This  is  my  song  in  praise  of  humani1es  data,  of  primary  sources  and  their  digital

  surrogates.     “Humani1es”  and  “data”  are  two  terms  that  sit  uneasily  beside  one  another,  because   in  the  humani1es  we  deliberately  and  on  purpose  do  not  prac1ce  the  scien1fic   method  (unless  we  do).  What  is  “the  humani1es  method”?  There  isn’t  one.  We   reserve  the  right  to  change  our  method  with  our  mood,  as  befits  a  human  studying   humans  in  a  human  way.    

1  

If  there  is  a  humani1es  method,  it  could  conceivably  consist  of  this:  a  person,  alone,   reading.  Not  conduc1ng  experiments  or  studies:  just  reading.  And  then,  wri1ng.   Philosophy,  history,  and  the  study  of  any  language’s  literature  all  see  this  as  their   archetypal  method,  I  think,  though  not  archaeology  (which,  yes,  was  classified  as   belonging  to  the  humani1es  by  no  less  than  the  Na1onal  Endowment  for  the   Humani1es  in  1965,  the  year  of  its  founding).  That  is  why  it  is  a  truism  and  cliché  to   say  that  the  library  is  the  humani1es  laboratory.     And  what  we  read  and  how  and  how  much  and  how  quickly  is  changing,  changing   uSerly.  We  live  in  an  age  where  answers  are  as  easy  to  come  by  as  parking  spaces.  It   wasn’t  always  so.    

2  

On  Saturday,  November  17,  1860,  the  masthead  of  the  semi-­‐scholarly  London   periodical  Notes  and  Queries  described  itself,  as  it  had  for  the  last  thirteen  years,  as   “A  Medium  of  Inter-­‐Communica1on  for  Literary  Men,  Ar1sts,  An1quaries,   Geneaologists,  Etc.”    

3  

Below  the  masthead  were  adver1sements  of  recently  published  or  soon-­‐to-­‐be-­‐ published  books,  such  as  Carthage  and  Its  Remains.    

4  

An  ad  for  the  London  Library  boasted  of  80,000  volumes,  a  reading  room  “furnished   with  the  principal  Periodicals,  English,  French,  German,”  and  a  catalogue  that  could   be  purchased  for  only  nine  shillings  and  sixpence.  “ This  EXTENSIVE  LENDING  LIBRARY,   the  only  one  of  its  kind  in  London,”  was  open  from  10  to  6.    

5  

As  usual,  the  journal  was  an  ac1ve  bulle1n  board  of  ques1ons,  answers,  and   miscellaneous  contribu1ons  to  knowledge,  such  as  the  correc1on  to  Forster’s  Lives  of   Eminent  Statesmen  concerning  the  mistaken  iden1ty  of  one  Lord  Wentworth.    

6  

 As  for  queries,  “A  Constant  Reader  and  Subscriber”  asked  for  “an  authen1c  account   of  Sawney  Bene,  the  Scotch  cannibal,”  

7  

while  X.  Y.  wondered  if  anyone  could  tell  him  who  wrote  the  1830  tragedy  called   “Wismar,”  and  Saxon  asked,  “Can  you,  or  one  of  your  correspondents,  inform  me  by   whom  the  term  ‘God's  Acre,’  as  applied  to  a  churchyard,  was  first  used  in  English   literature?  It  appears  in  the  wri1ngs  of  Longfellow,  who  seems  to  have  adopted  it   from  the  German;  but  I  have  some  doubts  whether  it  had  not  been  previously  used   by  one  of  our  early  writers  —  George  Herbert  for  instance.”    

8  

Most  of  these  queries  are  easily  answered  today,  by  means  other  than  this  clever  one   of  asking  people.  There’s  a  Wikipedia  page  for  Sawney  Bean,  of  course,  which  avers   that  he  probably  never  existed  (although  the  legend  “is  part  of  the  Edinburgh  tourism   industry”).    

9  

The  Oxford  English  Dic1onary  suggests  that  George  Herbert  never  used  the  term   “God’s  Acre,”  though  it  had  appeared  a  few  1mes  at  least  in  the  17th  century,    

10  

and  a  Google  Book  Search  generally  confirms  this,  although  the  term  also  turns  up  in   the  1828  Harvard  Register  -­‐-­‐  a  possible  source  for  New  Englander  Longfellow  -­‐-­‐  as   well  as  in  a  few  other  interes1ng  sources.    

11  

The  ques1on  of  who  wrote  the  1830  drama  “Wismar:  A  Tragedy”  is  a  bit  harder,   however,  and  it  may  forever  remain  unanswered.    

12  

Notes  and  Queries  is  s1ll  published  today,  by  Oxford  Journals,  but  it  has  changed,  as   you  can  see  by  the  table  of  contents.      

13  

14  

15  

16  

17  

18  

19  

20  

Notes,  notes,  notes,  notes  .  .  .  and,  finally,  a  reader’s  query.    

21  

The  notes  in  themselves  have  changed  somewhat,  as  well:  they  read  more  s1ffly  to   me,  they  sound  more  professional,  more  academic,  less  personal,  even  considering   the  more  formal  Victorian  dic1on  in  which  an  1860  author  explained  that  a  mistake   he  had  made  took  place  in  “a  1me  of  great  domes1c  anxiety.”   I  some1mes  wonder  whether  Victorian  humanists  are  staring  longingly  down  at  us   from  heaven,  longing,  just  longing,  to  get  their  hands  on  our  research  tools.  Of   course,  then,  as  now,  not  all  humani1es  ques1ons  –  perhaps  not  even  most  –  could   be  answered  with  data,  informa1on,  facts,  research,  and  then,  as  now,  there  are   scrupulous  researchers  and  not-­‐so-­‐scrupulous  researchers,  which  makes  a  big   difference  no  maSer  what  tools  you  have  at  your  disposal.    

22  

The  Victorian  translator,  cri1c,  poet,  librarian,  and  honorary  M.A.  Edmund  Gosse,  for   instance,  was  a  notoriously  bad  researcher.  In  the  fall  of  1886,  in  what  Gosse’s   biographer  Ann  Thwaite  calls  “the  central  episode  of  Edmund  Gosse’s  literary  career,”   the  cri1c  John  Churton  Collins  aSacked  Gosse’s  literary  history  From  Shakespeare  to   Pope  for  its  unscholarliness  (277).  “We  have  even  refrained  from  discussing  maSers   of  opinion,”  wrote  Collins  in  a  widely-­‐read  Quarterly  Review  piece.  “We  have  confined   ourselves  en1rely  to  maSers  of  fact–to  gross  and  palpable  blunders,  to  unfounded   and  reckless  asser1ons,  to  such  absurdi1es  in  cri1cism  and  such  vices  of  style  as  will   in  the  eyes  of  discerning  readers  carry  with  them  their  own  condemna1on”  (qtd.  in   Thwaite  282).  Gosse  was  just  about  to  take  up  a  faculty  posi1on  of  Clark  Lecturer  at   Cambridge  when  the  denuncia1on  appeared.     In  the  comments  of  Gosse’s  biographer  on  the  episode,  we  get  a  portrait  of  another   kind  of  Victorian  researcher.  Thwaite  writes,  “ There  is  no  ques1on  that  Collins  was  a   fana1c  and  a  pedant.  Later  in  life  he  would  search  the  registers  of  forty-­‐two  Norwich   churches,  trying  to  pin  down  the  elusive  birth-­‐date  of  Robert  Greene  for  an  edi1on  he   was  edi1ng.  But,  as  far  as  Gosse’s  book  was  concerned,  Collins  happened  to  be  right… From  Shakespeare  to  Pope  is  full  of  extraordinary  mistakes”  (278).  Gosse’s  career  did   survive  the  blow–he  took  up  his  posi1on  as  scheduled–but  his  reputa1on  as  a  scholar   was  never  the  same.  During  the  scandal,  Henry  James  remarked  in  a  leSer  that  Gosse   “has  [emphasis  original]  a  genius  for  inaccuracy  which  makes  it  difficult  to  dress  his   wounds”  (qtd.  in  Thwaite  339).  

23  

Gosse’s  research  inep1tude  or  carelessness,  however,  probably  contributed  to  the   existence  some  great  poetry,  for  instance  Dylan  Thomas’s  “Do  not  go  gentle  into  that   good  night.”  How,  you  ask?    

24  

“Do  not  go  gentle  into  that  good  night”  is  a  villanelle,  a  19-­‐line  6-­‐stanza  alterna1ng-­‐ refrain  poe1c  form  with  only  two  rhymes  that,  with  a  lot  of  help  from  Gosse,  had  for   over  a  century  the  reputa1on  of  being  an  ancient  French  poe1c  form.  In  1877,  Gosse   published  an  ar1cle  in  the  Cornhill  Magazine  1tled  “A  Plea  for  Certain  Exo1c  Forms  of   Verse,”  in  which  he  explained  the  rules  of  six  ancient  (or  “ancient”)  French  forms  and   gave  examples,  wri1ng  them  himself  when  necessary.     In  the  ar1cle,  Gosse  reprinted  a  16th-­‐century  poem  1tled  “J’ay  perdu  ma  tourterelle”   by  the  French  poet  and  professor  of  La1n  Eloquence  Jean  Passerat,  liSle  realizing  that   his  “example”  was  in  fact  the  only  early  poem  in  that  form.  Gosse  did  write  (in  a   slightly  puzzled  tone),  “I  do  not  find  that  much  has  been  recorded  of  [the  villanelle's]   history,  but  it  dates  back  at  least  as  far  as  the  fiueenth  century”  (64).  AdmiSedly,  at   this  point  Gosse  had  done  liSle  worse  than  rely  on  a  mistaken  source,  Théodore  de   Banville’s  PeBt  traité  de  poésie  of  1872,  but  he  was  later  to  repeat  his  error,  with  less   excuse.    

25  

In  1879,  two  years  auer  “A  Plea  for  Certain  Exo1c  Forms  of  Verse,”  a  Parisian   bibliophile  and  poet  named  Joseph  Boulmier  published  a  book  of  villanelles  in  French   all  modeled  auer  Passerat’s  poem.  It  is  likely  that  Boulmier  owned  a  copy  of  the   1606  work  in  which  Passerat’s  “J’ay  perdu  ma  Tourterelle”  first  appeared;  he  certainly   seems  to  have  been  the  first  nineteenth-­‐century  admirer  of  the  villanelle  to  consult  it.   But  Boulmier  the  book  collector  did  more  than  consult  that  single  volume,  a  volume   Gosse  couldn’t  have  goSen  hold  of:  he  searched  through  everything  he  had,  and   came  to  the  correct  conclusion:    

26  

“One  fine  day,  auer  having  spoken  successively  of  the  rondeau,  of  the  triolet,  of  the   ballade,  of  the  lai,  of  the  virelai,  of  the  chant  royal,  the  author  of  I  no  longer  know   which  trea1se  on  versifica1on,  bungled  to  hell  like  they  almost  always  are,  finally   tackled  the  villanelle,  having  the  idea,  or  perhaps  the  luck,  to  cite  as  a  model  of  this   last  genre–and  auer  all  he  wasn’t  wrong–a  certain  naïve  masterpiece  escaped,  God   knows  how,  from  the  pen  of  the  scholar  Passerat….The  turtledove  of  Passerat  once   launched  into  circula1on,  what  happened  to  it?  All  the  trea1ses  on  versifica1on  that   succeeded  one  another  and  copied  one  another  in  single  file,  accompanying  this  or   that  grammar,  this  or  that  rhyming  dic1onary,  did  not  fail  to  drag  it  back  on  the  scene,   and  especially  to  present  it  as  a  type  from  which  it  was  absolutely  forbidden  to   depart….Well,  I  say  it  without  fear:  you  can,  as  I  have  done  myself,  page  through  all   the  essays  on  versifica1on  from  the  fiueenth  and  sixteenth  century,  one  auer   another;  you  will  not  find  there  the  least  trace  of  Passerat’s  turtledove,  which  is  to  say   nothing  that  resembles  this  lovely  form.”  

27  

In  his  entry  on  the  villanelle  for  the  monumental  1911  Encyclopaedia  Britannica,   Gosse  (by  then  supposedly  a  wise  elder)  hardly  retreated  from  the  asser1ons  he  had   made  over  thirty  years  earlier.  Ci1ng  Boulmier,  Gosse  conceded  that  there  were  no   schema1c  double-­‐refrain  villanelles  before  Passerat,  yet  (like  Boulmier  himself)  he  did   not  conclude  that  it  was  he  and  his  contemporaries  who  were  responsible  for   defining  the  modern  form  of  the  villanelle  in  the  nineteenth  century:   “VILLANELLE,  a  form  of  verse,  originally  loose  in  construc1on,  but  since  the  16th   century  bound  in  exact  limits  of  an  arbitrary  kind.  .  .  .  It  appears,  indeed,  to  have  been   by  an  accident  that  the  special  and  rigorously  defined  form  of  the  villanelle  was   invented.  In  the  posthumous  poems  of  Jean  Passerat  (1534-­‐1602),  which  were   printed  in  1606,  several  villanelles  were  discovered,  in  different  forms.  One  of  these   became,  and  has  remained,  so  deservedly  popular,  that  it  has  given  its  exact  character   to  the  subsequent  history  of  the  villanelle.”   Gosse’s  plea,  you  see,  had  been  successful,  and  because  of  his  influence  there  had   been  a  small  villanelle  vogue  in  England  among  the  Parnassians  at  the  end  of  the   nineteenth  century.  James  Joyce,  eighteen  years  old  in  1900,  played  along,  and  later   reprinted  a  piece  of  his  poe1c  juvenilia  in  1914’s  Portrait  of  the  ArBst  as  a  Young  Man.   From  there,  and  helped  along  by  poetry  handbooks  quo1ng  one  another  in  single  file,   the  villanelle  became  entrenched  in  English  poetry  with  a  reputa1on  as  an  ancient   French  form,  leading  not  only  to  “Do  not  go  gentle  into  that  good  night”  but  also  to   Elizabeth  Bishop’s  “One  Art”  (recited  in  the  Cameron  Diaz  flick  In  Her  Shoes  –  “the  art   of  losing  isn’t  hard  to  master.”)  

28  

When  I  first  conducted  the  research  on  the  villanelle  that  led  to  this  tale  of  good  (but  ignored)  and  bungled  (but   influen1al)  research,  I  took  it  upon  myself  to  find  the  text  or  texts  that  caused  Banville  (Gosse’s  1872  source)  to   believe  that  the  villanelle  was  an  ancient  French  form.  Banville  had  begun  wri1ng  villanelles  himself  in  1845,  so  I   began  to  search  for  any  and  all  French  poetry  handbooks  and  anthologies  published  between  1606  and  1845,  with   special  aSen1on  to  early  nineteenth-­‐century  works  that  Banville  would  probably  have  had  to  hand.  I  worried  that  I   might  have  to  search  for  works  in  other  languages,  as  well,  but  it  was  surely  best  to  begin  with  works  in  French.     My  chief  resource  in  compiling  the  list  of  1tles  was  WorldCat,  which  I  regularly  plied  with  various  "poe*"  strings.   From  my  carrel  in  the  stacks  of  Alderman  library  at  the  University  of  Virginia,  I  began  to  make  forays  into  the  stacks   from  which  I  would  return  with  armfuls  of  books  that  I  would  then  page  through,  just  like  Boulmier  (how  much   had  changed  since  1879?),  looking  for  men1ons  of  the  villanelle  form  or  of  Passerat  or  of  "J'ay  perdu  ma   Tourterelle,"  and  looking  for  other  poetry  books  to  gather  or  to  order  from  Interlibrary  Loan.  Whenever  I  visited   the  shelves,  I  would  also  scan  the  proximate  volumes  and,  more  ouen  than  not,  scoop  them  up  to  take  back  to  my   carrel  -­‐-­‐  ouen,  I'm  sorry  to  say,  without  checking  them  out.  I  remember  that  it  was  a  week  or  so  into  this  process   that  I  discovered  a  1986  Slatkine  reprint  of  an  1844  work  by  an  author  named  Wilhelm  Ténint.  Standing  at  the   shelf,  I  paged  through  un1l  I  found  an  entry  that  both  cited  Passerat  and  claimed  that  the  villanelle  was  an  old   fixed  form.  Siegel  also  men1oned  that  Banville  himself  had  made  marginal  notes  on  the  manuscript  of  Ténint’s   Prosodie.     Remember,  now,  the  year  was  2003.  I  had  not  only  the  well-­‐stocked  stacks  of  an  excellent  research  library  at  my   disposal,  but  also  Google,  and  also  the  WorldCat  database.  Google  -­‐-­‐  the  regular  search  engine,  mind  you,  not   Google  Books  -­‐-­‐  gave  me  a  few  par1cularly  good  leads  at  other  points  in  my  research.  Auer  the  Ténint  discovery,  I   con1nued  to  look  for  other  men1ons  of  the  villanelle  form  in  early  19th-­‐century  French  texts,  but  I  found  very   liSle,  almost  nothing.     Flash  forward  five  years,  only  five  years,  and  imagine  me  now,  if  you  will,  engaged  in  co-­‐wri1ng  a  new  entry  on  the   villanelle  for  the  forthcoming  revised  edi1on  of  the  Princeton  Encyclopedia  of  Poetry  and  PoeBcs,  edited  by   Stephen  Cushman,  my  disserta1on  advisor.  This,  obviously,  was  our  big  chance  to  correct  the  record  about  the   villanelle  in  the  gold  standard  of  poetry  handbooks.  And  so  I  revisited  my  search  for  men1ons  of  the  villanelle  and   of  "J'ay  perdu  ma  Tourterelle"  between  1606  and  1845,  and  this  1me  I  used  Google  Books.    

29  

30  

Using  Google  Book  Search,  I  found  38  texts  that  might  have  influenced  Ténint,  38   more  sources  in  that  trail  of  textual  transmission,  more  evidence  of  what  was  known   and  thought  about  the  villanelle  in  that  fragile  1me  when  a  mistake  that  would   engrave  itself  in  the  record  for  more  than  a  century  was  just  beginning  to  flap  its   delicate  buSerfly  wings.  I  didn't  find  anything  that  directly  contradicted  my  claim  that   the  Ténint  work  can  be  considered  the  chief  entry  point  of  the  villanelle  error,  but   what  I  did  find  were  numerous  texts  that  smoothed  its  way.  To  sa1sfy  my  conscience,   I  included  two  of  the  more  popular  dic1onaries  and  encyclopedias  that  Google  Book   Search  turned  up  for  me  in  the  PEPP  entry.     I’m  not  sure  I  can  convey  properly  through  this  somewhat  procedural  narra1ve  the   thrill  I  felt  at  finding  the  Ténint  source  by  siuing  through  dozens  of  books  with  my   bare  hands,  and  the  dismay  I  felt  at  finding  ( just  a  liSle  too  late)  thirty-­‐eight  addi1onal   sources  by  siuing  through  millions  of  books  with  Google  Book  Search.   So  much  more  data,  so  suddenly.    

31  

That  “data  dismay”  is  something  researchers  have  always  felt,  of  course.  Witness   Virginia  Woolf’s  descrip1on  of  a  trip  to  the  Library  of  the  Bri1sh  Museum,  feeling  as   though  she  would  “need  claws  of  steel  and  beak  of  brass  even  to  penetrate  the  husk”   of  all  her  data,  as  though  she  were  some  kind  of  steampunk  clockwork  woodpecker.    

32  

One  of  the  chief  aims  of  digital  humanists  since  the  90s  has  always  been  simply  to  get   more  stuff  online,  preferably  in  a  scholarly  way.  We’re  just  now,  especially  but  not   exclusively  with  text,  beginning  to  say,  Okay,  we’ve  put  a  lot  of  stuff  online.  Our   primary  sources,  our  data,  are  now  digital.  Google  has  put  a  lot  online,  TwiSer  has   put  a  lot  online,  humanity  has  put  a  lot  online.  Now  what  do  we  do  with  it?     In  2009,  the  Na1onal  Endowment  for  the  Humani1es’s  Office  of  Digital  Humani1es   put  that  ques1on  to  researchers,  but  almost  as  a  dare.  What  can  you  do  with  all  that   data,  they  asked.  Show  us.  The  Digging  into  the  Enlightenment  project,  for  interest,   will  look  at  53,000  18th-­‐century  leSers.    

33  

The  Digging  into  Data  project  is  only  part  of  a  larger  trend,  some1mes  called  “distant   reading,”  in  a  term  taken  from  Franco  More{’s  Graphs,  Maps,  Trees,  shown  here  on   the  social  reading  site  GoodReads.    

34  

Examples  of  distant  reading  include  some  of  the  work  done  with  text  mining,  analysis,   and  visualiza1on  tools  such  as  the  MONK  project,  described  in  the  2008  ar1cle  “How   Not  to  Read  a  Million  Books.”  

35  

Tanya  Clement’s  work  with  Gertrude  Stein’s  Making  of  America  is  interes1ng  not  only   for  its  conclusion,  which  is  that  the  text  has  a  decided  structure  and  paSern  that  is   not  apparent  to  a  human  reader,  but  for  its  stated  premise:  that  the  work  is   unreadable  by  humans.  (A  neutral  observa1on,  not  an  aesthe1c  judgment.)  

36  

But  “distant  reading”  need  not,  perhaps,  entail  sta1s1cs  and  machines.  Speaking  with   a  journalist  at  the  New  York  Times  about  his  book  How  to  Talk  About  Books  You   Haven’t  Read,  Pierre  Bayard  described  some  very  human  and  qualita1ve  and   incomplete  and  yet  s1ll  valuable  modes  of  distant  reading:    

37  

“We  are  taught  only  one  way  of  reading,”  he  said.  “Students  are  told  to  read  the   book,  then  to  fill  out  a  form  detailing  everything  they  have  read.  It’s  a  linear  approach   that  serves  to  enshrine  books.  People  now  come  up  to  me  to  describe  the  cultural   wounds  they  suffered  at  school.  ‘You  have  to  read  all  of  Proust.’  They  were   trauma1zed.”     “They  see  culture  as  a  huge  wall,  as  a  terrifying  specter  of  ‘knowledge,’  “  he  went  on.   “But  we  intellectuals,  who  are  avid  readers,  know  there  are  many  ways  of  reading  a   book.  You  can  skim  it,  you  can  start  and  not  finish  it,  you  can  look  at  the  index.  You   learn  to  live  with  a  book.”    

38  

I  think  perhaps  that  large  sets  of  humani1es  data,  like  books,  can  be  read  in  the  way   Bayard  describes:  not  comprehensively,  but  by  living  with  them.  Their  sheer  size   suggests  but  need  not  entail  sta1s1cal  analysis  and  visual  display.  We  can  browse  very   large  humani1es  datasets,  skim  them,  live  with  them,  instead  of  reading  them  in  a   linear  fashion  with  computers.  Auer  all,  how  likely  is  it  that  the  database  itself  is   comprehensive?  Isn’t  it  very  likely  itself  simply  a  sample?  The  Reading  Experience   Database,  for  instance,  itself  a  compelling  example  of  a  very  large  humani1es  dataset,   admits  quite  charmingly  that  it  can  never  be  comprehensive:    

39  

“While  RED  may  never  be  the  comprehensive  database  that  would  allow  us  to  make   rigorously  sta1s1cal  arguments  for  reading  habits  in  given  places  or  1me  periods,  it   can  func1on  as  a  source  of  compelling  examples.  The  more  entries  that  go  in  it,  the   more  it  can  approach  the  ideal,  but  it  can  never  hope  to  be  a  comprehensive   database  of  every  archive,  every  annotated  page,  every  diary  manuscript,  in  the   Bri1sh  World,  1450-­‐1945,  much  and  all  as  we  may  want  it  to!”    

40  

Like  our  datasets,  our  methods  need  not  be  comprehensive.  Lately  I’ve  been   interested  in  the  possibili1es  of  what  I  think  of  as  “datum  love”:  the  selec1on   (random,  serendipitous,  affec1onate)  of  compelling  examples.  In  the  Reading   Experience  Database,  for  instance,  some  idling  through  the  byways  turns  up  the   interes1ng  fact  that  Dickens  was  once  at  least  read  by  “a  revolu1onary  Russian  rag   merchant.”  Isn’t  that  1dbit  a  spur  to  further  inquiry?  In  my  experience,  faced  with  the   “Fordist,  func1onalist”  impera1ve  to  write  that  Kathleen  men1oned  yesterday,   humani1es  scholars  of  any  rank  generally  begin  with  a  text,  a  topic,  a  theory,  or  a  text   and  a  topic  and  a  theory,  and  we  proceed  on  the  assump1on  that  we  must  produce   an  original  interpreta1on  or  argument.  What  I  wonder  is  whether  instead  we  can   begin  with  the  data,  or  with  a  datum,  and  simply  watch  for  what  it  may  tell  us,  even  if   what  it  tells  us  is  simply  a  story.  What  I  hope  is  that  all  our  data  will  bring  forth  a  new   age  of  humanis1c  induc1on,  induc1on  that  can  but  need  not  necessarily  rely  on   sta1s1cs  and  visualiza1ons.  

41  

And  what  I  hope,  too,  is  that  more  compilers  of  databases  will  recognize  that  they  are   at  least  as  well-­‐fiSed  as  anyone  to  tell  us  what  the  data  can  tell  us.  Archivists  and   librarians,  especially,  know  the  data,  because  they  feed  and  groom  it.  Tim  SherraS,   for  instance,  is  an  archivist,  historian,  and  programmer  in  Australia  who  has  recently   begun  a  project  called  Invisible  Australians.  When  he  worked  for  the  Na1onal   Archives  of  Australia,  SherraS  no1ced  that  there  were  a  great  many  print  records  that   could  be  converted  into  structured  data.    

42  

One  such  type  of  print  record  is  the  “Cer1ficate  Exemp1ng  from  Dicta1on  Test,”  or   CEDT.  The  CEDT  was  a  bureaucra1c  outgrowth  of  the  White  Australia  Policy,  which   restricted  non-­‐white  immigra1on  to  Australia  from  1901  to  1973:  it  was  a  form  that   enabled  exis1ng  non-­‐white  residents  of  Australia  to  leave  and  re-­‐enter  the  country   without  being  mistaken  for  immigrants.  Over  50,000  paper  CEDTs  reside  in  the   Na1onal  Archives  of  Australia,  and  these  forms  have  a  great  deal  to  tell  history  about   some  of  the  people  on  the  margins  of  history.  Obviously,  they  could  tell  more  if  their   data  were  digital:  enter  the  Invisible  Australians  project.  

43  

Choosing  a  CEDT  subject  surely  almost  at  random,  SherraS  narrates  some  of  the  life   of  Charlie  Allen,  a  half-­‐Chinese  man:     “Charlie  was  born  in  Sydney  in  1896.  His  mother  was  Frances  Allen  (some1me  sweet   shop  owner  and  brothel  keeper),  his  father  Charlie  Gum  (a  buyer  for  Wing  On   company).  Charlie  was  raised  by  his  mother,  but  in  1909,  at  the  age  of  thirteen,  he   was  taken  to  China  by  his  father.  His  father  returned  to  Sydney,  leaving  Charlie  in   China.  He  lived  with  rela1ves  in  the  town  of  Shekki  (inland  from  Hong  Kong)  for  six   years.  Charlie  was  homesick,  but  had  no  means  of  ge{ng  back  to  Australia.  His   mother  aSempted  to  enlist  government  help  but  to  no  avail.  Charlie  finally  returned   in  1915.  The  following  year  he  enlisted  in  First  AIF  (well,  actually  he  enlisted  three   1mes,  and  was  discharged  as  medically  unfit  each  1me).  Charlie  married  in  Sydney  in   1917  and  had  two  daughters  soon  auer.  He  returned  to  China  in  1922  for  seven   months.  Charlie  Allen  died  in  1938  as  the  result  of  an  industrial  accident.  He  was   forty-­‐one.”   To  my  mind,  SherraS  is  nearly  the  ideal  digital  humanist,  not  only  because  he  is  a   builder  of  databases,  but  because  his  ins1nct,  once  he  has  built  a  database,  is  to  use   it  to  tell  stories.  Few  or  no  graphs,  maps,  and  trees  for  him.    

44  

If  you  would  become  a  digital  humanist,  then,  what  I  would  encourage  you  to  do  is  to   go  and  find  compelling  examples.  Browse.  Play.  Observe.  Induce.  Roll  around  in  the   data.  Then  tell  me  which  pieces  stuck  to  you.  I  for  one  will  be  fascinated.    

45