  standards  and  infrastructure,  it  is  important  for  media  scholars  to  engage   with  legal  and  technical  texts  in  order  to  shine  a  light  on  procedures  that  establish  the   way  people  use  and  understand  media  and  communication.  Spam  is  a  good  case  study   to  focus  on  as  a  phenomenon  that  has  been  taken  for  granted  as  an  unwanted   behaviour  on  the  Internet.  This  category  is  usually  invisible  and  hidden  from  users,  but   while  most  people  think  it  means  Viagra  and  Nigerian  Princess  scam  emails,  it  actually   has  no  clear  definition  within  legislation.  In  European  Union  legislation,  spam  is  not   directly  addressed  or  defined.  This  makes  spam  unique  because  both  law  makers  and   computer  scientists  need  specific  definitions  of  categories  to  make  their  laws  and  codes   operational.     This  paper  examines  how  unwanted  forms  of  communications  are  categorized  as  spam   while  others  as  cookies,  and  what  are  the  consequences  of  such  a  process.  I  conduct   policy  analysis  of  European  Union  legislation  related  to  spam  and  cookies,  mainly  by   the  European  Commission,  the  Organization  for  Economic  Co-­operation  and   Development  (OECD),  and  the  Article  29  Working  Party1  (A29WP).  In  addition,  I   examine  technical  documents  of  the  Internet  standards  organisation  the  Internet   Engineering  Task  Force  (IETF)  and  the  World  Wide  Web  Consortium  (W3).         SpamALot       Media  law  literature  has  given  little  attention  to  spam.  The  most  notable  exception  is   Finn  Brunton  (2013),  who  gives  some  account  of  spam  related  laws,  focussing  on  North   America  and  especially  on  the  CAN  SPAM  Act  in  2003.  Milton  Mueller  (2010),   discusses  Internet  governance,  and  gives  some  account  to  spam  blacklists,  but  adopts   the  common  notion  that  spam  mainly  relates  to  emails.  Another  important  contribution  is   Lodewijk  Asscher  and  Sjo  Hoogcarspel’s  (2006)  piece  on  spam  regulation  in  the   European  Union  context,  which  however  analyzes  the  e-­Privacy  Directive  through  a   legal  lens  and  thus  provides  limited  understanding  of  spam  as  a  media  phenomenon.       Another  body  of  literature  related  to  spam  comes  from  media  studies  and  information   security,  and  involves  various  practices  conducted  on  the  Internet,  which  are  usually   associated  with  the  category  of  spam  as  a  verb;;  political  economy  (Parikka  &  Sampson,   2009;;  Brunton,  2013),  hacking  (Galloway,  2004;;  Coleman,  2013),  anti-­virus  companies   1  European  advisory  body  which  consists  data  protection  authority  representative  from  each  of  the   member  states,  and  was  founded  following  Article  29  of  the  Data  Protection  Directive  (95/46/EC).  The   main  goal  of  this  body  is  to  provide  opinions  and  recommendations  to  the  European  Commission  in   regards  to  data  protection.   4 (Helmreich,  2000),  and  socialbots  (Gehl,  2013;;  Bucher,  2014).  These  various  directions   do  not  always  touch  spam  directly  but  are  interpreted  as  such  in  different  discourses   and  consequently  point  to  avenues  in  categorizing  different  forms  of  information.         What’s  on  the  menu?       These  authors  provide  important  insights  to  the  field  of  spam;;  however,  they  tend  to   automatically  adopt  computer  scientists’  definitions  of  spam  without  questioning  them.   In  the  field  of  computer  science,  spam  is  usually  defined  in  a  binary  way  of  spam/non-­ spam,  but  as  will  be  shown;;  this  distinction  is  more  nuanced,  and  subjected  to  debate   and  power  struggles.  This  paper  will  address  these  issues  and  show  how  spam  and   cookies  are  important  to  the  construction  of  the  European  Union  Internet.  This  paper   argues  that  the  European  Union’s  soft-­law  approach,  that  prioritizes  private  actors’  self   regulation  by  adopting  their  standards,  protocols  and  enforcement,  contributes  to  the   institutionalization  of  e-­commerce.         References       Abbott,  K.  W.,  &  Snidal,  D.  (2000).  Hard  and  soft  law  in  international   governance.  International  organization,  54(03),  421-­456.       Asscher,  H.,  &  Hoogcarspel,  S.  (2006).  A  European  Perspective  after  the  Adoption  of   the  E-­Privacy  Directive.  The  Hague.       Brunton,  F.  (2013).  Spam:  a  shadow  history  of  the  Internet.  Mit  Press.       Bucher,  T.  (2014).  About  a  bot:  Hoax,  fake,  performance  art.  M/C  Journal,17(3).       Coleman,  E.  G.  (2013).  Coding  freedom:  The  ethics  and  aesthetics  of  hacking.   Princeton  University  Press.       DeNardis,  L.  (2009).  Protocol  politics:  The  globalization  of  Internet  governance.  Mit   Press.       Franklin,  M.  I.  (2013).  Digital  dilemmas:  Power,  resistance,  and  the  Internet.  Oxford   University  Press.     5   Galloway,  A.  R.  (2004).  Protocol:  How  control  exists  after  decentralization.  MIT  press.       Gehl,  R.  W.  (2013).  The  computerized  socialbot  Turing  Test:  new  technologies  of   noopower.  Available  at  SSRN  2280240.       Helmreich,  S.  (2000).  Flexible  infections:  computer  viruses,  human  bodies,  nation-­ states,  evolutionary  capitalism.  Science,  Technology  &  Human  Values,  25(4),  472-­491.       Mueller,  M.  L.  (2010).  Networks  and  states:  The  global  politics  of  Internet  governance.   Mit  Press.       Parikka,  J.,  &  Sampson,  T.  D.  (Eds.).  (2009).  The  spam  book:  on  viruses,  porn,  and   other  anomalies  from  the  dark  side  of  digital  culture.  Hampton  Press.       Savin,  A.  (2014).  How  Europe  formulates  internet  policy.  Internet  Policy  Review,  3.   Available  at:  http://policyreview.info/articles/analysis/how-­europe-­formulates-­internet-­ policy.         Senden,  L.  A.  (2005).  Soft  law,  self-­regulation  and  co-­regulation  in  European  law:  where   do  they  meet?.  Electronic  Journal  of  Comparative  Law,  9(1),  Available  at:   http://www.ejcl.org/91/abs91-­3.html.         6 2.  STANDARDS,  SELF-­REGULATION,  AND  SWARMS:  THE  CASE  OF   THE  DIGITAL  ADVERTISING  ALLIANCE     Robert  W.  Gehl   The  University  of  Utah       Abstract       This  essay  explores  the  Digital  Advertising  Alliance  (DAA),  an  online  advertising  and   marketing  trade  consortium  comprised  of  seven  organizations:  the  Better  Business   Bureau  (BBB),  the  Direct  Marketing  Association  (DMA),  the  Network  Advertising   Initiative  (NAI),  the  4As  (an  advertising  agency  trade  group),  the  Association  of  National   Advertisers  (ANA),  the  American  Advertising  Federation  (AAF),  and  the  Interactive   Advertising  Bureau  (IAB).  Together  under  the  aegis  of  the  DAA,  these  organizations   and  their  members  represent  a  massive  percentage  of  the  global  advertising  and   marketing  industry,  an  industry  based  on  sophisticated  tracking  technologies,  practices   of  profiling  consumers,  and  increasingly  precise  customization  of  ads,  sales  pitches,   pricing  schemes,  and  product  offerings  (Turow,  2011).       This  paper  considers  the  DAA  as  a  standards  organization.  The  DAA  sets  rules  and   protocols  for  the  U.S.'s  online  advertising  industry.  It  does  so  as  a  "self-­regulatory"   body,  meaning  it  regulates  its  members  to  avoid  having  the  U.S.  government  do  so.  The   DAA's  standards  are  thus  a  point  of  contact  between  it,  its  members,  online  consumers,   and  the  U.S.  government.     This  paper  thus  focuses  on  the  technical,  social,  and  political  aspects  of  the  DAA   standards,  specifically  its  opt-­out  cookie  system  (available  at  aboutads.info/choices).  I   will  argue  that  the  DAA's  opt-­out  system  is  designed  to  extend  and  maintain  the  power   of  advertising  in  the  political  economy  of  the  Internet  in  two  key  ways.  First,  the  DAA   seeks  to  present  its  system  as  a  singularity  when  it  is  arguing  for  the  effectiveness  of  its   self-­regulation.  The  DAA  refers  federal  regulators  to  its  opt-­out  system,  making  claims   about  it  being  a  single,  easy-­to-­use,  consumer-­friendly  system,  thus  negating  the  need   for  other  efforts  (such  as  Do  Not  Track  headers).  So  far,  this  argument  has  helped  stave   off  U.S.  government  regulation  of  online  advertising.     However,  I  will  also  argue  that  the  DAA's  standards  often  deploys  multiplicity  and   heterogeneity  in  service  to  its  overall  associating  practices.  Specifically,  I  will  closely   examine  the  structure  of  the  DAA  opt-­out  system,  noting  how  its  seeming  singularity   dissolves  into  a  swarm  when  actual  consumers  confront  it.  For  consumers,  the  DAA   system  is  baroque  and  confusing,  with  over  150  cookies  to  be  installed  on  each  and   7 every  end-­user  device.  Each  advertising  network  is  allowed  to  implement  the  opt-­out   standards  in  different  ways,  leading  to  diffusion  rather  than  rigidly  structured  practices.     At  the  core  of  this,  I  argue,  are  the  DAA  standards,  which  are  a  sophisticated  mix  of   specified  technical  procedures  and  flexibility,  allowing  for  the  DAA  to  support  its  claims   about  self-­regulation  while  allowing  its  members  enough  flexibility  to  continually   innovate  new  surveillance  techniques  even  in  the  face  of  consumer  rejection  of  online   tracking.     To  do  this  work,  I  will  deploy  two  concepts  from  Actor-­Network  Theory  (ANT),   punctualization  and  depunctualization,  as  key  lenses  through  which  to  see  the  shifting   contours  of  the  Digital  Advertising  Association  as  it  confronts  other  actor-­networks.  As   such,  in  addition  to  being  a  contribution  to  scholarship  on  standards,  the  paper   contributes  to  ANT.  Specifically,  I  argue  that  ANT  scholarship  on  organizations  tends  to   be  biased  towards  those  organizations  that  successfully  punctualize  –  that  is,   successfully  cohere  into  an  organization.  In  this  view,  an  organization  that  fails  is  said  to   have  "depunctualized."  Against  this,  drawing  on  the  work  of  John  Law  and  Annmarie   Mol,  I  argue  that  the  DAA  is  capable  of  depunctualizating  –  that  is,  becoming  a  swarm  –   at  key  moments.  In  other  words,  it  sometimes  use  messiness  and  incoherence  in  its   overall  mission  to  maintain  its  role  as  the  spokes-­thing  for  the  online  advertising   industry.  It  falls  apart  in  order  to  cohere.  This  theoretical  point  is  often  lost  in  ANT  work   on  organizations.   The  empirical  data  for  this  project  comes  from  press  releases,  legislative  testimony,  and   the  architectural  and  technical  structures  of  the  DAA  and  DAA-­member  Web  sites.  I  will   also  draw  on  the  scholarly  literature  on  online  advertising,  human-­computer  interaction,   and  standards  consortia.  My  ultimate  argument  is  that  ANT  scholarship  must  consider   how  standards  are  just  as  much  about  deploying  heterogeneity  as  they  are  about   cohering  and  channeling  practices  into  durable  singularities.       References       Turow,  J.  (2011).  The  daily  you:  how  the  new  advertising  industry  is  defining  your   identity  and  your  worth.  New  Haven:  Yale  University  Press.     8 3.  TECHNICAL  AND  ETHICAL  DISCOURSES  IN  WEB   STANDARDIZAION:  THE  CASE  OF  ENCRYPTED  MEDIA  EXTENSION   (EME)  AND  HTML5     John  L.  Sullivan   Muhlenberg  College       Technical  standardization  has  been  key  for  efficiency  and  ease  of  information  exchange   among  inter-­connected  systems.  Standards  are  “specifications  that  determine  the   compatibility  of  different  products”  (Stango,  2004,  p.  2).  As  Russell  (2014,  p.  16)   describes,  standards  can  also  describe  “a  social  process  by  which  humans  come  to   take  things  for  granted.  Through  standardization,  inventions  become  commonplace,   novelties  become  mundane,  and  the  local  becomes  universal.”  Standardization  is  the   term  that  therefore  “describes  the  process  of  making  standards  –  a  process  that  entails   a  wide  range  of  practices  and  ideas  with  distinct  political,  economic,  and  cultural   dimensions”  (p.  17).  In  line  with  this  socio-­technical  approach  to  standardization  (Busch,   2011;;  DeNardis,  2011;;  Russell,  2014;;  Sutor,  2011),  I  consider  the  process  of  web   standardization  and  the  complex  interplay  of  competing  discourses  about  the  nature  of   standardization  itself.       This  essay  explores  the  debates  surrounding  the  recent  introduction  of  a  controversial   feature  within  the  HTML5  web  standard,  the  latest  version  of  the  hypertext  markup   language  that  acts  as  a  universal  standard  for  rendering  webpages  on  the  World  Wide   Web.  The  HTML5  standard  was  developed  beginning  in  2003  under  the  auspices  of  the   World  Wide  Web  Consortium  (W3C),  a  non-­profit  consortium  founded  by  Tim  Berners-­ Lee  in  1994  to  guide  the  development  of  standards  for  the  web.  The  W3C  is  made  up  of   member  organizations,  including  universities,  non-­profit  foundations,  NGOs,  and   technology  corporations.       HTML5  is  the  central  linchpin  in  the  seamless  operation  of  the  web,  but  its  importance   goes  beyond  rendering  webpages.  HTML5  represents  the  vision  of  a  unified  online   content  experience  that  extends  from  computers  to  mobile  platforms.  As  Schrock  (2014,   p.  821)  has  argued,  HTML5  “creates  a  common  platform  for  producing  mobile  software   that  runs  across  multiple  smart  phones  and  tablets  that  alternately  pushes  against  and   aligns  with  corporate  desires  and  ideologies.”  Given  the  fact  that  the  HTML5  protocol  is   the  core  standard  at  the  heart  of  the  world  wide  web  –  the  “lingua  franca”  for  the  internet   –  it  is  notable  that  its  development  has  received  relatively  little  attention  by  scholars  (for   good  examples  see  Goggin,  2010;;  Horst,  2013;;  Russell,  2011;;  Schrock,  2014).  While   previous  research  by  Schrock  has  explored  the  concept  of  “openness”  in  the   development  of  the  HTML5  protocol  and  the  corporate  imperatives  for  the  development   of  a  cross-­platform  web  experience,  my  focus  here  is  on  the  competing  discourses  of   standardization  within  the  W3C.     9   A  Moment  of  Standardization:  Encrypted  Media  Extensions  (EME)       This  essay  considers  an  episode  in  the  development  history  of  the  HTML5  standard,   specifically  the  introduction  of  “Encrypted  Media  Extensions”  (EME)  in  2012-­13.  On   February  21,  2012,  a  proposal  to  implement  EME  within  the  HTML5  standard  was   distributed  via  email  to  members  of  the  HTML  Working  Group  of  the  W3C  (Dashevsky,   2013).  The  technical  document,  authored  by  four  representatives  from  Google,   Microsoft,  and  Netflix,  proposed  to  eliminate  the  need  for  end  users  to  install  browser   plugins  such  as  Flash  or  Silverlight  by  allowing  pockets  of  proprietary  software  to  run   within  the  open  HTML5  standard  for  the  purpose  of  protecting  copyrighted  audio  and   video  (Dorwin,  Smith,  Watson,  &  Bateman,  2015).  The  proposal  caused  considerable   controversy  among  the  members  of  the  working  group  (Constantin,  2012).  Proponents   argued  that  the  inclusion  of