Eat, Drink and Be Networked: Feasting and Bronze Age Networks

Colloquium on Digital History and the Transnational History of Nursing
Department of History, Carleton University - Ottawa

Transcript

1.   Colloquium  on  Digital  History  and  the  Transna6onal   History  of  Nursing     Eat,  Drink  and  Be  Networked:   Feas4ng  and  Bronze  Age  Networks       Zack  Ba6st    Directed  Interdisciplinary  Studies       Faculty  Supervisor,  Dr.  Shawn  Graham     March  23,  2012      
2. Project  Goals   •  To  learn  how  design,  implement  and  interpret   the  results  of  projects  that  u6lize  digital   components  and  methodologies   •  To  explore  the  use  of  network  analysis  in  an   archaeological  seNng   •  To  study  how  the  consump6on  of  intoxica6ng   substances  contributed  to  the  social   stra6fica6on  of  early  socie6es   •  This  was  accomplished  by  studying  the   distribu6on  of  poQery  rela6ng  to  feas6ng   across  the  Bronze  Age  Aegean  
3. The  Emergence  of  a     Social  Hierarchy   •  The  presence  of  luxury  vessels  that  were  reserved  for  the   ac6vi6es  of  the  wealthy  would  signify  the  presence  of  an  upper   class,  who  mobilized  resources  and  oversaw  the  centralized   economies  of  this  seNng   •  The  middle/late  Bronze  Age  was  a  period  of  transi6on  from  a   more-­‐or-­‐less  egalitarian  society  to  an  increasingly  hierarchical   chiefdom  structure  with  an  elite  class.   •  Feas6ng  is  an  interes6ng  example  of  conspicuous  consump6on,   through  which  the  leader  reinforced  his  leadership  and  links  to   the  his  allies,  while  also  emphasizing  his  dis6nc6on  from  the  rest   of  the  community.  
4. Na4onal  Archaeological  Museum,  Athens  
5. Network  Analysis  in  Archaeology  and  the  Social   Sciences   •  Network  analysis  is  a  method  to  find  rela6onships   between  en66es  that  are  not  plainly  obvious.     •  This  set  of  methods  is  especially  useful  in  archaeology,   since  the  accumula6on  of  intertwined  data  is  difficult   to  analyze  and  interpret   •  It  provides  a  systema6c  approach  to  examine  social   rela6onships  in  a  quan6ta6ve  way.     •  Network  analysis  can  be  used  to  examine  the   rela6onships  between  any  kinds  of  variables   •  In  the  social  sciences,  interpreta6on  requires  the   considera6on  of  the  nature  of  what  nodes  actually   represent   •  When  working  with  objects,  they  must  mean   something  to  the  people  who  used  them  
6. Dataset   •  Individual  vessels  were  recorded  from  excava6on  reports   – Problems  included  uneven  depths  of  excava6ons,   limited  access  to  reports   •  Total  of  5669  vessels  were  recorded   – Of  them,  2995  vessels  were  included  in  the  analysis.     •  Ten  sites  were  included  in  the  analysis   •  Varia6ons  within  a  poQery  type  were  ‘lumped’  together   •  Luxury  poQery  is  easily  classifiable  by  func6on  to  the   ac6vi6es  of  the  elite  within  a  hierarchal  society.    
7. Computa4onal  Methods     •  Gephi  –  www.gephi.org.     •  Metrics  used:     •  Degree  -­‐  The  degree  represents  the  number  of  connec6ons  that  a  par6cular   node  is  directly  associated  with.     •  Betweeness  Centrality  -­‐  The  betweeness  centrality  is  the  measure  of  how   oaen  a  par6cular  node  acts  as  an  intermediary  between  the  paths  of  any  two   other  nodes  in  a  given  network.  This  is  usually  expressed  as  an  index  value.     •  Modularity  -­‐  This  metric  iden6fies  small  sub-­‐communi6es  of  nodes  within  the   overall  network.  Densely  packed  groupings  are  oaen  connected  with  less   dense  intermediaries.  In  social  network  analysis,  the  iden6fica6on  of  these   sub-­‐communi6es  oaen  reflect  real-­‐world  applica6ons.       Ulrik  Brandes,  “A  Faster  Algorithm  for  Betweenness  Centrality,”  The  Journal  of  Mathema3cal  Sociology  25,  no.  2  (2001):  163–177.   Vincent  D  Blondel  et  al.,  “Fast  Unfolding  of  Communi6es  in  Large  Networks,”  Journal  of  Sta3s3cal  Mechanics:  Theory  and  Experiment   2008,  no.  10  (October  9,  2008):  P10008.    
8. Ariadne  Algorithm   •  Developed  by  Evans,  Rivers  and  Knappet   •  Applied  to  34  Bronze  Age  loca6ons  of  the  Aegean   •  Included  factors  that  ‘push’  or  ‘pull’  people  to  travel  to   certain  sites   –  Access  to  resources,  popula6on  dynamics,  carrying  capacity   •  Also  incorporates  wind  and  sea  currents,  sailing  technology   and  physical  loca6on   •  Tim  Evans,  Ray  Rivers  and  Carl  KnappeQ,  “Interac6ons  in  Space  for  Archaeological  Models,”   2011   •  Carl  KnappeQ,  Tim  Evans,  and  Ray  Rivers,  “Modelling  Mari6me  Interac6on  in  the  Aegean   Bronze  Age,”  An3quity  82,  no.  318  (2008):  1009–1024.  
9. Two-­‐Mode  Network  (PoJery  –  Sites)  
10. One-­‐Mode  Network  (PoJery)   •   Displays  the  rela6onships   between  luxury  poQery  types,   based  on  their  co-­‐presence  at   various  archaeological  sites   •   Two  separate  modules  are   displayed   •   Alabastron  &  S6rrup  jar  have   high  centrality  values  and  link   these  dis6nct  communi6es  
11. One-­‐Mode  Network  (Sites)   •   Displays  the  rela6onships   between  archaeological   sites  based  on  which  luxury   poQery  types  were  found  at   each  one   •   No  luxury  vessels  at   Amorgos,  according  to  this   dataset    
12. Some  Key  Lessons  Concerning  Network   Analysis  in  the  Social  Sciences   •  It  is  very  important  to  establish  goals  early  on   •  Data  collec6on  must  suit  these  goals   •  Good  quality  data  is  crucial,  since  the  analysis  and   interpreta6on  are  dependent  on  it   •  If  the  interpreta6ons  are  of  social  interac6ons,  the  dataset   must  be  representa6ve  of  this   –  Ar6facts  must  mean  something  to  the  people  who  used  them   •  The  digital  humani6es  are  about  extending  the  reach  of   tradi6onal  methods  of  academic  research  by  using   technology,  working  in  an  interdisciplinary  environment,   facilita6ng  collabora6on  with  other  scholars,  and  the   publica6on  of  results  in  an  open  and  accessible  manner  
13. Thank  You!   Merci!