/Users/andrea/_magisterarbeit/korpus/clean/testkorpus/1/file7.html NN ----------------------------------------- : Complementary JJ strategies NNS . SENT why WRB we PP use VVP our PP$ hands NNS when WRB we PP think VVP David NP Kirsh NP Department NP of IN Cognitive JJ Science NP University NP of IN California NP , , San NP Diego NP kirsh NP cogsci NP . SENT ucsd NN . SENT edu NN ABSTRACT JJ A DT complementary JJ strategy NN can MD be VB defined VVN as IN any DT organizing VVG activity NN which WDT recruits VVZ external JJ elements NNS to TO reduce VV cognitive JJ loads NNS . SENT Typical JJ organizing NN activities NNS include VVP pointing VVG , , arranging VVG the DT position NN and CC orientation NN of IN nearby JJ objects NNS , , writing VVG things NNS down RP , , manipulating VVG counters NNS , , rulers NNS or CC other JJ artifacts NNS that WDT can MD encode VV the DT state NN of IN a DT process NN or CC simplify VV perception NN . SENT To TO illustrate VV the DT idea NN of IN a DT complementary JJ strategy NN , , a DT simple JJ experiment NN was VBD performed VVN in IN which WDT subjects NNS were VBD asked VVN to TO determine VV the DT dollar NN value NN of IN collections NNS of IN coins NNS . SENT In IN the DT no DT hands NNS condition NN , , subjects NNS were VBD not RB allowed VVN to TO touch VV the DT coin NN images NNS or CC to TO move VV their PP$ hands NNS in IN any DT way NN . SENT In IN the DT hands NNS condition NN , , they PP were VBD allowed VVN to TO use VV their PP$ hands NNS and CC fingers NNS however RB they PP liked VVD . SENT Significant JJ improvements NNS in IN time NN and CC number NN of IN errors NNS were VBD observed VVN when WRB S's NP used VVD their PP$ hands NNS over IN when WRB they PP did VVD not RB . SENT To TO explain VV these DT facts NNS , , a DT brief JJ account NN of IN some DT commonly RB observed VVN complementary JJ strategies NNS is VBZ presented VVN , , and CC an DT account NN of IN their PP$ potential JJ benefits NNS to TO perception NN , , memory NN and CC attention NN . SENT Introduction NN A DT complementary JJ strategy NN can MD be VB defined VVN as IN any DT organizing VVG activity NN which WDT recruits VVZ external JJ elements NNS to TO reduce VV cognitive JJ loads NNS . SENT The DT external JJ elements NNS may MD be VB our PP$ fingers NNS or CC hands NNS , , pencil NN and CC paper NN , , movable JJ icons NNS , , counters NNS , , measuring VVG devices NNS , , or CC other JJ entities NNS in IN our PP$ immediate JJ environment NN . SENT Typical JJ organizing NN activities NNS include VVP pointing VVG , , arranging VVG the DT position NN and CC orientation NN of IN nearby JJ objects NNS , , Kirsh NP , , 95 CD , , writing VVG things NNS down RP , , manipulating VVG counters NNS , , rulers NNS or CC other JJ artifacts NNS that WDT can MD encode VV the DT state NN of IN a DT process NN or CC simplify VV perception NN . SENT An DT obvious JJ example NN of IN a DT complementary JJ strategy NN is VBZ using VVG pencil NN and CC paper NN to TO help VV add VV a DT list NN of IN several JJ two CD and CC three CD digit NN numbers NNS . SENT Most JJS of IN us PP find VVP it PP easier JJR , , faster RBR and CC more RBR reliable JJ to TO write VV down RP incremental JJ sums NNS , , and CC carry VV overs NNS , , than IN to TO do VV the DT summing VVG entirely RB in IN our PP$ heads NNS . SENT For IN long JJ lists NNS , , we PP tend VVP , , as RB well RB , , to TO recruit VV the DT pencil NN itself PP as IN a DT pointer NN to TO help VV keep VV our PP$ place NN . SENT Each DT of IN these DT actions NNS has VHZ its PP$ cognitive JJ benefit NN . SENT By IN writing VVG down RP numbers NNS we PP offload VV that DT portion NN of IN working VVG memory NN required VVN to TO store VV intermediate JJ results NNS , , by IN pointing VVG to TO particular JJ numerals NNS we PP help VVP direct JJ attention NN and CC offload VV that DT portion NN of IN working VVG memory NN required VVN to TO store VV knowledge NN of IN location NN , , and CC by IN recording NN carry VVP overs NNS we PP set VVD up RP the DT environment NN to TO simplify VV verification NN of IN our PP$ sum NN , , should MD we PP desire VVP to TO redo VV part NN of IN it PP . SENT In IN my PP$ terminology NN , , such JJ actions NNS complement VV the DT internal JJ processes NNS occurring VVG when WRB we PP add VVP . SENT They PP are VBP external JJ components NNS in IN an DT interactive JJ computation NN . SENT Hutchins NP , , 95 CD . SENT It PP is VBZ certainly RB no DT new JJ claim NN to TO argue VV that IN , , as IN intelligent JJ creatures NNS , , we PP have VHP techniques NNS for IN altering VVG our PP$ environment NN to TO enhance VV our PP$ cognitive JJ performance NN . SENT Cognitive JJ anthropologists NNS , , and CC situated VVN activity NN theorists NNS have VHP long RB discussed VVN some DT of IN the DT ways NNS we PP have VHP of IN changing VVG our PP$ environment NN to TO augment VV cognition NN . SENT Lave VV 88 CD . SENT Typically RB , , however RB , , the DT changes NNS discussed VVN are VBP cultural JJ , , they PP arise VVP when WRB new JJ technologies NNS are VBP introduced VVN , , or CC when WRB we PP learn VVP new JJ facts NNS , , methods NNS and CC concepts NNS . SENT They PP take VVP days NNS or CC weeks NNS or CC years NNS to TO evolve VV , , and CC they PP involve VVP sharing VVG resources NNS and CC frequently RB cooperating VVG with IN others NNS . SENT Moreover RB , , they PP are VBP rarely RB studied VVN experimentally RB . SENT Kirsh NP Maglio NP , , 94 CD The DT environmental JJ adaptations NNS I PP shall MD focus VV on IN , , however RB , , occur VVP moment NN by IN moment NN as IN we PP manage VVP our PP$ workspaces NNS . SENT They PP are VBP usually RB quick JJ to TO set VV up RP , , and CC their PP$ effect NN is VBZ brief JJ , , measured VVN in IN seconds NNS or CC fractions NNS of IN seconds NNS , , . SENT Moreover RB , , these DT strategies NNS are VBP often RB acquired VVN quickly RB , , as RB when WRB , , for IN instance NN , , in IN the DT course NN of IN an DT activity NN , , we PP discover VVP the DT value NN of IN pointing VVG , , or CC laying VVG down RP a DT ruler NN , , etc FW . SENT We PP often RB learn VVP these DT by IN ourselves PP , , and CC our PP$ improved JJ performance NN can MD be VB studied VVN both CC analytically RB and CC experimentally RB . SENT In IN this DT paper NN I PP explore VVP one CD pervasive JJ example NN of IN such JJ complementary JJ strategies NNS . SENT using VVG our PP$ hands NNS to TO help VV think VV , , remember VV and CC perceive VVP . SENT After IN briefly RB elaborating VVG the DT central JJ idea NN , , I PP introduce VV a DT pilot NN experiment NN to TO explore VV a DT few JJ of IN the DT functions NNS served VVN by IN pointing VVG and CC related VVN hand NN movements NNS . SENT I PP conclude VVP with IN a DT short JJ account NN of IN some DT of IN the DT principles NNS underpinning VVG complementary JJ strategies NNS . SENT Complementary JJ strategies NNS Imagine VV being VBG shown VVN an DT upside NN down IN photograph NN and CC asked VVD to TO identify VV the DT person NN depicted VVN . SENT Your PP$ natural JJ action NN is VBZ to TO reach VV out RP and CC turn VV the DT picture NN right NN side NN up RB . SENT Faces NNS are VBP more RBR readily RB recognized VVN when WRB upright JJ . SENT Apparently RB , , to TO facilitate VV perception NN , , we PP perform VVP an DT action NN that WDT adapts VVZ the DT world NN to TO our PP$ perceptual JJ capacities NNS . SENT This DT idea NN that IN sometimes RB the DT best JJS way NN to TO solve VV a DT cognitive JJ problem NN is VBZ by IN adapting VVG the DT world NN rather RB than IN adapting VVG oneself PP lies VVZ at IN the DT heart NN of IN complementary JJ strategies NNS . SENT I PP believe VVP we PP learn VVP these DT adaptational JJ strategies NNS by IN the DT thousands NNS . SENT For IN example NN , , if IN an DT agent NN were VBD given VVN the DT task NN of IN memorizing VVG the DT letters NNS of IN a DT string NN , , such JJ as IN QIUYOKJHUYTOGU NP , , first JJ without IN touching VVG the DT letters NNS , , then RB with IN touch NN and CC re NN arrangement NN allowed VVN , , it PP is VBZ likely JJ that IN he PP or CC she PP would MD discover VV a DT method NN of IN moving VVG the DT letters NNS to TO reliably RB increase VV performance NN . SENT One CD such JJ letter NN moving VVG technique NN would MD be VB to TO shift VV the DT letters NNS into IN groupings NNS , , such JJ as IN QIU NP YOK NP JHU NP Y NP TO TO GU NP . SENT Another DT , , more RBR radical JJ technique NN , , would MD be VB to TO re NN order NN the DT letters NNS in IN alphabetical JJ order NN , , such JJ as IN GHIJK NP OO NP Q NP T NN UUU NP YY NP . SENT In IN any DT such JJ activity NN there EX is VBZ a DT trade NN off RP . SENT the DT cost NN in IN time NN and CC effort NN to TO perform VV the DT complementary JJ activity NN in IN the DT world NN vs NP . SENT the DT time NN and CC effort NN to TO use VV existing VVG mental JJ procedures NNS and CC strategies NNS to TO accomplish VV the DT task NN without IN external JJ aid NN . SENT Kirsh NP , , 95 CD . SENT More JJR factors NNS are VBP involved VVN in IN the DT choice NN of IN a DT complementary JJ strategy NN than IN just RB speed NN , , however RB . SENT In IN addition NN to TO potentially RB faster JJR performance NN the DT virtue NN of IN such JJ strategies NNS is VBZ that IN by IN changing VVG the DT local JJ environment NN at IN the DT right JJ time NN and CC in IN the DT right JJ way NN agents NNS are VBP able JJ to TO reduce VV the DT probable JJ error NN rate NN , , to TO cope VV with IN larger JJR , , more RBR complex JJ problems NNS , , and CC to TO deal VV with IN interference NN more RBR successfully RB all DT typical JJ measures NNS of IN performance NN , , and CC indicators NNS of IN the DT cognitive JJ demands NNS a DT task NN imposes VVZ . SENT Complementary JJ strategies NNS , , therefore RB , , allow VV agents NNS to TO compensate VV for IN resource NN limitations NNS in IN working VVG memory NN and CC processing NN power NN , , and CC cognitive JJ limitations NNS in IN categorizing VVG skill NN , , and CC so RB on IN . SENT Backman NP et NP al NP , , 92 CD . SENT The DT objective NN of IN research NN on IN complementary JJ strategies NNS is VBZ to TO expose VV the DT ubiquity NN of IN these DT strategies NNS particularly RB those DT that WDT are VBP spontaneously RB displayed VVN by IN subjects NNS to TO describe VV the DT key JJ trade NN offs NNS , , such JJ as IN speed NN accuracy NN , , speed NN problem NN size NN , , speed NN robustness NN , , and CC to TO explain VV these DT in IN terms NNS of IN an DT underlying JJ processing NN account NN describing VVG the DT way NN mental JJ resources NNS are VBP used VVN . SENT If IN complementary JJ strategies NNS are VBP pervasive JJ , , there EX ought MD to TO be VB general JJ principles NNS governing VVG the DT shape NN of IN trade NN off IN curves NNS for IN successful JJ complementation NN strategies NNS . SENT A DT Simple JJ Coin NP Counting VVG Experiment NN To TO observe VV how WRB complementary JJ strategies NNS enhance VV performance NN , , a DT simple JJ pilot NN experiment NN was VBD performed VVN . SENT Three CD male NN and CC two CD female JJ subjects NNS age NN 23 CD 38 CD , , mean JJ 26 CD were VBD shown VVN two CD sets NNS of IN 30 CD images NNS , , each DT depicting VVG a DT different JJ arrangement NN of IN quarters NNS , , dimes NNS and CC nickels NNS . SENT Their PP$ task NN was VBD to TO determine VV the DT dollar NN and CC cents NNS amount NN present NN . SENT See VV figure NN 1 CD . SENT In IN condition NN one CD , , the DT no DT hands NNS condition NN , , subjects NNS were VBD told VVN not RB to TO point VV at IN the DT coin NN images NNS , , or CC to TO move VV their PP$ hands NNS . SENT In IN condition NN two CD , , the DT hands NNS condition NN , , they PP were VBD allowed VVN to TO use VV their PP$ hands NNS and CC fingers NNS to TO point VV or CC count VV . SENT They PP were VBD instructed VVN to TO sum VV the DT coins NNS as IN quickly RB as RB possible JJ , , but CC to TO make VV every DT effort NN to TO give VV the DT correct JJ answer NN . SENT The DT results NNS showing VVG the DT mean JJ time NN taken VVN to TO announce VV a DT sum NN , , hereafter RB time NN , , and CC the DT mean JJ number NN of IN mistaken JJ sums NNS , , hereafter RB errors NNS , , are VBP given VVN in IN figure NN 2 CD . SENT On IN average JJ subjects NNS took VVD 22 CD . SENT 5 CD sec NN in IN no DT hands NNS and CC 18 CD . SENT 7 CD sec NN in IN hands NNS to TO announce VV their PP$ answer NN , , and CC they PP were VBD mistaken VVN in IN no DT hands NNS 68 CD or CC 20 CD . SENT 3 CD out RB of IN 30 CD stimuli NNS , , p NN Figure NP 1 CD . SENT To TO get VV a DT sense NN of IN the DT problem NN , , count VV the DT coins NNS depicted VVN in IN figure NN 1 CD . SENT first JJ without IN using VVG your PP$ hands NNS and CC then RB count VV them PP using VVG your PP$ fingers NNS and CC hands NNS . SENT The DT difficulty NN in IN keeping VVG track NN of IN which WDT coins NNS one PP has VHZ counted VVN makes NNS pointing VVG a DT useful JJ complementary JJ strategy NN . SENT Subjects NNS who WP have VHP to TO count VV real JJ coins NNS can MD use VV more RBR powerful JJ complementary JJ strategies NNS such JJ as IN clustering VVG the DT coins NNS into IN denominations NNS , , or CC pushing VVG them PP off RP to TO one CD side NN as IN they PP are VBP counted VVN . SENT These DT are VBP familiar JJ complementary JJ strategies NNS which WDT occur VVP naturally RB . SENT But CC the DT experiment NN reported VVD here RB did VVD not RB use VV real JJ coins NNS or CC permit VV physical JJ re NN arrangement NN . SENT The DT results NNS of IN the DT experiment NN are VBP shown VVN in IN figure NN 2 CD . SENT Figure NN 2 CD . SENT Three CD features NNS of IN this DT simple JJ experiment NN deserve VVP special JJ mention NN . SENT First RB , , each DT subject NN was VBD tested VVN on IN a DT random JJ selection NN of IN 30 CD stimuli NNS in IN each DT condition NN . SENT The DT number NN of IN coins NNS displayed VVN was VBD a DT random JJ selection NN of IN nickels NNS , , dimes NNS and CC quarters NNS totaling VVG anywhere RB from IN 21 CD to TO 31 CD coins NNS , , mean VVP of IN 26 CD , , and CC the DT number NN of IN coins NNS was VBD matched VVN across IN conditions NNS , , so RB that IN each DT subject NN summed VVD three CD sets NNS of IN 21 CD , , 22 CD . SENT . SENT 31 CD coins NNS in IN no DT hands NNS totaling VVG 30 CD stimuli NNS and CC three CD sets NNS of IN 21 CD , , 22 CD . SENT . SENT 31 CD coins NNS in IN hands NNS totaling VVG 30 CD stimuli NNS . SENT All DT S's NP saw VVD the DT same JJ 60 CD stimuli NNS , , which WDT were VBD sufficient JJ to TO yield VV significant JJ differences NNS in IN mean JJ error NN p NN Second RB , , there EX was VBD clear JJ evidence NN that IN subjects NNS evolved VVD strategies NNS microgenetically RB . SENT During IN the DT first JJ 20 CD of IN trials NNS , , or CC so RB , , every DT subject NN appeared VVD to TO be VB experimenting VVG with IN different JJ techniques NNS , , this DT view NN was VBD confirmed VVN in IN verbal JJ reports NNS in IN the DT debriefing VVG . SENT Later RBR a DT dominant JJ strategy NN was VBD selected VVN , , and CC was VBD then RB used VVN for IN the DT rest NN of IN the DT trials NNS in IN that DT condition NN . SENT Once RB a DT comfortable JJ strategy NN was VBD found VVN , , subjects NNS usually RB continued VVD using VVG it PP even RB when WRB the DT condition NN changed VVD . SENT Thus RB , , when WRB subjects NNS were VBD tested VVN first RB on IN no DT hands NNS , , then RB on IN hands NNS , , the DT complementary JJ strategy NN used VVN in IN hands NNS was VBD geared VVN to TO help VV the DT basic JJ strategy NN settled VVN on IN in IN no DT hands NNS . SENT Hence RB the DT term NN complementary JJ strategy NN . SENT performing VVG external JJ actions NNS that WDT complement VV internal JJ actions NNS . SENT There EX was VBD verbal JJ support NN evidence NN of IN this DT pattern NN of IN setting VVG a DT strategy NN to TO be VB used VVN in IN both DT conditions NNS even RB for IN subjects NNS who WP were VBD given VVN the DT hands NNS condition NN first RB . SENT These DT subjects NNS reported VVD trying VVG to TO count VV without IN hands NNS in IN a DT manner NN that WDT resembled VVD the DT way NN they PP had VHD counted VVN with IN hands NNS . SENT Third JJ , , given VVN the DT complexity NN of IN the DT phenomena NNS being VBG studied VVN , , any DT claims NNS about IN what WP is VBZ occurring VVG during IN the DT microgenetic JJ phase NN , , and CC how WRB a DT complementary JJ strategy NN comes VVZ into IN being NN , , must MD be VB speculative JJ at IN best JJS . SENT Clearly RB , , subjects NNS are VBP aware JJ of IN trying VVG out RP new JJ strategies NNS , , both CC mental JJ and CC complementary JJ strategies NNS . SENT But CC we PP cannot NN say VVP how WRB they PP think VVP up RP new JJ strategies NNS , , or CC why WRB they PP settle VVP on IN one CD strategy NN rather RB than IN continue VV looking VVG for IN better JJR ones NNS . SENT Despite IN this DT limitation NN , , it PP is VBZ clear JJ that IN a DT minimal JJ theory NN of IN complementary JJ strategy NN should MD provide VV a DT set NN of IN theoretical JJ principles NNS powerful JJ enough RB to TO explain VV why WRB any DT particular JJ complementary JJ strategy NN succeeds VVZ or CC fails VVZ in IN terms NNS of IN the DT mental JJ resources NNS and CC mental JJ strategies NNS used VVD . SENT Complementing VVG Visual NP Strategies NNS In IN order NN to TO evaluate VV how WRB helpful JJ a DT complementary JJ strategy NN is VBZ , , we PP must MD have VH some DT idea NN of IN the DT mental JJ strategy NN it PP is VBZ complementing VVG . SENT The DT theory NN of IN visual JJ routines NNS Ullman NP 1985 CD provides VVZ the DT starting VVG place NN for IN a DT framework NN for IN discussing VVG certain JJ mental JJ strategies NNS . SENT It PP offers VVZ an DT account NN of IN the DT basic JJ computational JJ operations NNS available JJ to TO a DT subject NN for IN selecting VVG and CC manipulating VVG elements NNS of IN a DT scene NN , , and CC is VBZ therefore RB a DT natural JJ place NN to TO begin VV a DT theory NN about IN visually RB counting VVG . SENT The DT basic JJ idea NN is VBZ that IN visual JJ routines NNS are VBP procedures NNS or CC programs NNS that WDT use VVP primitive JJ visual JJ operations NNS to TO identify VV a DT target NN property NN . SENT The DT flexibility NN this DT gives VVZ the DT visual JJ system NN is VBZ that IN properties NNS invented VVN on IN the DT fly NN , , such JJ as IN a DT group NN of IN four CD quarters NNS not RB yet RB counted VVN , , can MD become VV targets NNS for IN systematic JJ visual JJ search NN . SENT In IN Ullman's NP study VVP there EX is VBZ no DT explanation NN of IN the DT processes NNS which WDT shape VVP the DT evolution NN of IN visual JJ routines NNS . SENT Nor CC is VBZ there RB any DT account NN of IN how WRB limitations NNS on IN non JJ visual JJ memory NN constrain VV the DT type NN of IN visual JJ routines NNS that WDT may MD exist VV . SENT The DT theory NN of IN visual JJ routines NNS , , accordingly RB , , does VVZ not RB explain VV how WRB the DT need NN to TO remember VV intermediate JJ values NNS , , such JJ as IN the DT dollar NN value NN of IN quarters NNS , , or CC dimes NNS , , helps VVZ to TO shape VV visual JJ strategies NNS for IN counting VVG . SENT Ultimately RB , , the DT plan NN an DT agent NN settles VVZ on IN must MD be VB responsive JJ to TO non JJ visual JJ constraints NNS as IN well RB . SENT Thus RB , , if IN the DT plan NN is VBZ to TO first JJ count NN quarters NNS in IN fours NNS , , then RB add VV dimes NNS incrementally RB to TO the DT dollar NN value NN of IN quarters NNS , , then RB add VV nickels NNS , , we PP should MD see VV this DT not RB as IN a DT purely RB visual JJ strategy NN , , but CC as IN a DT mixed JJ strategy NN , , one NN that WDT is VBZ sensitive JJ both CC to TO the DT visual JJ skills NNS and CC visual JJ memory NN limitations NNS of IN the DT agent NN , , and CC the DT non JJ visual JJ memory NN skills NNS and CC limitations NNS of IN the DT agent NN . SENT In IN figure NN 3 CD the DT mixed JJ strategy NN of IN subject JJ SR NN is VBZ depicted VVN . SENT Of IN all DT subjects NNS , , SR NN displayed VVD the DT most RBS significant JJ improvement NN in IN performance NN in IN the DT two CD conditions NNS . SENT What WP is VBZ revealing VVG about IN SR's NP approach NN is VBZ that IN his PP$ mental JJ strategy NN regularly RB called VVD for IN memory NN of IN more JJR visual JJ markers NNS than IN he PP could MD recall VV . SENT By IN substituting VVG certain JJ external JJ actions NNS he PP was VBD able JJ to TO reduce VV demands NNS on IN visual JJ memory NN enough RB to TO reduce VV errors NNS by IN 60 CD and CC increase VV speed NN by IN 20 CD . SENT In IN the DT debriefing VVG , , SR NN described VVD his PP$ strategy NN like IN this DT . SENT First RB , , I PP count VVP the DT quarters NNS by IN grouping VVG them PP into IN fours NNS in IN a DT sort NN of IN clockwise JJ manner NN , , if IN that WDT is VBZ natural JJ . SENT If IN the DT quarters NNS total VVP an DT odd JJ number NN , , such JJ as IN 2 CD . SENT 75 CD , , I PP look VVP around RP for IN a DT nickel NN to TO use VV to TO put VV the DT total NN to TO an DT even JJ number NN , , 2 CD . SENT 80 CD , , so RB that IN I PP can MD now RB add VV the DT dimes NNS easily RB , , by IN just RB adding VVG ten CD to TO what WP is VBZ really RB an DT easier JJR number NN to TO work VV with IN i NP . SENT e SYM . SENT 280 CD instead RB of IN 275 CD . SENT I PP then RB add VVP nickels NNS in IN groups NNS of IN two CD . SENT I PP found VVD that IN if IN I PP had VHD stolen VVN a DT nickel NN I PP would MD often RB forget VV where WRB it PP was VBD . SENT So RB when WRB I PP use VVP my PP$ hands NNS , , I PP put VVD my PP$ thumb NN on IN the DT stolen JJ nickel NN . SENT This DT helped VVD a DT lot NN . SENT Then RB , , when WRB it PP came VVD time NN to TO count VV nickels NNS , , I PP could MD use VV two CD fingers NNS from IN my PP$ other JJ hand NN to TO point VV to TO the DT two CD I PP was VBD on RP , , and CC so RB add VV my PP$ two CD nickels NNS easily RB . SENT In IN what WP follows VVZ I PP will MD discuss VV some DT of IN the DT benefits NNS SR NN reaped VVN with IN this DT curious JJ strategy NN , , and CC tie VV the DT discussion NN to TO more JJR general JJ notions NNS of IN memory NN , , perception NN and CC attention NN . SENT Figure NN 3 CD a DT No NP Hand NP Strategy NP . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT Figure NP 3 CD b SYM Hand NP Strategy NP Figure NP 3 CD a DT is VBZ an DT attempt NN to TO represent VV SR's NP no DT hand NN or CC mental JJ strategy NN . SENT Figure NP 3 CD b SYM portrays VVZ his PP$ hand NN or CC complementary JJ strategy NN . SENT The DT solid JJ triangle NN , , square JJ , , circle NN , , cross NN and CC X NP are VBP used VVN to TO indicate VV the DT visual JJ markers NNS proposed VVN by IN Ullman NP , , and CC the DT closed JJ curves NNS mark VVP subitized JJ regions NNS . SENT In IN 3 CD a DT , , SR NN has VHZ counted VVN all PDT the DT quarters NNS and CC is VBZ about RB to TO start VV on IN the DT dimes NNS . SENT He PP began VVD his PP$ mental JJ counting NN by IN subitizing VVG the DT four CD quarters NNS in IN the DT upper JJ left JJ quadrant NN and CC marking VVG the DT center NN of IN that DT set NN of IN 4 CD quarters NNS with IN his PP$ first JJ visual JJ marker NN , , represented VVN by IN . SENT His PP$ next JJ step NN was VBD to TO subitize VV a DT second JJ group NN of IN 4 CD quarters NNS , , this DT time NN in IN the DT upper JJ right NN quadrant NN , , and CC mark VV the DT set NN by IN . SENT Three CD quarters NNS remain VVP , , all RB in IN the DT lower JJR left VVN , , and CC these DT he PP counted VVN and CC marked VVN with IN . SENT Having VHG now RB found VVN 2 CD . SENT 75 CD worth NN of IN quarters NNS he PP proceeded VVD to TO steal VV a DT nickel NN to TO make VV a DT sum NN of IN 2 CD . SENT 80 CD , , and CC marked VVD the DT stolen VVN nickel NN with IN his PP$ fourth JJ marker NN . SENT He PP then RB turned VVD his PP$ attention NN to TO dimes NNS , , again RB beginning VVG in IN the DT upper JJ left JJ quadrant NN . SENT To TO keep VV his PP$ attention NN on IN that DT location NN he PP marked VVD that DT target NN with IN his PP$ final JJ marker NN , , X NP In NP 3 CD b SYM , , SR NN makes VVZ use NN of IN his PP$ left JJ forefinger NN to TO mark VV the DT location NN of IN the DT stolen VVN nickel NN and CC so RB to TO liberate VV one CD of IN the DT markers NNS for IN additional JJ use NN . SENT In IN his PP$ oral JJ account NN , , SR NN also RB described VVD using VVG his PP$ right JJ hand NN to TO help VV count NN nickels NNS in IN twos NNS . SENT But CC that DT action NN is VBZ not RB shown VVN here RB . SENT Memory NN The DT most RBS obvious JJ cognitive JJ burden NN subjects NNS encounter VV in IN the DT counting VVG experiment NN is VBZ to TO remember VV intermediate JJ sums NNS . SENT It PP is VBZ easy JJ to TO drop VV a DT digit NN in IN counting VVG am VBP I PP at IN 285 CD or CC 385 CD . SENT Some DT subjects NNS response NN to TO this DT problem NN was VBD to TO partially RB count VV with IN their PP$ fingers NNS . SENT Subject JJ JD NP , , for IN instance NN , , would MD encode VV the DT current JJ dollar NN value NN on IN her PP$ fingers NNS so RB that IN all DT mental JJ counting NN could MD be VB done VVN using VVG one CD or CC two CD digits NNS . SENT Thus RB , , rather RB than IN mentally RB counting VVG with IN three CD digits NNS as IN in IN 275 CD , , 285 CD , , 295 CD , , 305 CD , , she PP would MD extend VV two CD fingers NNS and CC count VVP 75 CD , , 85 CD , , 95 CD , , extend VV a DT third JJ finger NN and CC count VV 5 CD . SENT This DT reduced JJ JD's NP working VVG memory NN loads NNS . SENT SR's NP strategy NN was VBD equally RB effective JJ but CC more JJR baroque NN . SENT By IN stealing VVG a DT nickel NN to TO convert VV odd JJ valued VVN amounts NNS 275 CD to TO even VV ones NNS 280 CD , , SR NN achieved VVD the DT same JJ economizing VVG in IN working VVG memory NN as IN subject JJ JD NP did VVD but CC without IN reducing VVG digit NN length NN . SENT His PP$ trick NN was VBD to TO limit VV the DT phonological JJ complexity NN of IN the DT numeral NN he PP had VHD to TO continually RB update VV . SENT The DT number NN of IN syllables NNS kept VVN in IN the DT articulatory JJ loop NN Baddelly NP is VBZ greater JJR for IN two CD seventy CD five CD than IN for IN two CD eighty CD despite IN both DT 275 CD and CC 280 CD being NN three CD digits NNS long RB . SENT Thus RB , , although IN SR NN seemed VVD to TO be VB using VVG three CD digit NN numbers NNS whereas IN JD NP seemed VVD to TO be VB using VVG two CD digit NN numbers NNS , , both DT used VVN equi NN syllabic NN numbers NNS , , e NN . SENT g NN . SENT twen NP ty NP se FW ven NP vs NP . SENT two NN se FW ven NP ty NP . SENT Memory NN savings NNS in IN syllable NN length NN also RB translate VV fairly RB directly RB into IN savings NNS in IN processing NN time NN . SENT Presumably RB , , one CD of IN the DT potential JJ limitations NNS on IN counting VVG speed NN is VBZ the DT time NN needed VVD to TO mentally RB utter JJ twen NP ty NP se FW ven NP that WDT is VBZ , , to TO encode VV updated VVN sums NNS in IN articulatory JJ memory NN . SENT Accordingly RB , , how WRB fast JJ one PP can MD count VV partly RB depends VVZ on IN the DT length NN of IN time NN it PP takes VVZ to TO mentally RB utter JJ the DT numbers NNS . SENT Any DT reduction NN in IN syllable NN size NN translates VVZ to TO a DT reduction NN in IN counting VVG time NN . SENT Baddeley NP et NP al NP , , 84 CD . SENT Hence RB , , SR's NP technique NN saved VVD both DT time NN and CC memory NN . SENT Another DT savings NNS in IN working VVG memory NN produced VVN by IN SR's NP technique NN , , stems VVZ from IN shifting VVG items NNS out RP of IN working VVG memory NN to TO long JJ term NN memory NN . SENT SR NN suggested VVD that IN his PP$ greatest JJS memory NN savings NNS came VVD from IN placing VVG his PP$ thumb NN over IN the DT stolen JJ nickel NN , , for IN now RB he PP no RB longer RB had VHD to TO remember VV both DT whether IN he PP had VHD stolen VVN a DT nickel NN , , and CC which WDT particular JJ nickel NN to TO avoid VV counting VVG . SENT Prima JJ facie NN the DT function NN of IN this DT external JJ marker NN is VBZ to TO liberate VV an DT internal JJ marker NN . SENT But CC arguably RB the DT real JJ savings NN lies VVZ elsewhere RB . SENT For IN presumably RB , , regardless RB of IN whether IN an DT agent NN keeps VVZ an DT internal JJ visual JJ marker NN on IN a DT mental JJ representation NN of IN a DT nickel NN , , or CC an DT external JJ finger NN marker NN on IN a DT physical JJ nickel NN , , he PP still RB must MD keep VV in IN some DT portion NN of IN memory NN the DT meaning NN of IN the DT marker NN here RB lies VVZ the DT stolen JJ nickel NN . SENT In IN the DT case NN of IN internal JJ visual JJ markers NNS , , these DT labels NNS must MD be VB in IN some DT part NN of IN working VVG memory NN , , for IN the DT markers NNS are VBP created VVN on IN the DT fly NN with IN potentially RB ad FW hoc FW meanings NNS . SENT In IN the DT case NN of IN external JJ marking VVG , , however RB , , the DT meaning NN of IN one's JJ thumb NN on IN a DT coin NN may MD become VV conventional JJ , , hence RB drawn VVN from IN long JJ term NN memory NN . SENT Managing VVG Attention NN One NN easily RB forgettable JJ feature NN of IN attention NN is VBZ that IN its PP$ management NN is VBZ either RB highly RB practiced JJ and CC automatic JJ , , or CC is VBZ driven VVN by IN a DT program NN resident NN in IN WM NP . SENT It PP takes VVZ memory NN to TO remember VV the DT strategy NN one NN is VBZ currently RB following VVG . SENT A DT further JJR consequence NN of IN SR's NP technique NN of IN marking VVG the DT stolen JJ nickel NN is VBZ that IN it PP reduces VVZ the DT memory NN costs NNS associated VVN with IN running VVG the DT attentional JJ strategy NN . SENT To TO see VV this DT , , return VV to TO the DT function NN achieved VVN by IN pointing VVG to TO the DT stolen JJ nickel NN . SENT From IN a DT purely RB logical JJ point NN of IN view NN there EX is VBZ no DT reason NN to TO point VV to TO one CD nickel NN rather RB than IN another DT . SENT The DT informational JJ function NN of IN hiding VVG a DT nickel NN is VBZ to TO mark VV the DT fact NN that IN a DT nickel NN , , any DT nickel NN , , has VHZ been VBN counted VVN already RB . SENT This DT function NN could MD equally RB well RB be VB achieved VVN by IN holding NN one's NNS nose NN . SENT But CC , , from IN SR's NP oral JJ accounts NNS it PP was VBD clear JJ that IN he PP did VVD not RB point VV to TO a DT random JJ nickel NN , , and CC clear JJ , , moreover RB , , that IN when WRB it PP came VVD time NN to TO count VV nickels NNS , , he PP would MD intentionally RB avoid VV counting VVG the DT marked JJ nickel NN . SENT This DT is VBZ not RB simply RB idiosyncratic JJ . SENT The DT action NN is VBZ adaptive JJ . SENT For IN had VHD SR NN not RB hidden VVN a DT particular JJ nickel NN , , he PP would MD be VB forced VVN to TO choose VV a DT particular JJ nickel NN to TO overlook VV . SENT That DT is VBZ , , he PP would MD have VH had VHD to TO survey VV all DT nickels NNS , , and CC recall VVP that IN his PP$ finger NN being NN extended VVD meant VVN that IN one CD nickel NN should MD be VB ignored VVN . SENT The DT judgment NN of IN which WDT nickel NN to TO ignore VV , , however RB , , can MD be VB eliminated VVN if IN a DT specific JJ nickel NN is VBZ occluded VVN . SENT Accordingly RB , , if IN SR's NP finger NN serves VVZ an DT occluding VVG rather RB than IN a DT marking VVG function NN , , SR NN will MD have VH fewer JJR regions NNS to TO attend VV to TO since IN he PP will MD know VV which WDT region NN to TO ignore VV . SENT This DT saves VVZ him PP from IN following VVG a DT visual JJ instruction NN such JJ as IN , , go VV to TO that DT region NN but CC ignore VV the DT mentally RB marked JJ nickel NN . SENT That DT is VBZ , , by IN occluding VVG a DT nickel NN , , SR NN is VBZ able JJ to TO reduce VV the DT number NN of IN distracters NNS he PP must MD deal VV with IN , , and CC reduce VV the DT mental JJ overhead NN of IN following VVG an DT attentional JJ strategy NN . SENT Helping VVG Perception NN Although IN pointing VVG and CC marking VVG can NN obviously RB help VV to TO direct JJ attention NN and CC help VV to TO save VV the DT use NN of IN visual JJ markers NNS , , there EX are VBP several JJ purely RB perceptual JJ functions NNS they PP may MD serve VV as RB well RB . SENT Chief JJ among IN these DT is VBZ changing VVG the DT context NN of IN observation NN . SENT In IN placing VVG one's JJ finger NN on IN a DT surface NN , , the DT set NN of IN items NNS in IN view NN is VBZ altered VVN . SENT For IN example NN , , imagine VV a DT subject NN asked VVD to TO count VV the DT following VVG string NN of IN dots NNS . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT . SENT To TO help VV fixate VV on IN one PP dot VVP at IN a DT time NN , , and CC reduce VV the DT interference NN from IN neighboring JJ dots NNS , , a DT subject NN will MD naturally RB want VV to TO use VV a DT finger NN or CC pencil NN to TO help VV keep VV place NN . SENT Belying VVG this DT simple JJ complementary JJ strategy NN , , however RB , , is VBZ a DT sly JJ trick NN . SENT for IN by IN pointing VVG a DT pencil NN tip NN at IN a DT dot VVP , , and CC moving VVG it PP up IN one PP dot VVP at IN a DT time NN , , one PP is VBZ effectively RB counting VVG the DT number NN of IN pencil NN moves NNS , , rather RB than IN the DT number NN of IN dots NNS . SENT The DT dots NNS are VBP still RB too RB small JJ to TO count VV directly RB , , but CC not RB too RB small JJ to TO touch VV one CD by IN one CD with IN an DT instrument NN . SENT Even RB more RBR compelling JJ is VBZ the DT way NN adding VVG a DT feature NN can MD alter VV the DT static JJ gestalt NN properties NNS of IN a DT figure NN . SENT For IN instance NN , , in IN the DT Mueller NP Lyer NP illusion NN it PP is VBZ easy JJ to TO defeat VV the DT appearance NN of IN unequal JJ lines NNS by IN placing VVG one's JJ finger NN over IN the DT V NN portion NN of IN one CD of IN the DT lines NNS and CC visually RB lining VVG up RP the DT tops NNS . SENT Why WRB do VVP these DT actions NNS work NN . SENT How WRB is VBZ our PP$ visual JJ field NN affected VVN . SENT An DT adequate JJ theory NN of IN complementary JJ strategy NN would MD recruit VV enough RB psychophysical JJ theory NN to TO explain VV the DT mechanism NN at IN work NN , , and CC the DT likely JJ decrease NN in IN error NN rate NN . SENT Conclusion NN Intelligent JJ creatures NNS amplify VV their PP$ cognitive JJ abilities NNS by IN adapting VVG their PP$ environments NNS of IN action NN to TO environments NNS where WRB they PP can MD get VV the DT best JJS results NNS from IN their PP$ limited JJ cognitive JJ resources NNS . SENT A DT neglected JJ aspect NN of IN this DT adaptive JJ faculty NN concerns NNS the DT way NN hands NNS , , fingers NNS and CC surrounding VVG material NN objects NNS are VBP recruited VVN for IN cognitive JJ use NN . SENT In IN this DT paper NN a DT simple JJ experiment NN was VBD presented VVN to TO show VV that IN the DT performance NN benefits NNS of IN such JJ spontaneously RB created VVN strategies NNS complementary JJ strategies NNS can MD be VB readily RB measured VVN empirically RB . SENT To TO properly RB understand VV the DT basis NN of IN these DT performance NN improvements NNS , , it PP is VBZ necessary JJ to TO understand VV the DT way NN various JJ external JJ actions NNS fit VVP into IN an DT overall JJ strategy NN of IN computation NN . SENT This DT requires VVZ identifying VVG mental JJ functions NNS served VVN by IN external JJ actions NNS and CC changes NNS , , and CC enumerating VVG the DT resources NNS saved VVN in IN specific JJ cognitive JJ components NNS such JJ as IN visual JJ memory NN , , articulatory JJ loop NN , , attention NN and CC perceptual JJ control NN . SENT I PP have VHP done VVN no RB more RBR than IN gesture NN at IN the DT range NN and CC complexity NN of IN these DT savings NNS here RB , , but CC the DT need NN to TO explore VV complementary JJ strategies NNS should MD be VB evident JJ . SENT Acknowledgements NNS I PP thank VVP Marta NP Kutas NP , , and CC Paul NP Maglio NP for IN their PP$ helpful JJ comments NNS on IN earlier JJR drafts NNS . SENT This DT research NN is VBZ partially RB funded VVN by IN NIA JJ grant NN AG NP 11851 CD . SENT Bibliography NN Baddeley NP , , A NP . SENT D SYM . SENT and CC Lewis NP , , V NN . SENT J NP . SENT and CC Vallar NP , , G NP . SENT 1984 CD . SENT Exploring VVG the DT articulatory JJ loop NN , , Quarterly JJ Journal NP of IN Experimental JJ Psychology NP , , 36 CD . SENT 233 CD 252 CD . SENT Backman NP , , Lars NP , , Dixon NP , , Roger NP A NP . SENT 1992 CD . SENT Psychological JJ compensation NN . SENT A DT theoretical JJ framework NN . SENT Psychological JJ Bulletin NP , , v NN 112 CD n NN 2 CD . SENT 259 CD 283 CD . SENT Hutchins NP , , E NP . SENT 1995 CD . SENT Cognition NN in IN the DT wild JJ , , . SENT MIT NP Press NP Cambridge NP , , MA NN , , Kirsh NP , , D NP . SENT and CC Maglio NP , , P NN . SENT , , 1994 CD . SENT On IN distinguishing VVG epistemic JJ from IN pragmatic JJ action NN , , Cognitive JJ Science NP , , 18 CD . SENT 513 CD 549 CD Kirsh NP , , D NP . SENT , , 1995 CD . SENT The DT intelligent JJ use NN of IN space NN , , Artificial JJ Intelligence NN , , Lave VV , , J NP . SENT 1988 CD . SENT Cognition NN in IN practice NN , , Cambridge NP , , England NP , , Cambridge NP University NP Press NP . SENT Ullman NP , , S NP . SENT , , 1984 CD . SENT Visual JJ routines NNS Cognition NN , , 18 CD . SENT 97 CD 159 CD . SENT