Academia.eduAcademia.edu

Outline

SDSE 2013: why digital humanists should get out of textual scholarship

Abstract

This paper presents, somewhat polemically, the view that textual scholars need to immerse themselves in the digital world, and take full responsibility themselves for the digital editions they make. This requires a rethinking of the model of collaboration between textual scholar and digital humanist which has reigned for twenty years: what the paper calls the 'one scholar/one project/one digital humanist' model. The paper should be read with a companion blog, at http://scholarlydigitaleditions.blogspot.com/2013/07/why-digital-humanists-should-get-out-of.html, which elaborates an alternative model for collaboration between textual scholars and digital humanists (basically: many scholars/many projects/many digital humanists), and advocates too the widespread adoption of Creative Commons attribution share-alike licences (without the toxic 'non-commercial' restriction) for edition materials, and for their availability through open APIs, independent of any one interface. Joris van Zundert's blog at http://brandaen.huygensinstituut.nl/?p=497 (note his comment on his own blog) also contains arguments which should be read alongside this paper.

What  digital  humanists  don’t  know  about  scholarly  editing;  what  scholarly  editors   don’t  know  about  the  digital  world     I  am  a  simple  person,  and  when  I  first  started  thinking  of  this  conference,  I  had  a   very  simple  idea:  let’s  just  get  as  many  digital  humanists  and  as  many  textual   scholars  as  we  can  within  one  room  for  three  days  and  let’s  see  what  happens.    It’s   possible  that  no  area  of  the  humanities  has  as  much  to  benefit  from  the  digital   revolution  as  textual  scholarship,  and  digital  humanities  may  have  more  to  gain   from  textual  scholarship  than  any  other  area.    These  two  disciplines  have  already   had  a  lot  to  say  to  each  other.    We  have  three  days  now  to  talk  to  each  other.  But   every  conversation,  we  all  know,  carries  this  danger:  when  we  think  we  understand   the  other  person  but  actually,  we  don’t.    Even  more  dangerous,  I  think,  is  the   situation  when  not  only  does  the  other  person  think  he  or  she  understands,  but  you   think  the  other  person  understands  you  -­‐-­‐  but  actually,  they  do  not.    Then,  you  are   both  really  in  trouble.     Let’s  start  with  the  digital  humanist  side  of  the  conversation.    What  do  digital   humanists  think  they  know  about  textual  scholarship,  and  how  are  they  wrong?  I   think  digital  humanists  have  two  fundamental  misconceptions  about  textual   scholarship.    The  first  misconception  is  this:that  textual  scholarship  is  all  about  what   John  Unsworth  calls  the  ‘primitives’  of  humanist  scholarship:  discovering,   annotating,  comparing,  referring,  sampling,  illustrating,  and  representing.    We  find   texts,  we  record  information  about  them,  we  create  ways  of  representing  the  text,   we  compare  the  texts,  we  annotate  them.    Yes,  this  is  correct.    Yes,  we  do  catalogue   manuscripts,  we  describe  them,  we  transcribe  them,  we  collate  them.    All  these  are   indeed  parts  of  our  work.    And  all  these  are  highly  suitable  cases  for  computer   treatment,  which  is  why  digital  humanists  are  so  fond  of  textual  scholars.  But  it  is  an   error  to  think  that  these  activities  completely  describe  textual  scholarship.    They  do   not.    You  can  do  these  things  and  create  a  digital  archive.    But  if  this  is  all  you  do,  you   will  not  create  an  edition.    An  edition  requires  that  you  make  decisions:  who  is  the   edition  made  for?  What  text,  or  texts,  should  it  present?  How  should  these  texts  be   prepared;  who  should  do  the  preparation;  how  should  the  texts  be  presented;  what   should  be  the  basis  of  the  texts  presented?  These  decisions  are  often  difficult,  even   controversial.    They  are  usually  made  by  a  single  editor,  or  by  a  small  group  of   scholars.    Clearly,  the  editors  who  make  these  decisions  need  the  catalogues,  the   transcriptions,  the  collations;  they  need  the  digital  archives  that  digital  humanists   have  become  so  good  at  making.    In  turn,  the  digital  archives  must  be  shaped  by  the   needs  of  textual  scholars,  so  that  the  materials  they  make  are  fit  for  use.     Hence:  the  common  first  error  made  by  digital  humanists,  -­‐-­‐  and,  I  fear,  by  many   editors:  that  the  needs  of  textual  scholars,  and  indeed  the  interests  of  readers,  can   be  perfectly  served  by  digital  archives.    They  cannot.    An  edition  is  an  argument   about  a  text.    We  need  arguments;  without  arguments,  our  archives  are  inert  bags  of   words  and  images.         This  overconfidence  by  digital  humanists,  that  they  believe  that  they  understand   what  editing  is,  leads  to  the  second  common  error  digital  humanists  make.  It  is  this:   that  digital  humanists  may  usefully  collaborate  with  textual  scholars.    After  twenty   years  of  working  with  my  digital  humanist  hat  on,  as  a  collaborater  with  textual   scholars,  I  can  tell  you  now:  this  collaboration  is  a  mistake.    We  have  had  to  work   this  way  for  two  decades  because  the  digital  tools  needed  to  make  a  scholarly   edition  have  been  so  difficult  to  use  that  no  sane  editor    -­‐-­‐  indeed,  no  sane  person  -­‐-­‐   would  want  to  use  them.  We  had  to  work  this  way,  too,  because  we  needed  to   discover  what  digital  tools  could  do.  But  we  should  seek,  as  soon  as  we  can,  to  end   this  collaboration.    Textual  scholars  should  not  need  a  digital  humanist  to  be  able  to   catalogue  manuscripts,  to  make  transcriptions,  to  compare  texts,  to  analyze  the   differences  between  texts.    Now,  I  put  my  textual  scholar  hat  back  on,  and  assert:  I   want  to  do  these  things  myself.    I  want  to  decide  what  to  transcribe,  how,  and  what   should  be  compared,  and  then  I  want  to  do  it  myself  –  or  delegate  it  to  other  people   who  are  interested  and  expert  in  the  text.    I  want  tools  to  do  this  work,  certainly,  and   digital  humanists  will  be  needed  to  make  these  tools.    But  I  should  be  able  to  use   these  tools  directly,  myself,  without  unreasonable  expenditure  of  time  or  money  or   effort,  to  make  the  editions  I  want  to  make.    I  do  not  want  a  digital  humanist  as  a   collaborator,  to  mediate  this  work:  I  want  to  engage  with  it  directly,  myself  and  with   others  like  me.  And  what  I  want  to  do,  myself,  every  editor,  everyone  who  wants  to   make  an  edition,  should  be  able  to  do.    I  said  earlier,  that  the  most  dangerous   misconceptions  are  those  which  are  shared.    In  the  last  decades,  too  many  editors,   myself  included,  have  subscribed  to  both  notions:  that  digital  archives  may  stand  in   place  of  editions  (or  might  even  be  editions  –  social  editions,  perhaps);  and  that   editors  cannot  do  this  work  on  their  own.    It  is  time  to  assert  the  value  of  editions  in   the  digital  age,  as  fundamental  discourses  about  the  texts  which  are  important  to  us.     It  is  time  for  us,  as  editors,  to  take  back  control  over  the  texts  we  make,  and  how  we   make  them.    Digital  humanists  should  get  out  of  textual  scholarship:  and  if  they  will   not,  textual  scholars  should  throw  them  out.       The  misconceptions  do  not  lie  just  on  one  side.  There  is  one  thing  scholarly  editors   do  not  know  about  the  digital  world  and  which  they  have  to  learn.    It  is  this:  for  the   editions  they  make  to  go  on  living,  to  be  useful  and  to  continue  to  be  useful,  to  go  on   being  read  for  decades  and  centuries,  the  editor  must  give  up  control.    This  is   actually  the  reverse  of  what  is  the  case  for  the  print  world,  where  the  publisher’s   need  for  exclusive  copyright  and  the  editor’s  desire  for  control  join  in  a  cosy   conspiracy,  issuing  in  a  pristine  unchanging  printed  book.    Editors,  hear  this:  if  you   want  your  digital  edition  to  be  used  by  others,  to  become  the  centre  of  other   people’s  work,  to  be  transfigured  over  and  over  into  bright  new  forms  of  scholarship,   if  you  you  have  to  release  it  without  restriction,  to  permit  both  commercial  and  non-­‐ commercial  work,  to  license  the  making  of  any  and  all  derivative  works.    Scholars,  I   know  from  bitter  experience,  find  this  very  difficult  to  do.    Years,  decades  and  more,   of  close  devotion  to  their  text  leads  to  it  becoming  ‘my  text’  (or,  ‘my  manuscript’).    At   the  least,  they  do  not  want  their  mortal  enemy  (and  every  scholar  wears  his  enemies   as  a  badge  of  distinction)  to  use  their  text  in  his  or  her  own  edition.  Again,  a  cosy   conspiracy  reinforces  this  short-­‐sighted  choice:  the  digital  humanities  centre  which   collaborated  in  the  making  of  the  edition  would  like  to  keep  exclusive  control  of  the   text,  to  use  as  coin  towards  its  own  continued  funding.    Editors,  just  say  no  to  this.   Grant  funders  too  have  been  far  too  forgiving  of  this  attitude.    I  said  a  few  minutes   ago  that  we  textual  scholars  should  take  control  of  what  we  do.    But  if  what  we  do  is   continue  to  make  materials  which  we  treat  as  our  private  property,  while   pretending  that  they  are  available  to  everyone,  we  will  drive  ourselves  and  our   editions  into  a  cul-­‐de-­‐sac.    We  have  to  open  up  what  we  do  to  others.  We  may  think   others  may  misinterpret,  misunderstand  and  distort  what  we  make  –  but  the  only   edition  which  is  never  misused  is  the  edition  which  is  never  used.  We  need  to  make   the  best  editions  we  can:  and  then,  we  have  to  see  these  editions  not  as  the  end  of   the  dialogue,  but  as  moves  in  a  conversation  which  others  will  continue.     Here  is  my  purpose  in  rehearsing  this  catalogue  of  errors.    In  the  last  decades,  we   have  solved  almost  all  the  technical  problems  associated  with  the  making  of  editions.     We  can  transcribe,  encode,  capture  images,  link  images  and  text,  compare,  analyze,   and  present,  all  in  digital  form.  We  can  now  invite  others  to    join  us  in  this  work,  in   huge  numbers  perhaps,  so  that  many  hands  may  make  light  and  pleasant  work.    For   years,  in  the  early  days  of  the  marriage  of  digital  methods  and  textual  scholarship,   our  choices  were  severely  limited  by  what  was  possible.    Our  choices  are  no  longer   so  circumscribed:  rather,  we  now  have  all  the  choices  we  always  had  as  scholarly   editors  plus  now  all  the  choices  digital  technology  gives  us.    So,  what  choices  should   we  make?    Where  should  we  put  our  resources?  What  models  should  we  aim  at?     Rather  than  try  to  decide  these  choices,  I’d  like  to  set  out  some  assertions,  which   might  help  us  towards  these  decisions.    First,  what  we  decide  to  do  depends  on  what   we  think  a  scholarly  edition  is.    There  has  been  a  great  deal  of  rhetoric,  some  of  it   from  myself,  the  last  decades  about  how  scholarly  editions  and  editing  have  been   fundamentally  changed  by  the  digital  turn.  So  let  me  say  it  plainly.    I  don’t  think   there  has  been  any  such  change.    A  scholarly  edition  is  still,  as  it  has  been  for   centuries,  an  argument  about  a  text.    The  fundamental  players  in  this  argument  are   still  documents,  works,  and  the  editor’s  interpretation  of  them.    The  editor  is  the   editor,  and  not  a  “facilitator”.    There  are  still  many  more  readers  than  editors,  and   most  readers  do  not  want  to  be  editors.    Certainly,  there  have  been  changes  –  most   of  them,  I  think,  for  the  better.    The  materials  upon  which  editions  are  based  may  be   far  more  accessible,  and  the  roots  of  the  edition  in  the  materials  far  better   articulated.    The  editor’s  choices  may  be  examined  much  more  closely,  and  editors   accordingly  more  humble  about  their  choices.    Particularly,  we  are  seeing  the  rise  of   open  collaborations,  as  editions  become  grounded  in  communities  of  individuals   who  contribute  to  their  making,  under  the  leadership  of  editors.  We  may  also  see   materials  made  for  an  edition  taken  and  repurposed  in  new  editions,  even  against   the  original  edition,  in  long-­‐running  arguments  about  the  text.  We  are  seeing  many   new  tools,  and  we  need  yet  more,  particularly  for  analyzing  and  visualizing  variant   texts.    However,  some  things  we  are  not  seeing,  and  we  should  not  expect  to  see.    We   will  not  see  ‘social  text  editions’and  we  will  not  see  ‘social  editions.’     Here  is  what  we  should  do,  then.    We  should  work  towards  a  future  where  anyone   who  wants  to  be  an  editor  has,  fit  and  ready  for  their  hands,  the  tools  for  themselves   to  make  the  editions  they  want.    Included  in  this  toolset  should  be  the  ability  to   invite  others  to  contribute,  and  to  allow  others  to  invite  themselves  to  contribute.    In   this  future,  parts  of  the  edition,  and  the  materials  upon  which  it  stands,  may  be   endlessly  and  continually  repurposed  by  others.    For  this  to  happen,  we  need  far   better  systems  for  enabling  the  discovery  and  reuse  of  editorial  materials.     Particularly,  we  need  agreements  on  conventions  for  naming  things:  for  naming   documents,  the  works  they  instance,  the  texts  of  works  within  the  documents,  and   all  their  parts.    We  need  a  real  commitment  to  real  open  texts,  really  available  for   others  to  use  and  alter  and  hand  on  to  others,  and  not  the  feeble  gestures  towards   openness  with  which  most  projects  satisfy  themselves.     It  comes  down,  in  the  end,  towards  our  vision  of  what  we  want.  We  can  carry  on  as   we  are:  with  a  few  extremely  expensive  digital  scholarly  edition  projects,  made  by   the  collaboration  between  a  very  textual  scholars  and  a  very  few  digital  humanists,   leaving  everyone  else  to  work  as  best  as  they  can  with  inadequate  tools:  putting   editions  and  their  materials  into  wikis,  into  blogs,  or  into  closed  proprietary  systems,   or  into  a  range  of  free  systems  typically  using  home-­‐brewed  or  no  markup  at  all.     The  few  good  materials  will  be  fantastically  expensive  to  make,  and  will  sit  in  their   own  interfaces  inaccessible  to  everyone  else;  the  many  poorer  materials  will  be   effectively  invisible,  and  of  limited  value.    Put  this  way,  this  sounds  crazy.    But  that  is   what  we  are  doing  now,  and  powerful  traditional  and  institutional  forces  are   pushing  us  to  continue  in  this  direction.    This  model  suits  the  traditional  way   scholars  think  about  what  we  do:  we  would  all  like  to  be  at  the  helm  of  one  of  these   privileged  editions,  or  to  be  running  a  digital  humanities  centre  which  hosts  them.     Just  to  give  a  single  instance  of  what  is  wrong  with  this  model:  the  Transcribe   Bentham  project  is  wonderful,  and  rightly  praised  for  many  things.    But  consider   this.    The  cost  of  Transcribe  Bentham  is  over  a  million  dollars,  just  reckoning  on  the   two  major  tranches  of  funding  it  has  received  to  now.    So  far  the  project  has   transcribed  around  5000  of  10000  manuscript  pages.    Now,  you  could  have  had   those  pages  transcribed  in  the  Phillipines  for  a  total  cost  of  around  $50,000  –  or,  if   you  really  wanted  to  have  everything  checked  and  thoroughly  controlled,  say   $100,000  at  most.  I  don’t  think  we  can  afford  many  more  Transcribe  Bentham   projects.     We  don’t  have  to  work  this  way.  We  might  have  many  more  editions  made  directly   by  scholars,  who  form  all  kinds  of  new  partnerships  with  readers  using  all  the  social   tools  now  available  to  us,  with  more  readers  themselves  becoming  editors.    There   should  be  an  unbroken  continuum,  between  the  multiple  resources  created  by   libraries  as  they  produce  and  put  online  more  and  more  images  of  the  manuscripts   and  books,  through  the  massive  online  records,  with  all  their  flawed  OCR,  made  by   Google  and  others  of  these  materials,  then  through  to  the  scholarly  editions  we   might  build  within  and  on  top  of  all  these  materials,  where  every  word  is  scrutinized,   every  decision  weighed,  and  where  our  aim  is  not  just  to  record,  not  just  to  present,   but  to  understand.  We  have  great  work  to  do,  and  there  are  ways  digital  humanists   and  textual  scholars  can  help  each  other.  However,  I  argue  that  we  need  to  look  at   just  where  we  are,  where  we  are  heading,  and  ask  ourselves:  is  there  a  better  way.    I   think  there  is,  and  I  hope  this  conference  starts  us  out  along  this  better  way.  I  think   we  should  begin,  as  textual  scholars,  by  declaring  exactly  what  it  is  we  do:  we  do   textual  scholarship.    We  may  use  digital  humanities  to  be  better  textual  scholars,  but   we  do  not  pretend  to  be  digital  humanists.    In  return,  digital  humanists  might  also   declare:  we  do  digital  humanities,  and  we  try  to  help  textual  scholars  to  be  better   textual  scholars  through  digital  humanities,  but  we  do  not  pretend  to  be  textual   scholars.