WPC~ ـ FT =/FkT } %o"Hm(ԓ{4pO^ksQQ4w)@j!{)#b1!Oj gWUP3 k-S5^N4:Ydk\i( /fkZf};Ta2'C% 0? 0@ 0_A 0B 0B 0C 0oHD 0D 0oE 0'F 0F 0G 0wH 0I 0IU.J 0J 0:K 0@.MU:nN 0xN 0x O 0O 0>P 07PU@S 0:US 0 CSS 0 lVpPWUNXMY^ Yw@Y^Y\`#d %d mEf\fh 1iini BEvn 0(n4nnr)u C xM#xN'x()xfQxaSxfgxaixf}xaxxw4xnxv}^ }w@}} m} hB} h$2~ h$V~V~V~V~V~U Nz~V& 8Document[8]Document Style0..8` ..` V8Document[4]Document Style.. . V 8Document[6]Document Style8..V 8Document[5]Document Style0..V/8Document[2]Document Style 2A.3  Ԁ   V& 8Document[7]Document Style0..0` ..` zU :Right Par[1]Right-Aligned Paragraph Numbers..2I.3  Ԁ..0..zh :Right Par[2]Right-Aligned Paragraph Numbers..` ..2A.3  Ԁ..0` ..` V?8Document[3]Document Style.. 21.3  Ԁ   z{ :Right Par[3]Right-Aligned Paragraph Numbers..` ..`  ..P 21.3  Ԁ` ..` 0 .. z :Right Par[4]Right-Aligned Paragraph Numbers..` ..`  .. .. 2a.3  Ԁ .. 0..z :Right Par[5]Right-Aligned Paragraph Numbers..` ..`  .. ..h..2(1)3  Ԁ..0h..hz :Right Par[6]Right-Aligned Paragraph Numbers..` ..`  .. ..h..h..2(a)3  Ԁh..h0..z :Right Par[7]Right-Aligned Paragraph Numbers..` ..`  .. ..h..h....2i)3  Ԁ..0..z :Right Par[8]Right-Aligned Paragraph Numbers..` ..`  .. ..h..h....p..2a)3  Ԁ..0p..pVX8Document[1]Document Style  @..^  2I.3  Ԁ     Ԉ l2:Technical[5]Technical Document Style.. 2(1)3  Ԁ. l2:Technical[6]Technical Document Style.. 2(a)3  Ԁ. l/%:Technical[2]Technical Document Style 2A.3  Ԁ   .. l,!:Technical[3]Technical Document Style 21.3  Ԁ   .. l(!:Technical[4]Technical Document Style 2a.3  Ԁ   .. l:0:Technical[1]Technical Document Style  2I.3  Ԁ     .. l1:Technical[7]Technical Document Style.. 2i)3  Ԁ. l1:Technical[8]Technical Document Style.. 2a)3  Ԁ. %\  `$Times NewRoman&R& 8BibliogrphyBibliography0....fp2Doc InitInitialize Document StyleS !    I. 1. A. a.(1)(a) i) a)S ($0 ($0 0 (($0 0 0   A_ekqwDocumentDocument StyleI.1.A.a.(1)(a)i)a)jo4Tech InitInitialize Technical StyleS #  1 .1 .1 .1 .1 .1 .1 .1 S CuyTechnicalTechnical Document Style11.11.1.11.1.1.11.1.1.1.11.1.1.1.1.11.1.1.1.1.1.11.1.1.1.1.1.1.1x?t2PleadingHeader for numbered pleading paper % z &(  ~X&&XX6x''*d66x''*d6\\1\\2\\3\\4\\5\\6\\7\\8\\910111213141516171819202122232425262728  .+(2$ y!  <j:footnote texQ%\  `$Times NewRomanQd#(numPage numbering, except page 1#d#h:Default ParaDefault Paragraph FontXXXQ%\  `$Times NewRomanQ                Q%\  `$Times NewRomanQV:Bulleted LisBulleted List  C,%2A`ArialTTC                Q%\  `$Times NewRomanQ%2A`ArialTTR$1Default Paragraph Font        T-":endnote textendnote text       OZ  US     ^:endnote refeendnote reference        @-"$2footnote text       OZ  US     `:footnote reffootnote reference        L6Document 8Document 8        L6Document 4Document 4          L6Document 6Document 6        L6Document 5Document 5        L6Document 2Document 2        L6Document 7Document 7        PYK8Right Par 1Right Par 1  8.X` hp x 8        5+ ` hp x (#5PYK8Right Par 2Right Par 2  8.X` hp x 8        5+ ` hp x (#5L6Document 3Document 3        PYK8Right Par 3Right Par 3  8.X`  hp x 8        5+ ` hp x (#5PYK8Right Par 4Right Par 4  8.X` hp x 8        5+ ` hp x (#5PYK8Right Par 5Right Par 5  8.X` hp x 8        5+ ` hp x (#5PYK8Right Par 6Right Par 6  8.X` hp x 8        5+ ` hp x (#5PYK8Right Par 7Right Par 7  8.X` hp x 8        5+ ` hp x (#5PYK8Right Par 8Right Par 8  8.X` hp x 8        5+ ` hp x (#5LeW6Document 1Document 1&    8.X` hp x 8        5+ ` hp x (#5'  P\N8Technical 5Technical 5  8.X` hp x 8        5+ ` hp x (#5P\N8Technical 6Technical 6  8.X` hp x 8        5+ ` hp x (#5P8Technical 2Technical 2        P8Technical 3Technical 3        P\N8Technical 4Technical 4  8.X` hp x 8        5+ ` hp x (#5P8Technical 1Technical 1        P\N8Technical 7Technical 7  8.X` hp x 8        5+ ` hp x (#5P\N8Technical 8Technical 8  8.X` hp x 8        5+ ` hp x (#585K,toc 1toc 1           5+ ` hp x (#585K,toc 2toc 2   `         5+ ` hp x (#588K,toc 3toc 3   `         5+ ` hp x (#588K,toc 4toc 4            5+ ` hp x (#588K,toc 5toc 5   h        5+ ` hp x (#585K,toc 6toc 6           5+ ` hp x (#58!,toc 7toc 7          85K,toc 8toc 8           5+ ` hp x (#585K,toc 9toc 9           5+ ` hp x (#5@8K0index 1index 1   `         5+ ` hp x (#5@5K0index 2index 2   `         5+ ` hp x (#5P2K8toa headingtoa heading          5+ ` hp x (#5@-"0captioncaption       OZ  US     ^:_Equation Ca_Equation Caption        Pgg8Word222NullWord222Null  XXXCY<6X9`(CourierC  XXXCY<6X9`(CourierC<6X9`(Courier<ug.NormalNormal  XXXQ%\  `$Times NewRomanQ  XXXCY<6X9`(CourierC<u.FooterFooter    XXXQ%\  `$Times NewRomanQ  XXXCY<6X9`(CourierC`vj:Footnote refFootnote reference  O]<6X9`("Courier NewTTO  XXXCY<6X9`(CourierC<6X9`("Courier NewTTH04heading 1heading 1&        H04heading 2heading 2&        H//$3endnote reference                H,,$4_Equation Caption                64Body Text|(` hp x (#%'0*,.8135@8:<H?ACPFHKXMOR`TVYh[]`XXXQ%\  `$Times NewRomanQXXXUc(X   `(Times NewRomanTTU|(` hp x (#%'0*,.8135@8:<H?ACPFHKXMOR`TVYh[]`(X   `(Times NewRomanTT:8page number($      gk%  1  .  ݀ThisresearchprojecthasbeensupportedbyaUniversityofSouthCarolinaResearch&ProductiveScholarshipAward(#13570E120),whichinturnwasinstrumentalinsecuringaNationalScienceFoundationgrant(SBR9731056)tocontinuetheproject.IwouldalsoliketoacknowledgethecontributionoftheresearchassistantswhohavesosubstantiallycontributedtothisprojectthroughtheexcellenceoftheirGISprogramming:WillBain,DebThomas,RichardDeal,andGuojingShou.Inaddition,IwouldliketothankDaleThomasforhelpinanalyzingthedatageneratedfromthisproject.Anearlierversionofthispaperwaspresentedatthe1998annualmeetingoftheInternationalStudiesAssociation.(-2k$ y!    0  . f gk%  5  .  ݀AdetaileddescriptionoftheGISprogramminghasbeensetoutintheTechnicalAppendixtotheNationalScienceFoundationgrant(generatedfromtheongoingtechnicallogsofthethreeresearchassistants).ReadersinterestedinobtainingacopyoftheAppendixshouldcontacttheauthor.(9 Z6Times New Roman RegularYU( DPPEPPE  gk%  6  .  ݀OneofthemoreusefulaspectsofGISsystemsistheabilitytocreateanysize"buffer"oneithersideofaborder.ItshouldbenotedthattheoriginaltestsandanalyseswereperformedonIsrael(seeStarrandBain1995).GivenIsrael'ssize,a10,000meterbufferoneachsideoftheborderseemedappropiate. +  However,aftertestcaseanalysesonbordersbetweenCambodiaandThailandandEcuadorandPeru,itwasdecidedtousea50,000meterbuffer;thatsizebufferhadbeenemployedforthesaliencedimension(asdescribedbelow)evenintheIsraeli"pretest."Analysesofotherareas,suchasIranandIraq,mayleadustoemploybuffersthatarelargeryet.Suchbuffers,coveringmostorallofacountry,maybeneededforthosewishingtostudyinternalconflictandtheinternal"boundaries"betweengroups.  gk%  7  .  ݀WehaveestimatedthatthefinaldataforIsrael(asfoundinFigure5presentingIsrael'sVitalBordermeasures)requiredonlyonetenththecomputerspaceasthedatarequiredforIsrael'sVitalBordersusingthehexagonmethodology.Alloftheinformation,atallthestagesrequiredtoconstructtheVitalBorders H  inFigure5required2.5to3megs.Usingtheearlierhexagonmethodology,however,thefigurewascloserto15megs.TheNSFGrantTechnicalReportofAugust1996byDebThomas,discussesthechangeinmethodology,andpresentstherevisedproceduresfordeterminingtheopportunityforinteraction.(i &&    gk%  8  .  ݀TheOpportunityforInteraction/EaseofInteractionindexwasdevelopedasfollows:   ` 4= presenceofaroadandthepresenceofrailroad,andlowslope     ` 3= aroadorarailroad,andlowslope     ` 3= aroadandarailroad,andmediumslope h    ` 3= noroad,norailroad,andlowslope \B    ` 2= aroadorarailroad,andmediumslope 6    ` 2= aroadandarailroad,andhighslope     ` 2= aroadorarailroad,andhighslope     ` 1= noroad,norailroad,andmediumslope      ` 1= noroad,norailroad,andhighslope     gk%  9  .  ݀ThispointwasmadebyoneoftheanonymousrefereesoftheNSFgrantproposal.However,asthesamerefereenoted:"Starrhasnoapriorineedtodefinewhata'vitalborder'is;ratherhisresearchshouldinvestigatehowdifferentelementsinhisproposedindexrelatetorelevantdependentvariablesratherthantrying,exante,todefinetheanswer."Thisis,indeed,oneofthepurposesoftheproject,andaddressedbelowinthediscussionofIRhypotheses.U  gk%  10  .  ݀Boulding(1962,265)notes:0  Thelegalboundaryofanation,however,isnotalwaysitsmostsignificantboundary.Weneedtodevelopaconceptofacriticalboundary,whichmaybethesameasthelegalboundarybutwhich % maylieeitherinsideitoroutsideit...Thepenetrationofanalienorganizationinsidethiscriticalboundarywillproducegravedisorganization...War,thereforeisonlyusefulasadefenseofthenationalorganismifitiscarriedonoutsidethecriticalboundary(emphasisinoriginal). (#(#   =   CRight ParRight-Aligned Paragraph NumbersI.A.1.a.(1)(a)i)a)#|a%\  `$Times NewRoman&&a\  P6G;&P%\  `$Times NewRomanX\  P6G;P%\  `$Times NewRomanXXj\  P6G;XP%2A`ArialTTomanJ2PQP<6X9`(CourieromanXx6X@`7X@<6X9`("Courier NewTTZ6X@C@(X   `(Times NewRomanTTXXrX  PCXP d($LW1'zPleadi  gk%  2  .  ݀AccordingtoMarble(1990,10),tobeconsideredatrueGISasystemmustincludethefollowingfourmajorcomponents:0  1)Adatainputsubsystemwhichcollectsand/orprocessesspatialdataderivedfromexistingmaps,remotesensors,etc. (#(# 0  2)Adatastorageandretrievalsusbsystemwhichorganizesthespatialdatainaformwhichpermitsittobequicklyretrivedbytheuserforsubsequentanalysis. (#(# 0  3)Adatamanipulationandanalysissubsystemwhichperformsavarietyoftaskssuchaschangingtheformofthedatathroughuserdefinedaggregationrulesorproducingestimatesofparametersandconstraintsforvariousspacetimeorganizationorsimulationmodels. (#(# 0  4)Adatareportingsubsystemwhichiscapableofdisplayingallorpartoftheoriginaldatabaseaswellasmanipulateddataandtheoutputfromspatialmodelsintabularormapform.  gk%  3  .  ݀DataforthisprojectisfromtheDigitalChartoftheWorld(DCW),producedbyESRIfortheDefenseMappingAgencyin1992.ThedatacontainedintheDCWwasderivedprimarilyfrommapsintheDefenseMappingAgencyOperationalNavigationChartseriesthatwereusedtogeneratea1:1,000,000scalevectordatabasecoveringtheentiresurfaceoftheearth.Onemajor(perhaps"heroic")assumptionoftheproject,isthattheborderdatageneratedbythe1992DCWcanbeusefullyappliedbackwardfor2025years,andforwardforatleastadecade.Thatis,thedataretainvalidityasaroughsurrogatefortheeaseofinteractionandsalienceofareasforthistimeframe.  gk%  4  .  ݀ListofdatalayersintheARC/INFODigitalChartoftheWorld:PoliticalandOceans;PopulatedPlace;Railroads;Roads;Utilities;Drainage;DrainageSupplemental;Hypsography;HypsographySupplemental;OceanFeatures;Physiography;Aeronautical;CulturalLandmark;TransportationStructure;Vegetation;LandCover.Aseventeenthlayer,theDataQualitylayer,providesinformationontheparticularsourceofdataforagiventileandwhenthatsourcewaslastupdated. d,TABLE A,TABLE A,TABLE ADEVEE%\  `$Times NewRomanXXj\  P6G;XP*\  `(Times NewRomanTTd\  PCP\  `(Times NewRomanTTC\  PQP%2A`ArialTTomanTTJ2PQP<6X9`("Courier NewTTTTXx6X@CX@<6X9`("Courier NewTTTZ6X@C@%\  `$Times NewRomanT&&a\  P6G;&P  = MvU(  D{D[DU @@@(9 Z 6Times New Roman Regular y!      XX&&@OPPORTUNITY,WILLINGNESSANDGEOGRAPHICINFORMATIONSYSTEMS(GIS):@@ P  RECONCEPTUALIZINGBORDERSININTERNATIONALRELATIONS f e  1       @  @.HarveyStarr@= = DepartmentofGovernment&InternationalStudies@&UniversityofSouthCarolina@gg+Columbia,SC29208@+(803)7772676/7292@nn+starrharvey@sc.edu0  ABSTRACT !(#(# 0  ThispapercontinuesaprojectwhichhasdevelopedamajorreconceptualizationandrevisionofhowbordersmaybeseenandmeasuredthroughtheuseofGISbyallowingustotalkaboutthespecificqualitiesofbordersintermsofopportunityandwillingness:easeofinteractionandsalience,respectively.Thisreconceptualizationhasalsoleadtothecreationofadatasetwhichletsusgobeyondsimplyobservingthenumberofbordersastatepossesses,whetherornotaborderexistedbetweentwostates,orthelengthofthatborder. (#(# Paperpreparedfortheconference,"NewMethodologiesfortheSocialSciences:TheDevelopmentandApplicationofSpatialAnalysisforPoliticalMethodology,"InstituteofBehavioralSciences,UniversityofColorado,Boulder,Colorado,March1012,2000.  -P(+ 7XXd&&d7 GEOPOLITICS,BORDERSANDIR:ABRIEFINTRODUCTION   Areviewoftheliteratureonwar,militarizeddisputes,enduringrivalries,andalliances,indicatesthatoverthepast20yearstherehasbeenarenewedattentiontotheroleandimpactofgeographyinthestudyofinternationalrelations.Muchofthisisrelatedtothe"newgeopolitics"whichtreatsgeographyasanessentialpartofthecontextofpossibilitiesandconstraintsthatfaceforeignpolicydecisionmakers(seeStarr1991a;Ward1992;GoertzandDiehl1992;andevenGoertz1994).Studiesofthediffusionofbehavioralphenomenaaswellastheinvestigationoftherelationshipsbetweenproximity,contiguity,locationandterritorytointernationalinteractions,haveburgeoned.Theseactivitiesbystudentsofinternationalrelationshaveparalleledthoseofgeographerswhoarefocusingona"newgeopolitics"basedonthepossibilitiesthatthegeopoliticalenvironmentprovidetohumandecisionmakers.Muchofmyownworkhasaddressedquestionswhichfitwithinthisbroadgeopoliticalperspective.Beginningwiththedevelopmentoftheopportunityandwillingnessframework(Starr1978),Ihaveengagedincollaborativeresearchwhichhassoughtbothtorefinethetheoreticalbasisoftheseconceptsaswellastospecifytherelationshipsbetweenthem(e.g.MostandStarr1989;CioffiRevillaandStarr1995;FriedmanandStarr1997).Theseconceptshavealsobeenclarifiedthrougheffortstooperationalizetheminthestudyofinternationalpoliticalphenomena.Thelatterinvestigationsincludethestudyofthediffusionofinternationalphenomena:violentconflict(e.g.MostandStarr1980;SiversonandStarr1991),anddemocracy(e.g.Starr1991b,1999).Thecentralfeatureoftheanalyseshasbeenafocusonthenatureandeffectsofspatialproximityasoperationalizedbyinternationalborders.Onemajoraimofthepresentprojectcanbeseenasaconceptclarificationexercisetoreviseandreconceptualizehowwethinkaboutbordersandhowtheyshouldbemeasured.Whiletherearemanyfacetstoageopoliticalapproachtointernationalrelations,bordershavebeenofprimaryinterestto,andthemainfocusof,geopoliticalscholarship.Forexample,asoneimportantcomponentinthemappingoffactorsrelatedtotheonsetofwar,theCorrelatesofWarprojecthasdevelopedthemostextensiveandcompletedatasetonbordersavailableforinternationalsystemicactorssince1816(e.g.seeGochman1992,foradescriptionoftheCOWmeasurementrules;seealsoStarrandMost1976).Bordershavebeenstudiedaspartoftheanalysisofmanyofthecentralconcernsofinternationalrelations.Abrieflistoftheseconcernswouldinclude:thenumberandtypesofinteractionsamongstates;interdependenceamongstates,withinregionalgroupingsandthelevelofinterdependencewithintheinternationalsystemasawhole;regionalintegration;theprobabilityofwaramongstates;thediffusionofwarandotherformsofinternationalconflict;thediffusionofadditionalinternationalphenomena,suchasthespreadofdemocracy;understandingwhyandhowterritoryaffectstheonsetofwar;theprocessesunderlyingthelackofwarbetweenpairsofdemocracies;theeffectsofalliances;thequestionofinternationalregionsandthestructureoftheinternationalsystem. CONCEPTUALIZING(ANDRECONCEPTUALIZING)BORDERS  0*%( Thus,thelocationofstates,theirproximitytooneanother,andespeciallywhetherornottheyshare"borders,"emergestimeandagainaskeyvariablesinstudiesofinternationalconflictphenomena: ,`'* frommajorpowergeneralwar,tothediffusionofinternationalconflict,totheanalysisofpeacebetweenpairsofdemocracies.FromBoulding's(1962)ideasof"behaviorspace,""lossofstrengthgradient"and"criticalboundary"tothesimplebutprofoundconcernofgeographersthathumansinteractmostwiththosetowhomtheyareclosest,therearepowerfultheoreticalreasonstobeinterestedinborders,andhowtheyaffectinternationalrelations.Buthowexactlydobordersaffect p internationalinteraction?Clearly,akeydimensionformanyresearchersisproximity(seeforexample,Gochman's1992 @  discussionofbordersastheyrelatetotheoverallCOWproject).Myowndiffusionresearchmovedtothestudyofbordersafterconcludingthatthediffusionofcertainphenomenacouldonlybestudiedbylookingatunitsthatwere"relevant"tooneanotherandthatsuchrelevancecouldbeindicatedbygeographicalproximity(seealsotheworkofLemke1995,1996).Proximity,inturn,couldbeoperationalizedthrough"borders."BorderswereseenasimportantindicatorsofproximitybecausetheyhadimportantrelationshipstoboththeopportunityandwillingnessofstateactorsasconceptualizedbyMostandStarr(1976).Onekeyaspectofbordersisthattheyaffecttheinteraction   opportunitiesofstates,constrainingorexpandingthepossibilitiesofinteractionthatareavailable  tothem.Statesthatshareborderswilltendtohaveagreatereaseofinteractionwithoneanother,and  thuswilltendtohavegreaternumberofinteractions.Thisideadevelopedfrommultidisciplinarysources,suchaseconomistKennethBoulding's(1962)conceptofthelossofstrengthgradient;orgeographerG.K.Zipf's(1949)"lawofleasteffort."Theimportantissueraisedhereisthatborderscreatetheopportunityforinteraction(seeStarrandMost1976,MostandStarr1980,andSiversonandStarr1991forafulldiscussionofgeographicopportunity).Suchopportunitymightbeseenintermsofthenumberofothercountrieswithwhichanysinglestate @ hasinteractionopportunities.Itmightalsobeseeninthedegreetowhichsuchopportunityexistsbetweenanyparticularpairofstates.So,forexample,Wesley(1962)arguesthatthelengthofacommonborderbetweentwocountriesisabettermeasureof"geographicopportunity"thansimplythenumberofborders.And,presagingtheGISdiscussiontobepresented,hegoesontosuggestthanlengthshouldbemeasurednotin"actualphysicallength"butintermsofpopulationunits.Ihavearguedthattheopportunitiesforinteractionviewofbordersgetsattheimportantconceptualcoreofproximityinawaythatothermeasuresof"distance"donot.Suchmeasureshaveincludedtheuseoftheairmileagebetweenthecapitalsofstatestomeasuredistance(e.g.GleditschandSinger1975,Garnham1976).Onepurposeofthepresentprojectwouldbetotesttheutilityofdifferentconceptionsofproximityasinteractionopportunties.Secondly,bordersalsohaveanimpactonthewillingnessofdecisionmakerstochoosecertainpolicyoptions,inthattheyactasindicatorsofareasofgreatimportanceorsalience.Becauseotherstates p&!$ areclose,havinggreatereaseofinteractionandtheabilitytobringmilitarycapabilitiestobear,theyarealsokeyareasofexternalcues(ordiffusion).Accordingly,activitiesintheseareasareparticularlyworrisome,cancreateuncertainty,andthusdeserveattention.ThenotionthatchangesinborderingareascreateuncertaintybecauseoftheirproximitywasbasedonargumentsdevelopedbyMidlarsky(e.g.1970,1975),andappliedinMostandStarr(1980). ,`'* StarrandMost(1976:10)werealsoparticularlyconcernedwiththe"rolesthatdifferenttypesofbordersappeartoplay"inwarinvolvement.Differenttypesofbordermighthavedifferentialimpactsonbothopportunityandwillingness.Thus,bordersweredifferentiatedintermsofhomelandbordersandbordersgeneratedbycolonialterritories.Thisdifferentiationallowedustotestwhetherallterritorywasseenasequallyimportant,orwhetherhomelandterritorygeneratedgreaterwillingnessthanmoredistantlyheldcolonial/imperialterritories.Implicitlytestedinsuchanalyseswasthenotionthatitwashomelandterritoryperse,thatwasimportant:thattheproximityofanyhomeland P  territoryofonestatetoanyhomelandterritoryofanotherstatewastheimportantfactor.Whilesome @  ofourdiffusionanalysesindicatedthatthiswasprobablythecase,otheranalysesalsodemonstratedthestrongimpactofcolonialterritorialbordersonthediffusionofwar.Simply,colonialterritorieswereresponsibleforcreatingagreaternumberofopportunitiesforconflictualinteraction(inaway `  thatisusefullycomplementarytoChoucriandNorth's[1975]notionoflateralpressure"intersections").Thus,wewerecorrectintheinclusionofcolonialborders(atwistuniquetointernationalrelationsresearchofthe1970s,asfoundinouroriginalborderdataset).StarrandMost(1976)alsodistinguishedbetweenlandbasedcontiguityandacrosswaterproximity.Again,suchadistinctionimplicitlydealtwithpossiblevariationsineaseofinteractionandsalience.  Thepresentprojectseekstobuilduponthesetwodimensionsofbordersasindicatorsofproximity,toreviseandreconceptualizehowbordersmaybeseenandmeasured.TheuseofGIS(GeographicInformationSystems)willpermitamuchfullerandclearerspecificationofbordersbyallowingustotalkaboutthespecificqualitiesofbordersintermsofopportunityandwillingness.The p reconceptualizationwillpermitustogobeyondsimplyobservingthenumberofbordersastatepossesses,whetherornotaborderexistedbetweentwostates,orthelengthofthatborder.Bysodoing,anumberofquestionsraisedinstudyingtheissuesnotedabovecannowbemorefullyaddressed. METHODOLOGY:ARC/INFOANDTHECONCEPTUALIZATION P OPERATIONALIZATIONOFBORDERS  @ Geographicinformationsystems,developedthroughtheearlytomid1960s,arenowthefocusofalargeliteratureproducedbygeographersandregionalscientists(seeabriefreviewinStarrandBain1995).Ascouldbeexpected,therearemanyapproachesandperspectivesonGIS.ItisimportanttounderstandthataGISisatool,foundedonavarietyofcomputertechnologies,that "  permitstheintegrationofdataaboutthespatialityofphenomenaalongwithdataaboutothercharacteristicsofthosephenomena. { e  2      ׀ItisimportanttonotethatGISismorethanmerecomputer $" mapping.AccordingtoCowen(1990,57),theheartofaGISsystemisitsabilitytooverlayvariouslayersorcoveragesofdatasothattheGISwouldhave"creatednewinformationratherthanjust p&!$ haveretrievedpreviouslyencodedinformation"(emphasisadded).ThereconceptualizationofbordersusingGISderivesfromtheabilitytogeneratenewmeasuresornewinformationaboutthenatureofborders.TheGISsystemutilizedinthisstudyisARC/INFO,developedandsuppliedbyEnvironmentalSystemsResearchInstitute,Inc.(ESRI).ItisoneofthemostwidespreadcommercialGISsinuse ,`'* globally.Itsstrengths,inpart,resideinitsabilitytointegratemanykindsofdata,aswellas"anopenarchitecturewhichallowsittobelinkedtoanumberofrelationaldatabasemanagementsystems"(PeuquetandMarble1990,91).ARC/INFOemploysa"georelational"approach,whichabstracts"geographicinformationintoaseriesofindependentlydefinedlayersorcoverages,eachrepresentingaselectedsetofcloselyassociatedgeographicfeatures(e.g.,roads,streams,andforeststands)"(ESRI1992,14). | e  3      ׀Thelargescaledatabaseconsistsofthesixteenlayersofdata.Theselayers `  containdatarangingfromphysicalcharacteristicssuchasdrainagenetworks,hypsography(elevationandtopographicrelief),andlandcover,tomanmadefeaturessuchasroadnetworks,railroadnetworks,andaeronauticaldata. } e  4      ׀ 0  StarrandBain(1995)describehowvariouscoverageswithintheARC/INFOsystemwereusedtorevisittheopportunityandwillingnessdimensionsofborders. h e  5      ׀Insubsequentwork(withthe P  researchassistanceofDebThomasandRichardDeal),thespecificmethodologyusedforextractingandcombiningtheborderdatafromthevariousARC/INFOcoverageshaschangedsubstantially(seetheTechnicalAppendixreferredtoinnote5).However,thedatasources,coverages,andoverallcriteriafordeterminingtheopportunityforinteractionandsaliencedidnotvary.TheGISmethodology,however,mustbedrivenbytheoreticalconsiderations.ThevariouslayersoftheARC/INFOGIScontainagreatnumberofvariables,andthekeyquestionmustbewhichof  thesevariablesshouldbeselectedtocreatevalidindexestorepresentopportunityandwillingness(easeofinteractionandsalience).Thereisalargeliteratureonthenatureofboundariesproducedbygeographers;(andbygeopoliticianswritingearlierthiscentury).Thisliteratureoftenfocusesonthemeaningofboundariesforgroupformation,groupidentity,andgroupmaintenance;(e.g.seeFalahandNewman1995,aswellassuchseminalworksasPrescott1987orGlassner1992).Asnoted,themereexistenceofabordercantellusmanythingsandgenerateresearchhypotheses.Yet,"aborderisnotaborderisnotaborder."Bordersservemanyfunctions,andobviouslytakeondifferentmeaningsindifferentspecificcontexts(e.g.seesuchdiscussionsasGiddens1984,orGoertz1994).Nosingleoperationalizationofborder,andnodatasetbasedonsuchanoperationalization,willbeabletoprovidethetotalhistoricalpoliticalcontextforallpairsofstates.Nevertheless,drawingontwodecadesofattentiontotheseidea,Iarguethattheopportunityandwillingnessconceptualizationdoestapkeyelementsoftheproximityborderconcept,andconstitutesaprogressivestepinthemoregeneralconceptualizationofborders,context,andtheanalysisofinternationalinteraction.OpportunityforInteraction(EaseofInteraction) % # Regardingopportunity,thevariableselectiondecisionswerebasedontheoreticalconsiderationspresentedbybothgeographersandIRscholars.ThenotionofeaseofinteractionderivesfromBoulding's(1962)concernwiththe"lossofstrengthgradient,"andtheabilitytoprojectconventionalmilitarypower.Outofthewelterofpossiblevariables(andtakingvarioustechnical/analyticconstraintsintoaccount),threecentralfactorsforthemovementoflandbasedmilitarycapabilitywereselectedtheexistenceofroads,railroads,andthesteepnessofterrain.ThisprojectdrawsfromWesley(1962),whosuggeststhatscholarsshouldusecrossborderroadsto ,`'* operationalizegeographicopportunity.Similarly,Lemke(1996),usingBuenodeMesquita's(1981)operationalizationofthelossofstrengthgradient,isconcernedwiththedistanceoverwhichmilitaryforcescanmoveinspecificperiodsoftime,andhowthisrelatestothepropensityforwarsorMIDs(militarizeddisputes).Lemkeisspecificallyconcernedwithpavedroadsandrailroadsinattemptingtoestimatethelossofstrengthgradient,andthosecountriesthatmakeup"relevant"dyads/neighborhoodsforAfricanstates.BasedontheworkofBoulding,BuenodeMesquita,andLemke,anindexcreatedfromthesethreefactorsbothreflectseaseofinteraction,andisapplicable(valid)acrossalargesetofinternationaldyadicboundaries.UsingdataavailableintheDCW'sdatalayersanindexwasconstructedforeaseofinteractionwhichaggregatesvaluesgeneratedfromARC/INFO.Itcanbeusedtocharacterizeanyborder(termedan `  "arc"inARC/INFO)orbordersegmentontheglobe.Afterreviewingallthevariableswithinallof P  thelayers,thefollowingwereselectedtocreatean"easeofinteraction"index.Thefirstvariablelookedforthepresenceorabsenceofroadswithinthelocationsbeingstudied.Roadsincludedmultilanedividedroads,aswellasprimaryandsecondaryroads(Layer4,RoadLayer,RDLINE).Thesecondvariablewasthepresenceorabsenceofrailroads(Layer3,RailroadLayer,RRLINE).Thethirdvariableinvolvedtheslopeofanarea,whichwasbasedontheelevationvaluesofcontourlines(inmeanfeetabovesealevel;Layer8,HypsographyLayer,HYNET),andwasderivedfromadigitalterrainmodelbyconvertingthehypsographyintoatriangulatedirregularnetwork.Eachofthesevalueswasinvestigatedforabufferareaof10,000metersoneachsideofallinternational  borders. m e  6       p ThemethodologyusedbyStarrandBaininvolvedtheuseofhexagons,andthecombineddensityofthedifferentcoveragesforroads,railroadsandhypsographyproducedbyapolygonattributetableforeachhexagon.Thisgeneratedametric(aneaseofinteractionindexrunningfrom0to373) 0 wherebyhigh"saturation"ordensityindicatedthegreatesteaseofinteraction.Theleastsaturated  p hexagonsarethosewhicharemostdifficulttomoveacross;themostheavilysaturatedhexagonsarethemosteasilytraversed.@{{)[Figures1and2here]However,despiteseveraladvantagestotheuseofhexagons,thismethodologyincreasedboththecomputingcapacityandtimerequiredforgeneratingthedatafromARC/INFO.Amethodologyusingthevectordatadirectlywasdevelopedwhichappearedtosimplifythegenerationofthedataconsiderably,whileproducingroughlyanalogousresults. n e  7      ׀Therevisedcolormapofopportunityfor $" interactionforIsraelisseeninFigure1;asimilarrepresentationoftheIranIraqborderispresentedinFigure2.Withthisformulationwesimplynotethepresenceorabsenceofroadsandrailroads(ratherthanthesaturationofahexagon),andrepresentedthehypsographyorslopeasfollows:coded1iftheslopewas05degrees,coded2iftheslopewas520degrees,andcoded3ifitwasgreaterthan20.Thiscreatedasimplecombined1to4index,with4(redinthecolormaps)representingthe @)$' greatesteaseofinteraction,and1(darkgreeninthecolormaps)themostdifficultareastomove 0*%( across. o e  8      ׀  +p&)  ,`'* Notethattherevisedproceduresdescribedherewerespecificallydevelopedtogenerateafour  categoryscheme.Thiswaspurposelydoneinordertofacilitatethetranslationofthecolormapsintoblackandwhiterepresentations(asfoundinStarr1998).Whilemapsareanimportantmediumforthepresentationofresults,itmustbestressedthatanysectionofanybordercannowberepresented  byavaluefrom1to4:valuesthatcanbeusedindataanalyseswithintheGIS,bothwithotherGISvariablesoranyotherdatasetsthatareimportedintotheARC/INFOGIS.Suchadatasetwillbedescribedbelow.Figures1and2areimportantindemonstratingthattheeaseofinteractioncanvaryalonganysingle 0  borderthatastatemighthavewithacontiguousneighbor(thevalues,asrepresentedbydifferentcolors,vary).Theopportunityforinteractionvariablecanbeusedtoindicatethisvariationalonganysingleborder(or"arc",forexampleIsrael'sborderwithLebanon).Thiswouldcapturethevariationthatmightoccuronaverylongborderanyparticularportionofabordercanbethusbe @  characterizedastoitsdegreeofpermeability.Thus,wearenowabletogobeyondthesimpleideathatinsomewaycontiguityprovidesthepossibilityforinteractionwhilesomepartsofsomeborderswouldmakethishighlylikelyorpossible,otherpartswouldmakeinteractionmuchlesslikely.Wecouldmakesuchjudgmentsirregardlessofthelengthofaborder,orthenumberofdifferentbordersthatastatemighthave.Salience(Willingness)  Thesaliencedimensionofproximity/bordersisconcernedwiththeimportanceorvalueofterritoryalongorbehindaborder.Again,thequestionishowimportance/valueistobemeasured.BelowIwilldiscussanalternativehypothesisthatitisterritorypersewhichisimportant.However,herewe P mustbeconcernedwithindicatorswhichwoulddiscriminatethelevelofvalueorconcernoverterritory(asMostandStarrdidbydifferentiatingbetweenhomelandterritoryandcolonialpossessions;andbetweencontiguouslandbordersandacrosswaterborders).Drawingoncemoreongeographers,demographicsareseenasimportant:theterritoryonwhichastate'spopulationlives.Thiswastobeoperationalizedbyareasofpopulationconcentration.Acapitalcity,thelocusofgovernmentalactivityandthesymbolofthestate,shouldalsobeusedtoindicatetheimportanceofterritory.WhileIdisagreethatdistancebetweencapitalcitiesshouldbeconsideredaprimaryindicatorofproximity,suchstudies(e.g.GleditschandSinger1975)dohighlightthecentralimportanceofcapitalcities.Notethatinselectingareasofpopulationconcentrationandtheseatsofgovernmentwehavenowcapturedallthreeofthecentralelementsofthestatefoundintheinternationalrelationsliterature:territory,population,andgovernment.AreasofurbanconcentrationincludingurbanizedareasandcapitalcitieswereextractedfromthePopulatedPlaceLayer,Layer2(PPPOLYandPPPOINT).Othercoveragesprovidedthelocationofitemsthatwouldindicatetheimportanceofanarea.Forinstance,fromLayer13,theAeronauticalLayer,activecivilandmilitaryairportswereidentified(AEPOINT).TheCulturalLandmarkLayer(Layer14)providesacatalogueofsuchitems,including:militarycamps,forts,oilwellsandrefineries,powerplantsofvariouskinds,watertanks,factories,industrialcomplexes,hospitals,telecommunicationsstations,etc.ThewidevarietyofitemsfoundinLayer14wasusedbecausethesubstantiveimportanceofanysingletypeofinstallationcouldvary ,`'* considerablyacrossstates.Byidentifyingthelocationofkeyaspectsofastate'stransportation,communication,energyproduction,industrial,agricultural,andsecurityinfrastructures,wehaveitemsthattap"importance"inamannergenerallyrelevanttoallstates.@{{)[Figures3and4here]Thesalienceindexwasdevelopedinmuchthesamefashionastheindexforopportunityforinteraction.Afterreviewingthevariouscoverages,thesalienceorimportanceofaborderareawasdeterminedbyplacesofpopulationconcentration,statecapitals,airfields,andselectedculturalfeatureslocatedwithina50,000meterbufferoftheregion'sborders.Onerulethatwasusedacrossmethodologies,wasthatacapitalcitywasautomaticallycodedwiththehighestvaluefoundinanyoftheunitsofanalysis.Therevised,vectorapproachtosalienceisdescribedthusly:"Thewillingnessforinteractioncomponentofthestudyprovidedmoreofachallengeinusingavectorapproachbecausethehexagonswereusedto'count'thenumberofculturalfeatures,numberofpopulationcenters,andnumberofairportswithineachone.UsingthePOINTDISTANCEcommandincombinationwithaFREQUENCYcommandallowedeachfeaturetobegivenavaluebasedonthenumberofotherfeaturesthatwerewithin4kilometers.Thesecouldthenbemappedbasedonthevalue,showingwhereclustersarose"(NSFTechnicalReport#2byThomas;seealsoDeal,TechnicalReport#3onhowthesalienceindexeswereconstructuted).TheresultsoftheserevisedanalysesareshowninFigures3and4,whichdisplaytheclusteringofpointcoveragesindicatingtheimportanceofanarea,withgraphicsrepresentingthenumbersofpointsthatoverlapwithinfourkilometerranges.Notethatwhilereducingthetimeandcapacityrequirementsforanalysis,thevectorbaseddatasets/methodsalsoproducemoregraphicallypleasingandinterpretableresults.WhilethedatageneratingstrengthsofGIShavebeenstressed,itwouldbeusefulheretoemphasizeaswellthevisualutilityofGISitsabilitytogeneratemapswhichhelpinvestigatorstolookatdata 0 differently,andprovideheuristicsforgeneratinghypothesesandmodelspecification.Thus,whilecolormaprepresentationsofsalienceareuseful,rememberthatanyareainabufferaroundabordercannowbecharacterizedbyavaluefrom1to4,whichcanbeutilizedindataanalyses.Afourvaluescalewascreatedtoindicateareaswithoneimportantfeatureupthroughthosewithfourormore.Thesizeofthecirclesalsohelpsindicatethelevelofsalience.And,again,Figures3and4areimportantbydemonstratinghowbordersmaydifferintheirimportanceintermsofwherepeoplelive,wherethecapitalcityislocated,wheresignificantelementsofthetransportation,military,oreconomicsystemsaresituated.Portionsofborderswheremoreoftheseitemsarelocated,(withina50,000meterbufferoftheborder),couldbeseenasmoreimportantorsalient(theareasinredororange)thansegmentswithoutpopulationcenters,oreconomic,militaryortransportationfacilities.Aswithopportunityforinteraction,thisrepresentationofthesalienceofborderspermitsustodifferentiatewholeborders,todifferentiateportionsoflongborders,andtomakesenseastowhysomebordersmightbeseenasmoreimportantthanothers;whychangesoreventsacrosssomebordersmightgeneratemoreuncertaintythanoccurrencesacrossotherborders.@-[Figure5here] ,`'* VitalBorders  TheuseofaGISdataset,then,permitsanewmechanismforoperationalizingastate'sborders.AGISsystemhasbeenusedtocreatenewdata.Throughtheindexesgenerated,wecanattachvaluestoasingledyadicborderorbordersegment.Thesevalueswillindicatetheeaseofinteractionprovidedbythatborder,and/ortheimportanceofanyparticularborderorbordersegment.Thesetwodimensionscanbeusedseparatelyorcombined.RecallthatMostandStarr(1989)arguethatopportunityandwillingnessarejointlynecessaryconditionsforcertaintypesofbehavior,andthat @  theyarerelatedtoeachotherincomplexways(seealsoCioffiRevillaandStarr1995).Forwantofabetterterm,Ihavesuggestedthataborderwithhighvaluesonbothcouldbeconsidereda"vitalborder,"(aspresentedinFigures5).Thecoreofthisconceptisthatanarcorabordersegmentmaycombinehighorlowvaluesreflectingbothopportunityandwillingness.Theremayindeedbesome P  military/securityaspectsofborderswhicharenotcapturedbytheschemepresentedhere,orconfoundit:e.g.becauseaborderisstrategicallyimportant,roadsandrailwaysmaynotbebuilt! q e  9       0  Figure5indicatesthatfourcategorieshavebeenusedincombiningopportunityforinteractionandsalience.Giventhatbothopportunityforinteractionandsaliencewerepresentedas4pointscales,theirjointcombinedvaluecanrunfrom28.Forabordertobeconsidered"vital"itmusthaveajointvalueof7or8demandingavalueofeither3or4oneachdimension(againrepresentedasred  areasonthecolormaps).Onthecolormapsdarkgreenareasrepresentbordersegmentswhicharethe"leastvital."Vitalbordersthusrepresentareasthatarebothhighlypermeableeasytocrossandalsoencompasspopulationcentersand/orfeatuesofeconomic,politicalorsocialimportance.Oncemore,theyarerepresentedbyvaluesthatcanbeusedinstatisticalanalysesorrepresentedonmaps.@-[Table1here] ANEWDATASETONTHE"NATURE " OFBORDERS  0 AsshowninTable1,theGISgeneratedmapscanbereducedtoarelativelycompactdatasetusefulforquantitativeanalysis.Foreachofthe301contiguouslandbordersbetweenstates17variableshavebeendeveloped,whichcanbetransformedintoavarietyofnominal,ordinal,andintervalmeasures.Foranydyadborder(seetheexampleofIsrael'sborderusedinTable1)wecanpresentthelengthofthatborderinkilometers,andtheareaunderthebufferscreatedfromthatborder.Fromthesetwovariableswecanpresentthepercentageofeachborderthatfallsintocategories4through1(or,redthroughdarkgreeninthecolorfigures).Thiscanbedoneforeaseofinteraction,saliency,orvitalness.Knowingthelengthofthearc,theareaunderthebufferalongit,andthepercentageofeachcategory,permitstheanalysttouseintervaldata(asnotedbelow)orbroadlybasedcategoriessuchashighsalienceorlowsalience.NotealsothatTable1providesaweightedaverageforeachborderintermsofeaseofinteraction,salience,orvitalness,showingtheaveragevalueacrossthewholeborder. ,`'* @-[Table2here]Thisdatasetincludes151stateswithlandborders,whichgenerate301separatecontiguouslandbordersbetweenstates.Thestatesinthisgroupthusaveragealmostfourborderseach.Table2providesdescriptivedataonthetotalsetofborders,usingtheweightedaverages.Forexample,weseethattheaveragesalienceisquitelow,barelygettingabove1.000(withamaximumvalueof1.369onthe4.000scale).Thismeansthatalthoughwefindmany"red"areas(scaleof4.000)onthemaps,theyconstituteonlyverysmallportionsofthetotalborder.Thevaluesforeaseofinteractionaremuchhigher(thus,soarethevaluesforvitalness).Interestingly,theborderwiththehighestsaliencescoreexistsbetweenMoldovaandtheUkraine.Inmanywaysthisshouldnotbesurprisingsinceuntillessthanadecadeago,thiswasonlyaninternalborder,ortheequivalentoftheborder  p  betweenConnecticutandMassachusetts.Withaweightedaverageof1.342,theGermanDutchborderisthenexthighest.AclusterofrelativelyhighsaliencebordersarefoundamongtheoriginalmembersoftheEEC.And,becauseofahighdensityofroadandrailfacilities,theyalsohaveamongthehighestweightedaveragesintermsofeaseofinteraction.ThatisthebordersoftheoriginalcorecountriesoftheEUalsohavebordersthatlookliketheinternaljurisdictional   boundariesofstates. THESCIENTIFICUTILITYOFAGISGENERATEDBORDERDATABASE   OnebasicpointraisedinMostandStarr(1989)wasthatresearchersneededtobemuchclearerastothebroaderconceptswhichwerereallyunderinvestigation,sothattheirmodelsandtheresultingresearchdesignscouldbemorelogicallyandfullyspecified.Perhaps"borders"canbeusedinsomeresearchforreasonsthatareinnateto"borderness"thattheyseparateentitiesfromoneanother. P However,asdiscussedabove,mostusesofbordersinvolvetheirrepresentationofproximitythatis,entitiesareclosetooneanother,importanttooneanother,andhaveanenhancedabilityto 0 interactwithoneanother.But,doestheexistenceofaborderactuallyrespresentthesenotions?  p Bordersthataredifficulttotraverse,eithercommerciallyormilitarilymaynotfitthisideaofproximity.Borderswhichare"buffered"byemptyandmeaninglessspacesmaynotfitthisideaofproximity.Conversely,legalbordersinthecontemporaryworldmaybemeaninglessintermsoffullpermeabilityandhighlevelsoftransactionsasintheEuropeanUnion.Theconceptofavitalborderwithitstwosubcomponentsspecifiesmorecompletelyandpreciselyhowabordermightrepresent"proximity"andallowsustoinvestigatethemeaningofaborder,traditionalor ! transnational.RevisitingIRHypotheseswithMoreFulltSpecifiedBorders $" Awidearrayofresearchquestionsbasedontheassumptionthatbordersindicateproximity,salience,andeaseofinteractionmaybeaddressedbyavitalbordersdataset.Forexample,whichtypesofbordersaremostorleastrelatedtospatialdiffusion?IntheMostandStarranalysesofthediffusionofconflict(e.g.1980),theyinvestigatedwhetherstateswhichweresubjectedtothe"treatment"ofhavingaWarringBorderNation(WBN)weremorelikelytobecomeinvolvedinconflictthanthosewithoutsuchatreatment.Isitthenatureoftheborderratherthannumbersofbordersthataffect 0*%( conflictbehavior?Boulding(1962)alsointroducedtheconceptofcriticalboundaries,aspartofhisconcernwiththeviabilityofstatesinregardtoneighbors. s e  10      ׀OnesuggestiveanalysisinStarrand ,`'* Bain(1995)wasthattheGISgeneratedmapsofsalienceandopportunityforinteraction,indicatedthatin1967IsraelhadchangeditslegalborderswithJordanandSyriatomatchwhatcouldbeconsideredtobeitscriticalboundaries.Whatistherelationshipbetweenvitalbordersandcriticalboundaries?AreattemptstomatchthetwobehindtheeventslistedintheMIDsdataset?causesofwaringeneral?causesofwarsoverterritoryinparticular?OnefindingofSiversonandStarr(1991,5455)isthattherelationshipbetweenjoininganongoingwarandbeingsubjectedtoWBNsandWAPsisoneof"loosenecessity."Manystateshavetreatments,butdonotjoinwars.Infact,havingonlyonetreatment(onlyoneWBNand/orWAP) 0  appearedtohavealmostnoeffectonthebehaviorofstates.Thus,onepotentialresearchprojectusingthenewVitalBorderdatasetwouldbetoinvestigatethenatureofthebordersthatseparate `  thestatefromitsWBN.Iftheborderisavitalborder,doesitincreasetheprobabilityofthediffusionofwar/conflict.Thatis,bothsetsofstudieswereconcernedwith"treatments."AnewnullhypothesiscouldbethatanyWBNtreatmentenhancestheprobabilityofwardiffusion.Morefully 0  specifiedhypotheseswouldproposethatvitalborderWBNshaveagreaterprobabilityofenhancingdiffusion;(orthataborderneedstoscorehighoneithersalienceoropportunityforinteraction).Is  itsimply"borderness"insomevaguesense,orthesemorespecificqualitiesthatareinvolved?Lemke(1995,1996)raisesthesamequestion.AswithMostandStarr,Lemkeisconcernedwiththekeyquestionofidentifying"relevantdyads."Indeed,hegoesbeyondthistotrytoidentify"relevantneighborhoods."Usingthisproject'snewlycreatedopportunityforinteractionindex,Lemke'squestionscanbespecifiedevenmoreclosely.AnotherresearchprojectcouldinvolvethecomparisonofLemke'sfindingstothosebasedonGISgeneratedborderdataonopportunityforinteraction.Furtherrefinementoftheproceduresused(particularlythesizeofthebuffersemployed)forgeneratingvitalbordersareclearlyneeded,asisanindependentoperationalizationforcriticalboundary.Butevensettingasidethequestionofcriticalboundariesdostates(especiallythosewithlargenumbersofborders)demonstrateahigherincidenceofconflictalongvitalborders?alongsalientborders?alongborderswithgreatesteaseofmovement?RevisitingBordersasPartofTheWarPuzzle 0 iXXXXPartoftheconsiderableresearchdevotedtobordersandterritory(e.g.StarrandMost1976,1978;   Vasquez1993)suggeststhatterritorialcontiguityisamajordeterminantofwhetherornotastatewillgotowarwithanotherstate.Indeed,asasignificantpieceofthewarpuzzle,Vasquezsuggeststhatterritorialcontiguityisthe sourceofconflictmostlikelytoresultinwar(Vasquez1993,307).However,perhapssimplecontiguitymaynotbethecriticalfactor.Droppingonelevelofanalysislower,Vasquez(1993)alsohypothesizesthatthenatureoftheborderbetweentwostatesalsoaffectstheprobabilitythatstateswillgotowar.Specifically,hehypothesizesthatbordersthatcoincidewithnaturalfrontiersorthattraverseuninhabitedregionsorareseenashavinglittlevaluearemuchlesslikelytoprovokewarsthandissimilarbordersandborderareas(thisisathemepickeduplaterbyLemke).ThedatasetgeneratedbytheGISprojectprovidesanidealwaytotestthislatterhypothesisregardingthenatureofborders.Threedifferentgroupsofconflictdyadswereselectedforthis ,`'* analysis:enduringrivalries(_Goertz_ԀandDiehl1993;1995),territorialdisputes(_Huth_Ԁ1996),andmilitarizedinterstatedisputes(usingthe1996oftheMIDSdataset).Thesecaseswereselectedforpairsofstatesthatsharedacontiguouslandborder,andwheretheconflict(orseriesofconflicts)involvedfellintothebroadtemporalbandcoveredbytheGISdata(refertoendnote#3).Foralistoftheconflictdyadsidentifiedbythesedatasetsandthatareusedinthisanalysispleaserefertotheappendix.Inordertotestthenullhypothesisthatborderscoincidingwithnaturalfrontiers(agreaterdifficultyininteraction)haveeithernoeffectonthelikelihoodthatstateswillgotowarwithneighborsormakestatesmorelikelytogotowar,onemustcomparethenatureofborderswhereconflicthasoccurredwiththosewhereconflicthasnotoccurred.Thesameistruefortheeffectoftheperceivedimportance(salience)ofaborderarea.Onestrategywouldbetocomparethenatureofconflictborderswithallotherbordersinthesystem.However,thisstrategyisflawedsinceitfailstoaccountfordifferencesingovernmentandfordifferencesinthepropensityofindividualstatestoenterintowars.Instead,weemployamoreconservativestrategythatcantakethesedifferencesintoaccount;wetestforstatisticallysignificantdifferencesbetweenconflictdyadborders(thesharedconitguoushomelandborderbetweentwostates)andtheremainingbordersofthetwostateswhichformtheconflictdyad.Analternativehypothesisisalsoexamined;itpositsthatratherthanthenatureoftheborderbeingimportant,thelengthoftheborderistheprimarydistinguishingfactorbetweenconflictbordersandnonconflictborders.Thelongeraborderis,thegreatertheopportunityforinteractionand,therefore,conflict.  Asstatedearlier,eachborderhasbeenmeasuredintermsofeaseofinteraction,salience,andajointmeasureofthetwo,vitalness.Borderareashavebeenassignedtooneoffourcategories!categoryonebeingthemostdifficultforinteractionorleastimportantandcategoryfourbeingtheeasiestforinteractionandthemostimportant.Similarly,vitalnessisalsomeasuredinfourcategorieswithcategoryonebeingtheleastvitalandcategoryfourbeingthemostvital.Thesemeasuresarethengivenaspercentagesofthetotalborderlength(easeofinteraction)andarea(salience).Theyhavealsobeencalculatedinabsolutetermsofkilometersandkilometerssquaredforeachofthefourcategories.Finally,aweightedaverageforeaseofinteraction,salience,andvitalnessisgivenforeachborder.  Theresultsoftheanalysisofenduringrivalrydyads(N=22dyads)aregiveninTable3.Themeansforenduringrivalrybordersarecomparedtotheremainingbordersofstatesintheenduringrivalry.Specifically,atestofindependentmeanshasbeenconductedforeachcategorybothinrelativeterms(percentage)andabsoluteterms(kilometersorkilometerssquared).Thechoiceofatestofindependentmeansisappropriatesincethenatureofanystateborderisindependentofthenatureoftheremainingbordersforthatstate.Surprisingly,notasinglecategorymeasuringthenatureofthebordershowsastatisticallysignificant(__Ԁ=.10)differenceinmeans.Onthisbasis,onecan P(#& concludethatthenatureofthebordersofastateengagedinanenduringrivalrydoesnotsignificantlyimproveourchancesofpredictingwhichneighboringstatewillbetheenduringrival.Ontheotherhand,thealternativenullhypothesisthatnostatisticaldifferenceexistsbetweenthemeanborderlengthforenduringrivalryconflictdyadsandthemeanlengthofremainingenduring ,`'* rivalborderscanberejected.Theprobabilityofseeingasampledifferencethislarge(520.81KM),ifthedifferenceinthepopulationequalszeroisp=.09.Thissuggeststhatbasedsolelyontheknowledgeofthelengthofthebordersofastateengagedinanenduringrivalry,wecanconfidently  predictwhichstatewillbetheenduringrivalryconflictdyadpartner.Conflictdyadstakenfrom_Huth_s(1996)territorialdisputesdataset(N=27dyads)showsimilarresults.SeeTable4.Whereaswiththeenduringrivalrydataset,nostatisticallysignificantdifferencesexistedbetweentheconflictdyadbordersandtheotherbordersofenduringrivalsintermsofeaseofinteraction,salience,orvitalness,twodifferentcategoriesofsalienceshowstatisticallysignificantdifferences.Thesearethenumbersquarekilometersrankedcategoryoneandthepercentageoftheborderrankedcategorythree.Oddly,thedifferencebetweentheterritorialdisputedyadsandtheremainingbordersofthedisputantshastheoppositesignfromwhichtheoriginalhypothesiswouldsuggest.Theconflictdyadsfrom_Huth_sterritorialdisputedatasetactuallyhavemorekilometersoflowimportanceborderareathandotheircorrespondingnonconflictborderareas.Similarly,thehypothesiswouldleadonetoexpectthatconflictdyadswouldhavealargerpercentageofhighimportanceareasthancorrespondingnonconflictdyads;however,thisisalsoshowntobefalse.Therefore,onecannotrejectthenullhypothesesthatnaturalbordersandlowimportanceborderareasdonotreducethelikelihoodthatstateswillgotowar.Nonetheless,thenullhypothesisassociatedwithborderlengthcanbeeasilyrejectedtheprobabilityofseeingthislargeofameandifference(465.19KM)inthesampleifthepopulationdifferenceisactuallyzeroisverysmall(p=.03).Thus,supportfortheargumentthatthelongeraborderisthemorelikelythatconcernedstateswillgotowarappearstobegrowingstronger.  Thefinalsetofanalysesusesthemilitarizedinterstatedispute(MIDS)data(N=61dyads).SimilartestsregardingthedifferencesbetweenthemeansofconflictdyadsandtheremainingbordersofconflictpartieshavebeenconductedandtheresultsareshowninTable5.TheMIDSdataismuchmoreinclusivethanthatofeither_Goertz_ԀandDiehl(__1993;1995)or_Huth_Ԁ(1996).Consequently,61usableconflictdyadshavebeenidentifiedasopposedtothe27for_Huth_Ԁand22for_Goertz_ԀandDiehl.Nevertheless,aphenomenonsimilartothatfoundinthe_Huth_ԀterritorialdisputedataexistsintheMIDSdata.Severalcategoriesshowstatisticallysignificantdifferences(__Ԁ=.05),butinevery 0 casethesignofthedifferenceisoppositeofthatsuggestedbythehypotheses.Oneisforcedonceagaintoacceptthenullhypothesesthatmoredifficulttocrossbordersandlessimportantborderareasdonotlowerthelikelihoodthatstateswillgotowarwiththatneighbor.However,differencesinthemeanlengthofbordersforconflictdyadsandnonconflictbordersforconflictpartiesremainstatisticallysignificant(p=.03).  Consequently,thisanalysissubstantiallyunderminestheVasquez(1993)hypothesesthatthenatureofaborderintermsofeaseofinteractionandimportanceaffectswhetherornotstatesaremorelikelytofightacrossit.Statisticallysignificantdifferenceswiththeexpectedsigndonotappearinanexaminationoftwentythreecategoriesmeasuringthenatureofborders.Nonetheless,thealternativehypothesisthatthenatureoftheborderdoesntmatterbutthatwhatdoesmatteristhelengthoftheborderreceivesbroadsupportfromananalysisofthreemajorconflictdatasets.#XXXiXߟ#  +p&)  ,`'* Ї CONCLUSION   TheseinitialanalysesiXXXXdonotsupportVasquez'snotionthatdroppingtolowerlevelsof"interaction  opportunity"increasestheabilitytoexplainwar.Thefindingthatthelengthofthebordermatters,however,doessuggestthatthegeneralconcernbyMostandStarrwith"opportunitiesforinteraction"(e.g.1980)wasgenerallycorrectintermsofterritorialcontiguity#XXXiXd#iXXXX(andthatWesley `  [1962]wasontherighttrack)#XXXiX#iXXXX.ItalsosupportsotheranalysesbyVasquezinwhichhedemonstrates P  theimportanceofanyterritorytostates.Territoryappearstobeimportant;theopportunitiesfor @  territorytobecomepartofconflictareincreasesbythelengthofcontiguousterritory,andnotbymorespecificmeasuresofopportunityandwillingness.#XXXiXߞ#TheseanalysesprovidesomeindicationoftheutilityoftheGISbasedconceptualizationanddataset.Theydemonstratethatsuchadatasetcanbeusedtoinvestigateanumberofrelatedquestions,forexample:Whatsortsofborderscanbefoundbetweenstatesinenduringrivalries?Whatisthenatureoftheterritoryoverwhichconflictsarise?GoertzandDiehl(1992),Holsti(1991),Huth(1996),forexample,focusonterritoryperseasacauseofwar;asboththeissueoverwhichwar  breaksout,andasafactorwhichincreasesthestakesofawar.Suchanalysesprovideuswithaveryimportantalternativehypothesis:itisterritoryanyterritorywhichcreatesanopportunityfor  conflict,whichservesastheissueforwar,andwhichmakesthestakesworthfightingover.TheGISbaseddatasetnowpermitsanalyststotestthesecompetinghypotheses.Insum,boththeoreticallyandsubstantivelygeopoliticalfactorsareimportanttoawiderangeofissuesinthestudyofinternationalrelations.Onekeyaspectofbordersisthattheyaffecttheinteractionopportunitiesofstates,constrainingorexpandingthepossibilitiesofinteractionthatareavailabletostates.Statesthatshareborderswilltendtohaveagreatereaseofinteractionwithoneanother.Secondly,bordersalsohaveanimpactonthewillingnessofdecisionmakerstochoosecertainpolicyoptions,inthattheyactasindicatorsofareasofgreatimportanceorsalience.TheARC/INFOGISpermitsustooperationalizeandinvestigatethesetwodimensions--opportunityaseaseofinteraction,andwillingnessassalience/importance.UsingdataavailableinthedifferentdatalayersfoundinARC/INFO'sDigitalChartoftheWorld,wehaveconstructedindexesofbotheaseofinteractionandofsalience.Theycanbeusedtocharacterizeanyborder(orarc)orbordersegmentontheglobe.TheuseofaGISdataset,then,permitsanewmechanismforoperationalizingastate'sborders.Wecannowgobeyondsimplynotingtheexistenceofaborder,oritslength.Throughtheindexesgenerated,wecanattachvaluestotheeaseofinteractionaborderorbordersegmentprovides,and/ortheimportanceofanyparticularborderorbordersegment.Thesetwodimensionscanbeusedseparatelyorcombined.Aborderwithhighvaluesonbothcouldbeconsidereda"vitalborder."TheGISgeneratedindexespermitustotapbothdimensions,andtousethemsinglyorcombinedgiventheresearchquestionunderconsideration.  `'"% _  APPENDIX(X(#(#(RecentEnduringRivalryDyads(_Goertz_Ԁand  Diehl1993;1995)Afghanistan ` Pakistan p Algeria ` Morocco `  Argentina0 ` ChileP ` `  Cambodia ` Thailand @  China0   ` India0  China0   ` Russia p  Congo  0 ` Zaire` ` `  Egypt0   ` IsraelP  Ethiopia ` Somalia @  Ethiopia ` Sudan 0  Greece ` Turkey   India  0 ` Pakistan` `  Iran0   ` Iraq Iraq0   ` Kuwait Israel  0 ` Jordan` `  Israel0   ` Syria Jordan0   ` Syriap Kenya0   ` Uganda` Laos0   ` ThailandP Norway ` Russia @ NorthKorea ` SouthKorea 0 SaudiArabia ` Yemen  p RecentTerritorialDisputes(_Huth_Ԁ1996) P Afghanistan ` Pakistan @ Argentina0 ` Chile0` `  Argentina ` Uruguay   _Bukina_ԀFaso ` Mali ! Bhutan ` China "  Bangladesh ` India #! Bolivia ` Chile $" Brazil0   ` Uruguay% # Cambodia ` Vietnam p&!$ Chad  0 ` Libya`'"%` `  China  0 ` VietnamP(#&` `  China0   ` India@)$' China0   ` Russia0*%( Egypt  0 ` Sudan +p&)` `  Ethiopia ` Somalia ,`'* Ghana p 0  Togo*(#(# Guyana  Surinam + Guyana  Venezuela p, Honduras  Nicaragua ` - Honduras  ElSalvador P . India0 p   Pakistan@ /p(#p(# Ireland  UnitedKingdom 0 0 Iraq p 0  SaudiArabia p1(#(# Laos p 0  Thailand` 2(#(# XXXXLesotho#XXXX#  SouthAfrica P 3 Oman p 0  UnitedArabEmirates@ 4(#(# NorthKorea0  SouthKorea0 5(#(# RecentMilitarizedInterstateDisputes(1996) 7 Afghanistan  Pakistan 8 Algeria  Morocco 9 Argentina0  Chile:(#(# Armenia  XXXXAzerbaijan#XXXX# ; _Bukina_ԀFaso  Mali p< Bangladesh  India `= Bangladesh  Myanmar P> Bosnia  0  Yugoslavia@? (# (# XXXXBotswana#XXXXw#  SouthAfrica 0@ Cambodia  Thailand  pA XXXXCameroon#XXXX#  Nigeria `B Chad p 0  LibyaPC(#(# Chad p 0  Nigeria@D(#(# China p 0  Vietnam0E(#(# China0 p   India Fp(#p(# China0 p   Russia!Gp(#p(# Columbia  Venezuela "H Congo p 0  Zaire#I(#(# CostaRica  Nicaragua $J Croatia  Yugoslavia % K XXXXCzechRepublic#XXXX߲#  Germany p&!L Ecuador  Peru `'"M Egypt0 p   IsraelP(#Np(#p(# Egypt p 0  Libya@)$O(#(# Estonia  Russia 0*%P Ethiopia  Somalia  +p&Q Ethiopia  Sudan ,`'R Ghana  0 ` Togo` `  XXXXGuinea-Bissau#XXXXߦ#ԀSenegal  Greece ` Turkey  XXXXGeorgia#XXXX?# ` Russia  Guyana ` Venezuela p Honduras ` Nicaragua `  Honduras ` ElSalvador P  Hungary ` Yugoslavia @  India0   ` Pakistan0  XXXXIndonesia#XXXX߻# ` XXXXPapuaNewGuinea#XXXX#  p  Iran  0 ` Iraq` ` `  Iran0   ` TurkeyP  Iraq  0 ` Kuwait@ ` `  Iraq  0 ` SaudiArabia0 ` `  Iraq  0 ` Syria ` `  Iraq  0 ` Turkey` `  Israel0   ` Jordan Israel  0 ` Syria` `  Jordan0   ` Syria Kenya0   ` Somalia Kenya0   ` Ugandap Laos0   ` Thailand` Libya  0 ` SudanP` `  Libya  0 ` Tunisia0` ` @ XXXXMauritania#XXXX)# ` Morocco 0 XXXXMauritania#XXXXߝ# ` Senegal  p Myanmar ` Thailand ` Mozambique ` SouthAfrica P Norway ` Russia @ NorthKorea ` SouthKorea 0 XXXXRwanda#XXXX# ` Uganda   Sudan  0 ` Uganda!` `  Uganda ` Zaire "  Zaire  0 ` Zambia#!` `   ! (X(  % ! @2NOTES %  й)  * @ww/REFERENCES0   Boulding,KennethE.1962.ConflictandDefense.NewYork:Harper&Row.(#(# 0   BuenodeMesquita,Bruce.1981.TheWarTrap.NewHaven:YaleUniversityPress.` (#(# 0   Choucri,NazliandRobertC.North.1975.NationsinConflict.SanFrancisco:W.H.Freeman.@ (#(# 0   CioffiRevilla,ClaudioandHarveyStarr.1995."Opportunity,Willingness,andPoliticalUncertainty:TheoreticalFoundationsofPolitics."JournalofTheoreticalPolitics7 `  (January):447476. (#(# 0   Cowen,DavidJ.1990."GISversusCADversusDBMS:WhataretheDifferences."InIntroductory 0  ReadingsinGeographicInformationSystems,eds.DonnaJ.PeuquetandDuaneF.Marble.   London:TaylorandFrancis. (#(# 0   0(#(#  EnvironmentalSystemsResearchInstitute.1992.ARC/INFO:GISTodayandTomorrow.New  York:ESRI. (#(# 0   Falah,GhaziandDavidNewman.1995."TheSpatialManifestationofThreat:IsraelisandPalestiniansSeeka'Good'Border."PoliticalGeography14(8):689706.`(#(# 0   Friedman,GilandHarveyStarr.1997.Agency,StructureandInternationalPolitics:FromOntology @ toEmpiricalInquiry.London:Routledge.0(#(# 0   Garnham,David.1976."DyadicInternationalWar,18161965:TheRoleofPowerParityandGeographicProximity."WesternPoliticalQuarterly29(2):231242.P(#(# 0   Giddens,Anthony.1984.TheConstitutionofSociety.Berkeley:UniversityofCaliforniaPress.0(#(# 0   Glassner,M.J.1992.PoliticalGeography.NewYork:Wiley.!(#(# 0   Gleditsch,NilsPetterandJ.DavidSinger.1975."DistanceandInternationalWar,18161965."ProceedingsoftheIPRA5thConference,IPRAStudiesinPeaceResearch.Oslo:InternationalPeaceResearchAssociation. (#(# 0   Gochman,CharlesS.1992."InterstateMetrics:Conceptualizing,OperationalizingandMeasuringtheGeographicProximityofStatesSincetheCongressofVienna."InTheNewGeopolitics, P(#& ed.MichaelD.Ward.Philadelphia:GordonandBreach. (#(# 0   Goertz,Gary.1994.ContextsofInternationalPolitics.Cambridge:CambridgeUniversityPress. +p&)(#(#  ,`'* 0   Goertz,GaryandPaulF.Diehl.1992.TerritorialChangesandInternationalConflict.London:  Routledge. (#(# 0   Goertz,GaryandPaulF.Diehl.1993."EnduringRivalries:TheoreticalConstructsandEmpiricalPatterns."InternationalStudiesQuarterly37(2):147171.p(#(# 0   Goertz,GaryandPaulF.Diehl.1995."Taking'Enduring'OutofEnduringRivalry:TheRivalryApproachtowarandPeace."InternationalInteractions21(3):291308.@ (#(# 0   Holsti,K.J.1991.PeaceandWar:ArmedConflictsandInternationalOrder16481989.Cambridge:  p  CambridgeUniversityPress. (#(# 0   Huth,PaulK.1996.StandingYourGround:TerritorialDisputesandInternationalConflict.Ann @  Arbor:UniversityofMichiganPress. (#(# 0   0(#(#  Lemke,Douglas.1995."TheTyrannyofDistance:RedefiningRelevantDyads."International  Interactions17(1):113126.(#(# 0   Lemke,Douglas.1996."WhySoMuchPeace?:InterstateWarinAfrica."PresentedattheAnnualMeetingoftheAmericanPoliticalScienceAssociation,SanFrancisco,CA. (#(# 0   0(#(#  Marble,DuaneF.1990."GeographicInformationSystems:AnOverview,"InIntroductoryReadings ` inGeographicInformationSystems,eds.DonnaJ.PeuquetandDuaneF.Marble.London: P TaylorandFrancis. (#(# 0   Midlarsky,ManusI.1970."MathematicalModelsofInstabilityandaTheoryofDiffusion."InternationalStudiesQuarterly14:6084.`(#(# 0   Midlarsky,ManusI.1975.OnWar.NewYork:FreePress.@(#(# 0   Most,BenjaminA.andHarveyStarr.1980."Diffusion,Reinforcement,Geopolitics,andtheSpreadofWar."AmericanPoliticalScienceReview74(December):932946.!(#(# 0   Most,BenjaminA.andHarveyStarr.1989.Inquiry,LogicandInternationalPolitics.Columbia, #! SC:UniversityofSouthCarolinaPress. (#(# 0   0(#(#  Peuquet,DonnaJ.andDuaneF.Marble.1990."ARC/INFO:AnExampleofaContemporaryGeographicInformationSystem."InIntroductoryReadingsinGeographicInformation `'"% Systems,eds.DonnaJ.PeuquetandDuaneF.Marble.London:TaylorandFrancis.P(#&(#(# 0   Prescott,J.R.V.1987.PoliticalFrontiersandBoundaries.London:AllenandUnwin.0*%((#(#    +p&) 0   SiversonRandolphM.andHarveyStarr.1991.TheDiffusionofWar.AnnArbor:Universityof   MichiganPress. (#(# 0   Starr,Harvey.1978."'Opportunity'and'Willingness'asOrderingConceptsintheStudyofWar."  InternationalInteractions4(4):363387.p(#(# 0   Starr,Harvey.1991a."JoiningPoliticalandGeographicPerspectives:GeopoliticsandInternationalRelations."InternationalInteractions17(1):19.@ (#(# 0   Starr,Harvey.1991b."DemocraticDominoes:DiffusionApproachestotheSpreadofDemocracyintheInternationalSystem."JournalofConflictResolution35(2):356381.` (#(# 0   0(#(#  0(#(#  Starr,Harvey.1998."GeographicInformationSystems(GIS)andtheStudyofBordersinInternationalRelations."PresentedattheAnnualMeetingoftheInternationalStudiesAssociation,Minneapolis. (#(# 0   Starr,Harvey.1999."DemocraticDominoesRevisited:DemocraticDiffusion,19741996."PresentedattheAnnualMeetingoftheInternationalStudiesAssociation,Washington,DC. (#(# 0   Starr,HarveyandWillBain.1995."TheApplicationofGeographicInformationSystems(GIS)toInternationalStudies."InternationalStudiesNotes20(2):18.p(#(# 0   Starr,HarveyandBenjaminA.Most.1976."TheSubstanceandStudyofBordersinInternationalRelationsResearch."InternationalStudiesQuarterly20(December):581620.@(#(# Vasquez,John.1993.TheWarPuzzle.Cambridge:CambridgeUniversityPress.  p 0   0(#(#  Ward,MichaelD.,ed.1992.TheNewGeopolitics.Philadelphia:GordonandBreach.P(#(# 0   Wesley,JamesP.1962."FrequencyofWarsandGeographicalOpportunity."JournalofConflict 0 Resolution6(4):387389. (#(# 0   Zipf,G.K.1949.HumanBehaviorandthePrincipleofLeastEffort.Cambridge,MA:Addison "  Wesley. (#(# t  P(#&  t =  ==   XXXXTABLE1.ComponentsofaNewDataset:TheExampleofIsrael'sBorders  Israel'sBorderWith:   Egypt h   Jordan   p Lebanon  Syria   Length(km)  `    220 h   410   p 110    92 P   Area(sqkm)  `    21,108 h   39,282   p 8,944    10,184 0   Percent `  EaseofInteractionCategory:    ` 1   7.41 h   28.97   p 29.58    19.06 0     ` 2   4.44 h   18.13   p 24.78    12.43      ` 3   86.03 h   50.24   p 42.56    61.55     ` 4   2.12 h   2.67   p 3.08    6.96   Percent  SalienceCategory:    ` 1   91.62 h   92.34   p 81.50    90.23 `    ` 2   6.18 h   6.31   p 11.48    7.22 P    ` 3   2.20 h   1.35   p 6.13    2.56 @    ` 4   0.00 h   0.00   p 0.88    0.00 0  Percent ` VitalBorderCategory:    ` 1   10.80 h   45.60   p 49.30    29.90 0    ` 2   80.50 h   46.70   p 36.80    56.80      ` 3   8.56 h   7.49   p 11.90    12.90 !    ` 4   0.90 h   0.22   p 02.03    3.76 "   WeightedAverage p&!$ ofEaseofInteraction   2.83 h   2.27   p 2.19    2.56 `'"%  WeightedAverage @)$' ofSalience  `    1.11 h   1.09   p 1.26    1.12#XXXXM # 0*%(  WeightedAverage ,`'* ofVitalBorder    1.98 h   1.62   p 1.67    1.84 -P(+   -@),= TABLE2 `  GISBasedDataset:SummaryStatistics     `  Easeof p    `  Interaction h   Salience p Vitalness   Length  `  Minimum `  1.195   h   1.000   p 1.097    3.0 0  Maximum `  3.296   h   1.369   p 2.462    6900.0 P  Median `  2.800   h   1.013   p 1.918    520.0   Mean   `  2.597   h   1.044   p 1.818    792.8  StandardDeviation `  0.500   h   0.071   p 0.299    863.8 ` N=301casesWeightedAverages(exceptforlength)      p _TABLE30 ` AssessingtheBordersofEnduringRivals` (#` (# *wxddd Xdd Xdd X(#(#w,Ydd ,dd ,dd ,dd ,~dd ,>dd ,edd +  $xx  $(i XX  Variable Fx90 vvxx  FJEnduringRivalryBorder /x"  x  /  AllOtherBordersforEnduringRivalryStates <xx+"   x <Difference(Std.Error) Axx0'L  xx AObservedProbability 7-'L  xx 7 "x |  "  Mean 5x$ 0 x 5KStandardDeviation ,x x ,  Mean 5x$ 0 x 5StandardDeviation & x &  |    |  %Interaction1 QG T L4@20.30L4@Q20.30 ~tB T3 L4@20.30 L4@ uV3@19.806uV3@~19.806 ~tC T3 uV3@19.806 uV3@ 8@24.008@~24.00 ~tB T3 8@24.00 8@ Mb88@24.219Mb88@~24.219 [QC T3 Mb88@24.219 Mb88@ [-3.70(4.83)# 7- 3 7p=.450 .$ T 3 .Interaction1KM RH(x ! {G^z@421.93{G^z@R421.93 uC(x "3 {G^z@421.93 {G^z@ ̂@848.35̂@848.35 uC(x #3 ̂@848.35 ̂@ Qh@192.21Qh@192.21 uC(x $3 Qh@192.21 Qh@ GzVr@293.38GzVr@293.38 [QC(x %3 GzVr@293.38 GzVr@ [229.72(183.12)# 7-@ '3 7p=.223 .$(x (3 .%Interaction2 PFL ) (\@5.14(\@P5.14 |rAL *3 (\@5.14 (\@ o!@5.033o!@|5.033 |rBL +3 o!@5.033 o!@ ffffff@5.60ffffff@|5.60 |rAL ,3 ffffff@5.60 ffffff@ \(@6.965\(@|6.965 ZPBL -3 \(@6.965 \(@ Z-.45(1.57) 7-d /3 7p=.772 .$L 03 .Interaction2KM QGp 1 QR@74.63QR@Q74.63 ~tBp 23 QR@74.63 QR@ 33333a@136.6033333a@~136.60 ~tCp 33 33333a@136.60 33333a@ GzNA@34.61GzNA@~34.61 }sBp 43 GzNA@34.61 GzNA@ (\/G@46.37(\/G@}46.37 ZPBp 53 (\/G@46.37 (\/G@ Z40.02(29.47)# 7-873 7p=.188 .$p 83 .%Interaction3 QG9  ףp=:R@72.91 ףp=:R@Q72.91 ~tB:3  ףp=:R@72.91  ףp=:R@ y&1,7@23.172y&1,7@~23.172 ~tC;3 y&1,7@23.172 y&1,7@ 333333Q@68.80333333Q@~68.80 ~tB<3 333333Q@68.80 333333Q@ F=@29.097F=@~29.097 [QC=3 F=@29.097 F=@ [4.11(5.70)# 7-\?3 7p=.475 .$@3 .Interaction3KM RHA Q@894.14Q@R894.14 vCB3 Q@894.14 Q@ B`"0@806.017B`"0@806.017 vDC3 B`"0@806.017 B`"0@ (\@643.67(\@643.67 uCD3 (\@643.67 (\@ \(@821.52\(@821.52 [QCE3 \(@821.52 \(@ [250.47(192.02) 7-G3 7p=.196 .$H3 .%Interaction4 PF,I Gz?1.63Gz?P1.63 {qA,J3 Gz?1.63 Gz? ffffff?1.90ffffff?{1.90 {qA,K3 ffffff?1.90 ffffff? ?1.60?{1.60 {qA,L3 ?1.60 ? Q@2.69Q@{2.69 YOA,M3 Q@2.69 Q@ Y.03(.60) 7-O3 7p=.963 .$,P3 .Interaction4KM QGPQ (\(@12.03(\(@Q12.03 }sBPR3 (\(@12.03 (\(@ zGa3@19.38zGa3@}19.38 }sBPS3 zGa3@19.38 zGa3@ Q'@11.56Q'@}11.56 }sBPT3 Q'@11.56 Q'@ Gz8@24.58Gz8@}24.58 ZPBPU3 Gz8@24.58 Gz8@ Z.47(5.58) 7-W3 7p=.933 .$PX3 .WeightedAverageInteraction PF<Z (\@2.57(\@P2.57 {qA$t[3 (\@2.57 (\@ jt?.439jt?{.439 {qA$t\3 jt?.439 jt? Q@2.49Q@{2.49 {qA$t]3 Q@2.49 Q@ Q?.535Q?{.535 YOA$t^3 Q?.535 Q? Y.08(.107)# 7-<`3 7p=.472 .$$ta3 .%Salience1 QGHb )\HX@97.14)\HX@Q97.14 |rBHc3 )\HX@97.14 )\HX@ q= ףp @3.18q= ףp @|3.18 |rAHd3 q= ףp @3.18 q= ףp @ pX@97.75pX@|97.75 |rBHe3 pX@97.75 pX@ Q @3.59Q @|3.59 YOAHf3 Q @3.59 Q @ Y-.61(.826) 7- `h3 7p=.460 .$Hi3 .Salience1KM2 UKl!j  ףpyg@ 120439.59 ףpyg@U120439.59 {Fl!k3  ףpyg@ 120439.59  ףpyg@ ̐@ 109049.05̐@109049.05 zFl!l3 ̐@ 109049.05 ̐@ ffffK@74929.90ffffK@74929.90 yEl!m3 ffffK@74929.90 ffffK@ \H@74052.56\H@74052.56 ]SEl!n3 \H@74052.56 \H@ ]45509.69(24346.62)# 7-4"p3 7p=.073 .$l!q3 .%Salience2 RH#r (?1.9474(?R1.9474 }sC#s3 (?1.9474 (? @2.20@}2.20 {qA#t3 @2.20 @ Q?1.62Q?{1.62 {qA#u3 Q?1.62 Q? @2.50@{2.50 YOA#v3 @2.50 @ Y.32(.575) 7-X$x3 7p=.571 .$#y3 .Salience2KM2 SI%!z Q6@1421.58Q6@S1421.58 wD%!{3 Q6@1421.58 Q6@ {Gѝ@1908.47{Gѝ@1908.47 wD%!|3 {Gѝ@1908.47 {Gѝ@ Q@1057.63Q@1057.63 wD%!}3 Q@1057.63 Q@ Rk9@2588.71Rk9@2588.71 \RD%!~3 Rk9@2588.71 Rk9@ \363.95(583.25) 7-|&!3 7p=.534 .$%!3 .%Salience3 OE'(# ?.80?O.80 zp@'(#3 ?.80 ? Q?1.47Q?z1.47 zpA'(#3 Q?1.47 Q? (\?.58(\?z.58 zp@'(#3 (\?.58 (\? ?1.35?z1.35 YOA'(#3 ?1.35 ? Y.22(.32) 7-(#3 7p=.498 .$'(#3 .Salience3KM2 RH)L% GzVw@373.38GzVw@R373.38 uC)L%3 GzVw@373.38 GzVw@ (\]@523.67(\]@523.67 uC)L%3 (\]@523.67 (\]@ ףp= o@254.97ףp= o@254.97 uC)L%3 ףp= o@254.97 ףp= o@ ؀@539.00؀@539.00 [QC)L%3 ؀@539.00 ؀@ [118.41(125.78) 7-*&3 7p=.348 5x$)L%3 5%Salience4 VxEdx )\(?.11)\(?V.11 xo@d3 )\(?.11 )\(?x Q?.48Q?.48 xo@d3 Q?.48 Q?x ?.05?.05 xo@d3 ?.05 ?x )\(?.22)\(?.22 _xN@d3 )\(?.22 )\(?x _.06 (.06) >x-,3x >p=.347 2(d3x 2XX (iTABLE3Continued#(i XX>K#Ԁ(XX (iAssessingtheBordersofEnduringRivals#(i XXߝK#) 8   ,"(x  ,  Variable Fx90 d  vv  FILEnduringRivalryBorder /x"L  x  /  AllOtherBordersforEnduringRivalryStates 5+" d  x 5LDifference(Std.Error) :0' ,  :ObservedProbability 7-' ,  7   \   Mean .$p   .5NStandardDeviation %8  %  Mean .$p   .NStandardDeviation &8   &   \     \  Salience4KM2 QG ! = ףp=@@32.48= ףp=@@Q32.48 ~tB "3 = ףp=@@32.48 = ףp=@@  ףp= \@112.16 ףp= \@~112.16 ~tC #3  ףp= \@112.16  ףp= \@ Q.@15.46Q.@~15.46 }sB $3 Q.@15.46 Q.@ K@55.25K@}55.25 ZPB %3 K@55.25 K@ Z17.02(24.72)# 7-\ '3 7p=.498 .$ (3 .WeightedAverageSalience PF * p= ף?1.04p= ף?P1.04 zpA +3 p= ף?1.04 p= ף? ?.05?z.05 zp@ ,3 ?.05 ? {Gz?1.03{Gz?z1.03 zpA -3 {Gz?1.03 {Gz? ?.05?z.05 XN@ .3 ?.05 ? X.01(.01) 7- 03 7p=.436 .$ 13 .%Vital1 QG,2 )\(9@25.16)\(9@Q25.16 }sB,33 )\(9@25.16 )\(9@ \(7@23.11\(7@}23.11 }sB,43 \(7@23.11 \(7@ zGa=@29.38zGa=@}29.38 }sB,53 zGa=@29.38 zGa=@ Q+=@29.17Q+=@}29.17 ZPB,63 Q+=@29.17 Q+=@ Z-4.22(5.691)# 7-83 7p=.464 .$,93 .%Vital2 QGP: (\Q@71.09(\Q@Q71.09 }sBP;3 (\Q@71.09 (\Q@ \(6@22.86\(6@}22.86 }sBP<3 \(6@22.86 \(6@ HzP@67.37HzP@}67.37 }sBP=3 HzP@67.37 HzP@ Q<@28.72Q<@}28.72 ZPBP>3 Q<@28.72 Q<@ Z3.72(5.62) 7-@3 7p=.512 .$PA3 .%Vital3 PF$tB {Gz @3.56{Gz @P3.56 {qA$tC3 {Gz @3.56 {Gz @ zG @3.61zG @{3.61 {qA$tD3 zG @3.61 zG @ (\@3.12(\@{3.12 {qA$tE3 (\@3.12 (\@ 333333@4.30333333@{4.30 YOA$tF3 333333@4.30 333333@ Y.44(.983) 7-<H3 7p=.657 .$$tI3 .%Vital4 OEHJ RQ?.19RQ?O.19 yo@HK3 RQ?.19 RQ? = ףp=?.57= ףp=?y.57 yo@HL3 = ףp=?.57 = ףp=? p= ף?.13p= ף?y.13 yo@HM3 p= ף?.13 p= ף? (\?.34(\?y.34 XN@HN3 (\?.34 (\? X.06(.091) 7-`P3 7p=.517 .$HQ3 .WeightedAverageVital PFlR  p= ף?1.79p= ף? P1.79 {qAlS3 p= ף?1.79 p= ף?  \(\?.245\(\? {.245 {qAlT3 \(\?.245 \(\?  ףp= ?1.74ףp= ? {1.74 zpAlU3 ףp= ?1.74 ףp= ?  333333?.30333333? z.30 XN@lV3 333333?.30 333333?   X.05(.06)# 7-4X3  7p=.435 .$lY3 .Length SIZ zG@1402.82zG@S1402.82 wD[3 zG@1402.82 zG@ ̒@1316.70̒@1316.70 vD\3 ̒@1316.70 ̒@ Gz@882.01Gz@882.01 uC]3 Gz@882.01 Gz@ 33333y@943.1533333y@943.15 [QC^3 33333y@943.15 33333y@ [520.81(295.43)# 7-X`3 7p=.090 2(a3  2#UnequalVariances(p>.95))b    )   b XX (iTABLE40 ` AssessingtheBordersofHuthsTerritorialDisputes` (#` (# *xddYdd dd dd dd ~dd >dd edd x(#(#,Ydd ,dd ,dd ,dd ,~dd ,>dd ,edd +  $xx  $#(i XX߬e#  Variable Fx90 vv}xx  Fg  TerritorialDisputeBorder 8x+"   x  8jh  AllOtherBordersforTerritorialDisputeStates <xx+"   x <hDifference(Std.Error) Axx0'L  xx AObservedProbability 7-'L  xx 7 "x |  "  Mean 5x$ 0 x 5)jStandardDeviation ,x x ,  Mean 5x$ 0 x 5jStandardDeviation & x &  |    |  %Interaction1 QG T R7@23.82R7@Q23.82 }sB T3 R7@23.82 R7@ ףp= 8@24.59ףp= 8@}24.59 }sB T3 ףp= 8@24.59 ףp= 8@ Q+3@19.17Q+3@}19.17 }sB T3 Q+3@19.17 Q+3@ L8@24.30L8@}24.30 ZPB T3 L8@24.30 L8@ Z4.65(5.24) 7- 3 7p=.377 .$ T 3 .Interaction1KM RH(x ! Qy@400.02Qy@R400.02 uC(x "3 Qy@400.02 Qy@  @772.10 @772.10 uC(x #3  @772.10  @ Gzh@198.94Gzh@198.94 uC(x $3 Gzh@198.94 Gzh@ u@345.00u@345.00 [QC(x %3 u@345.00 u@ [201.08(152.26)# 7-@ '3 7p=.197 .$(x (3 .%Interaction2 PFL ) (\@4.99(\@P4.99 {qAL *3 (\@4.99 (\@ p= ף@6.66p= ף@{6.66 {qAL +3 p= ף@6.66 p= ף@ Q @3.59Q @{3.59 {qAL ,3 Q @3.59 Q @ (\@5.89(\@{5.89 YOAL -3 (\@5.89 (\@ Y1.40(1.30) 7-d /3 7p=.284 .$L 03 .Interaction2KM QGp 1 \(Q@68.44\(Q@Q68.44 ~tBp 23 \(Q@68.44 \(Q@ ffffff_@125.60ffffff_@~125.60 ~tCp 33 ffffff_@125.60 ffffff_@ q= ף0;@27.19q= ף0;@~27.19 }sBp 43 q= ף0;@27.19 q= ף0;@ lF@44.85lF@}44.85 ZPBp 53 lF@44.85 lF@ Z41.26(24.55)# 7-873 7p=.104 .$p 83 .%Interaction3 QG9 )\xQ@69.89)\xQ@Q69.89 }sB:3 )\xQ@69.89 )\xQ@ (\u<@28.46(\u<@}28.46 }sB;3 (\u<@28.46 (\u<@ RS@76.03RS@}76.03 }sB<3 RS@76.03 RS@ Gz;@27.83Gz;@}27.83 ZPB=3 Gz;@27.83 Gz;@ ZЄ6.14(6.02) 7-\?3 7p=.309 .$@3 .Interaction3KM RHA zG@895.16zG@R895.16 uCB3 zG@895.16 zG@ = ףpc@748.43= ףpc@748.43 uCC3 = ףpc@748.43 = ףpc@ RA@680.24RA@680.24 uCD3 RA@680.24 RA@ 33333Ո@794.6533333Ո@794.65 [QCE3 33333Ո@794.65 33333Ո@ [214.91(169.08) 7-G3 7p=.206 .$H3 .%Interaction4 PF,I p= ף?1.29p= ף?P1.29 {qA,J3 p= ף?1.29 p= ף? )\(?1.76)\(?{1.76 {qA,K3 )\(?1.76 )\(? Q?1.22Q?{1.22 {qA,L3 Q?1.22 Q? ףp= @2.48ףp= @{2.48 YOA,M3 ףp= @2.48 ףp= @ Y.08(.51) 7-O3 7p=.878 .$,P3 .Interaction4KM QGPQ Q/@15.86Q/@Q15.86 }sBPR3 Q/@15.86 Q/@ Gz=@29.83Gz=@}29.83 |rBPS3 Gz=@29.83 Gz=@ HzG @8.14HzG @|8.14 |rAPT3 HzG @8.14 HzG @ Q4@20.52Q4@|20.52 |rBPU3 Q4@20.52 Q4@ zG@7.72zG@|7.72 YOAPV3 zG@7.72 zG@ Yp=.115 .$PW3 .WeightedAverageInteraction PF$tY Q@2.49Q@P2.49 zpA\Z3 Q@2.49 Q@ HzG?.54HzG?z.54 zp@\[3 HzG?.54 HzG? p= ף@2.58p= ף@z2.58 zpA\\3 p= ף@2.58 p= ף@ p= ף?.52p= ף?z.52 XN@\]3 p= ף?.52 p= ף? XЄ.10(.11) 7-$t_3 7p=.399 .$\`3 .%Salience1 QGa q= ףX@98.51q= ףX@Q98.51 |rBb3 q= ףX@98.51 q= ףX@ (\?1.66(\?|1.66 |rAc3 (\?1.66 (\? X@98.25X@|98.25 |rBd3 X@98.25 X@  @3.20 @|3.20 YOAe3  @3.20  @ Y.27(.64) 7-Hg3 7p=.677 .$h3 .Salience1KM2 UK i G~@ 114743.93G~@U114743.93 zF j3 G~@ 114743.93 G~@ ( <@95168.76( <@95168.76 yE k3 ( <@95168.76 ( <@ p= @77705.44p= @77705.44 yE l3 p= @77705.44 p= @ zw0@74503.48zw0@74503.48 ]SE m3 zw0@74503.48 zw0@ ]37038.48(16991.58) 7-l!o3 7p=.031 .$ p3 .%Salience2 PF"q ?1.25?P1.25 {qA"r3 ?1.25 ? 333333?1.45333333?{1.45 {qA"s3 333333?1.45 333333? Gz?1.23Gz?{1.23 {qA"t3 Gz?1.23 Gz? Q@2.34Q@{2.34 YOA"u3 Q@2.34 Q@ Y.03(.47) 7-#w3 7p=.954 .$"x3 .Salience2KM2 SI$< y Gz@1286.37Gz@S1286.37 wD$< z3 Gz@1286.37 Gz@ RT@1941.23RT@1941.23 vD$< {3 RT@1941.23 RT@ Qފ@859.84Qފ@859.84 vC$< |3 Qފ@859.84 Qފ@ ףp=̣@2534.27ףp=̣@2534.27 \RD$< }3 ףp=̣@2534.27 ףp=̣@ \426.53(522.79) 7-%!3 7p=.416 .$$< 3 .%Salience3 OE'`" )\(?.22)\(?O.22 yo@'`"3 )\(?.22 )\(? ?.25?y.25 yo@'`"3 ?.25 ? q= ףp?.46q= ףp?y.46 zp@'`"3 q= ףp?.46 q= ףp? RQ?1.27RQ?z1.27 YOA'`"3 RQ?1.27 RQ? YЄ.24(.13)# 7-'(#3 7p=.064 .$'`"3 .Salience3KM2 RH4)$ Gzk@221.44Gzk@R221.44 uC4)$3 Gzk@221.44 Gzk@ QWv@357.47QWv@357.47 uC4)$3 QWv@357.47 QWv@ Qk@221.86Qk@221.86 uC4)$3 Qk@221.86 Qk@ = ףpE@552.68= ףpE@552.68 [QC4)$3 = ףpE@552.68 = ףpE@ [Є.42(111.95) 7-)L%3 7p=.997 5x$4)$3 5%Salience4 VxEX+&x {Gz?.02{Gz?V.02 xo@X+&3 {Gz?.02 {Gz?x {Gz?.08{Gz?.08 xo@X+&3 {Gz?.08 {Gz?x Q?.06Q?.06 xo@X+&3 Q?.06 Q?x {Gz?.32{Gz?.32 _xN@X+&3 {Gz?.32 {Gz?x _Є.04(.06) >x- ,p'3x >p=.570 2(X+&3x 2XX (iTABLE4Continued#(i XX #Ԁ(XX (iAssessingtheBordersofHuthsTerritorialDisputes#(i XXj#) d  ,"T ,  Variable Fx90@ vv   F/  TerritorialDisputeBorder 8x+"(x  x  8  AllOtherBordersforTerritorialDisputeStates 5+"@   x 5 Difference(Std.Error) :0'    :ObservedProbability 7-'    7   8   Mean .$L   .EStandardDeviation % d %  Mean .$L   .ʛStandardDeviation & d &   8    8 Salience4KM2 QGp  {Gz'@11.74{Gz'@Q11.74 }sBp 3 {Gz'@11.74 {Gz'@ Q@@33.99Q@@}33.99 }sBp 3 Q@@33.99 Q@@ = ףp/@15.87= ףp/@}15.87 }sBp 3 = ףp/@15.87 = ףp/@ HzWQ@69.37HzWQ@}69.37 ZPBp 3 HzWQ@69.37 HzWQ@ ZЄ4.13(13.77) 7-8 3 7p=.765 .$p  3 .WeightedAverageSalience PF\ " RQ?1.02RQ?P1.02 zpA #3 RQ?1.02 RQ? {Gz?.02{Gz?z.02 zp@ $3 {Gz?.02 {Gz? RQ?1.02RQ?z1.02 zpA %3 RQ?1.02 RQ? ?.05?z.05 XN@ &3 ?.05 ? XЄ.01(.01) 7-\ (3 7p=.519 .$ )3 .%Vital1 QG * )\<@28.66)\<@Q28.66 }sB +3 )\<@28.66 )\<@ q= ף<@28.69q= ף<@}28.69 }sB ,3 q= ף<@28.69 q= ף<@ Q6@22.62Q6@}22.62 }sB -3 Q6@22.62 Q6@ zG;@27.88zG;@}27.88 ZPB .3 zG;@27.88 zG;@ Z6.04(6.03) 7- 03 7p=.319 .$ 13 .%Vital2 QG,2 q= ף@Q@69.01q= ף@Q@Q69.01 }sB,33 q= ף@Q@69.01 q= ף@Q@ q= ף0<@28.19q= ף0<@}28.19 }sB,43 q= ף0<@28.19 q= ף0<@ QR@74.83QR@}74.83 }sB,53 QR@74.83 QR@ Gzn;@27.43Gzn;@}27.43 ZPB,63 Gzn;@27.43 Gzn;@ ZЄ5.82(5.93) 7-83 7p=.329 .$,93 .%Vital3 PFP: RQ@2.29RQ@P2.29 {qAP;3 RQ@2.29 RQ@ Gz@2.46Gz@{2.46 {qAP<3 Gz@2.46 Gz@ q= ףp@2.43q= ףp@{2.43 {qAP=3 q= ףp@2.43 q= ףp@ \(\@4.09\(\@{4.09 YOAP>3 \(\@4.09 \(\@ YЄ.15(.82) 7-@3 7p=.860 .$PA3 .%Vital4 OE$tB ?.05?O.05 yo@$tC3 ?.05 ?  ףp= ?.09 ףp= ?y.09 yo@$tD3  ףp= ?.09  ףp= ? )\(?.11)\(?y.11 yo@$tE3 )\(?.11 )\(? RQ?.38RQ?y.38 XN@$tF3 RQ?.38 RQ? XЄ.07(.07) 7-<H3 7p=.372 .$$tI3 .WeightedAverageVital PFHJ  ףp= ?1.74ףp= ? P1.74 zpAHK3 ףp= ?1.74 ףp= ?  (\?.29(\? z.29 zp@HL3 (\?.29 (\?  p= ף?1.79p= ף? z1.79 zpAHM3 p= ף?1.79 p= ף?  (\?.29(\? z.29 XN@HN3 (\?.29 (\?   XЄ.06(.062) 7-`P3  7p=.363 .$HQ3 .Length SIlR Q@1379.63Q@S1379.63 wDlS3 Q@1379.63 Q@  ףp=S@1172.81 ףp=S@1172.81 vDlT3  ףp=S@1172.81  ףp=S@ Q@914.44Q@914.44 uClU3 Q@914.44 Q@ (\=@935.72(\=@935.72 [QClV3 (\=@935.72 (\=@ [465.19(212.27) 7-4X3 7p=.030 2(lY3  2#UnequalVariances(p>.95))Z    )  Z XX (iTABLE50 ` AssessingtheBordersofMilitarizedInterstateDisputes` (#` (# *xddYdd dd dd dd ~dd >dd edd x(#(#,Ydd ,dd ,dd ,dd ,~dd ,>dd ,edd +  $xx  $#(i XXߜ#  Variable Fx90 vv?3xx  F  MilitarizedInterstateDisputeBorder 8x+"   x  8[  AllOtherBordersforMilitarizedInterstateDisputeStates <xx+"   x <ԵDifference(Std.Error) Axx0'L  B3xx AObservedProbability 7-'L  C3xx 7 "x |  "  Mean 5x$ 0 x 50StandardDeviation ,x x ,  Mean 5x$ 0 x 5÷StandardDeviation & x &  |    |  %Interaction1 QG T Gz3@19.08Gz3@Q19.08 }sB T3 Gz3@19.08 Gz3@ q= ף04@20.19q= ף04@}20.19 }sB T3 q= ף04@20.19 q= ף04@ ףp= W3@19.34ףp= W3@}19.34 }sB T3 ףp= W3@19.34 ףp= W3@ 6@22.606@}22.60 ZPB T3 6@22.60 6@ ZЄ.26(3.28) 7- 3 7p=.937 .$ T 3 .Interaction1KM RH(x ! (\o@255.18(\o@R255.18 uC(x "3 (\o@255.18 (\o@ R?@551.99R?@551.99 uC(x #3 R?@551.99 R?@ Rb@149.56Rb@149.56 uC(x $3 Rb@149.56 Rb@ q@280.00q@280.00 [QC(x %3 q@280.00 q@ [105.62(73.87)# 7-@ '3 7p=.157 .$(x (3 .%Interaction2 PFL ) Gz@5.27Gz@P5.27 {qAL *3 Gz@5.27 Gz@ Q@7.13Q@{7.13 {qAL +3 Q@7.13 Q@ 333333@4.55333333@{4.55 {qAL ,3 333333@4.55 333333@ 333333@6.30333333@{6.30 YOAL -3 333333@6.30 333333@ Y.71(.98) 7-d /3 7p=.464 .$L 03 .Interaction2KM QGp 1 lI@50.85lI@Q50.85 }sBp 23 lI@50.85 lI@ YW@93.40YW@}93.40 }sBp 33 YW@93.40 YW@ 8@24.758@}24.75 }sBp 43 8@24.75 8@ = ףp}B@36.98= ףp}B@}36.98 ZPBp 53 = ףp}B@36.98 = ףp}B@ Z26.10(12.29)# 7-873 7p=.037 .$p 83 .%Interaction3 QG9 lR@73.70lR@Q73.70 }sB:3 lR@73.70 lR@ 9@25.509@}25.50 }sB;3 9@25.50 9@ Q+R@72.68Q+R@}72.68 }sB<3 Q+R@72.68 Q+R@ Q:@26.67Q:@}26.67 ZPB=3 Q:@26.67 Q:@ Z1.02(3.94) 7-\?3 7p=.795 .$@3 .Interaction3KM RHA {G@733.86{G@R733.86 uCB3 {G@733.86 {G@ (\#@676.47(\#@676.47 uCC3 (\#@676.47 (\#@ {G@568.36{G@568.36 uCD3 {G@568.36 {G@ I@681.20I@681.20 [QCE3 I@681.20 I@ [165.50(101.15) 7-G3 7p=.104 .$H3 .%Interaction4 PF,I 333333?1.95333333?P1.95 {qA,J3 333333?1.95 333333? (\@3.87(\@{3.87 {qA,K3 (\@3.87 (\@ \(\ @3.42\(\ @{3.42 {qA,L3 \(\ @3.42 \(\ @ = ףp=@6.31= ףp=@{6.31 YOA,M3 = ףp=@6.31 = ףp=@ YЄ1.47(.69)# 7-O3 7p=.035 .$,P3 .Interaction4KM QGPQ p= ף)@12.82p= ף)@Q12.82 }sBPR3 p= ף)@12.82 p= ף)@ Q:@26.02Q:@}26.02 }sBPS3 Q:@26.02 Q:@ ףp= W-@14.67ףp= W-@}14.67 }sBPT3 ףp= W-@14.67 ףp= W-@ p= ;@27.89p= ;@}27.89 ZPBPU3 p= ;@27.89 p= ;@ ZЄ1.85(4.09) 7-W3 7p=.651 .$PX3 .WeightedAverageInteraction PF<Z p= ף@2.58p= ף@P2.58 zpA$t[3 p= ף@2.58 p= ף@ ?.45?z.45 zp@$t\3 ?.45 ? @2.60@z2.60 zpA$t]3 @2.60 @ RQ?.51RQ?z.51 XN@$t^3 RQ?.51 RQ? XЄ.02(.07) 7-<`3 7p=.771 .$$ta3 .%Salience1 QGHb \(\X@97.44\(\X@Q97.44 |rBHc3 \(\X@97.44 \(\X@ Gz@4.02Gz@|4.02 |rAHd3 Gz@4.02 Gz@ Q+X@96.68Q+X@|96.68 |rBHe3 Q+X@96.68 Q+X@ @5.75@|5.75 YOAHf3 @5.75 @ Y.76(.68)# 7- `h3 7p=.264 .$Hi3 .Salience1KM2 TJl!j p= k@88790.69p= k@T88790.69 yEl!k3 p= k@88790.69 p= k@ )\It@79684.61)\It@79684.61 yEl!l3 )\It@79684.61 )\It@ 3333=@63982.103333=@63982.10 yEl!m3 3333=@63982.10 3333=@ (\@63710.58(\@63710.58 ]SEl!n3 (\@63710.58 (\@ ]24808.59(11312.34)# 7-4"p3 7p=.031 .$l!q3 .%Salience2 PF#r 333333?1.95333333?P1.95 {qA#s3 333333?1.95 333333? ףp= @3.23ףp= @{3.23 {qA#t3 ףp= @3.23 ףp= @ RQ@2.54RQ@{2.54 {qA#u3 RQ@2.54 RQ@  ףp= @4.51 ףp= @{4.51 YOA#v3  ףp= @4.51  ףp= @ YЄ.60(.539)# 7-X$x3 7p=.269 .$#y3 .Salience2KM2 SI%!z \('@1097.79\('@S1097.79 wD%!{3 \('@1097.79 \('@ zGə@1650.32zGə@1650.32 wD%!|3 zGə@1650.32 zGə@ (\<@1039.14(\<@1039.14 wD%!}3 (\<@1039.14 (\<@ q= #@2374.57q= #@2374.57 \RD%!~3 q= #@2374.57 q= #@ \58.64(329.53) 7-|&!3 7p=.859 .$%!3 .%Salience3 OE'(# ?.55?O.55 zp@'(#3 ?.55 ? ?1.10?z1.10 zpA'(#3 ?1.10 ? Q?.71Q?z.71 zp@'(#3 Q?.71 Q? \(\?1.46\(\?z1.46 YOA'(#3 \(\?1.46 \(\? YЄ.16(.21) 7-(#3 7p=.438 .$'(#3 .Salience3KM2 RH)L% p@269.60p@R269.60 uC)L%3 p@269.60 p@ HzG|@452.48HzG|@452.48 uC)L%3 HzG|@452.48 HzG|@ q= ףm@239.52q= ףm@239.52 uC)L%3 q= ףm@239.52 q= ףm@ \(@507.01\(@507.01 [QC)L%3 \(@507.01 \(@ [30.08(73.63) 7-*&3 7p=.683 5x$)L%3 5%Salience4 VxEdx Q?.06Q?V.06 xo@d3 Q?.06 Q?x (\?.29(\?.29 xo@d3 (\?.29 (\?x Q?.06Q?.06 xo@d3 Q?.06 Q?x zG?.21zG?.21 _xN@d3 zG?.21 zG?x _Є.00 (.04) >x-,3x >p=.990 2(d3x 2XX (iTABLE5Continued#(i XX#Ԁ(XX (iAssessingtheBordersofMilitarizedInterstateDisputes#(i XXV#) 8   ,"(x  ,  Variable Fx90 d  vv>  F  MilitarizedInterstateDisputeBorder 8x+"L   x  8~  AllOtherBordersforMilitarizedInterstateDisputeStates 5+" d  x 5Difference(Std.Error) :0' , > :ObservedProbability 7-' , > 7   \   Mean .$p   .>StandardDeviation %8  %  Mean .$p   .StandardDeviation &8   &   \     \  Salience4KM2 QG ! Hz6@22.53Hz6@Q22.53 }sB "3 Hz6@22.53 Hz6@ 33333R@74.3033333R@}74.30 }sB #3 33333R@74.30 33333R@ Q0@16.52Q0@}16.52 }sB $3 Q0@16.52 Q0@ \(J@53.22\(J@}53.22 ZPB %3 \(J@53.22 \(J@ Z6.01(8.88) 7-\ '3 7p=.499 .$ (3 .WeightedAverageSalience PF * {Gz?1.03{Gz?P1.03 zpA +3 {Gz?1.03 {Gz? ?.05?z.05 zp@ ,3 ?.05 ? p= ף?1.04p= ף?z1.04 zpA -3 p= ף?1.04 p= ף? {Gz?.08{Gz?z.08 XN@ .3 {Gz?.08 {Gz? XЄ.01(.01)# 7- 03 7p=.185 .$ 13 .%Vital1 QG,2 8@24.108@Q24.10 }sB,33 8@24.10 8@ {Gz9@25.48{Gz9@}25.48 }sB,43 {Gz9@25.48 {Gz9@ )\7@23.66)\7@}23.66 }sB,53 )\7@23.66 )\7@ fffff:@26.65fffff:@}26.65 ZPB,63 fffff:@26.65 fffff:@ Z.44(3.93) 7-83 7p=.911 .$,93 .%Vital2 QGP: QR@72.23QR@Q72.23 }sBP;3 QR@72.23 QR@ 9@25.509@}25.50 }sBP<3 9@25.50 9@ QQ@70.83QQ@}70.83 }sBP=3 QQ@70.83 QQ@ (\:@26.81(\:@}26.81 ZPBP>3 (\:@26.81 (\:@ Z1.40(3.95) 7-@3 7p=.723 .$PA3 .%Vital3 PF$tB Gz @3.51Gz @P3.51 {qA$tC3 Gz @3.51 Gz @ \(\@5.84\(\@{5.84 {qA$tD3 \(\@5.84 \(\@ (\@5.24(\@{5.24 {qA$tE3 (\@5.24 (\@ !@8.55!@{8.55 YOA$tF3 !@8.55 !@ YЄ1.74(.994)# 7-<H3 7p=.083 .$$tI3 .%Vital4 OEHJ (\?.17(\?O.17 yo@HK3 (\?.17 (\? ?.45?y.45 yo@HL3 ?.45 ? p= ף?.26p= ף?y.26 yo@HM3 p= ף?.26 p= ף? {Gz?.64{Gz?y.64 XN@HN3 {Gz?.64 {Gz? XЄ.10(.08)# 7-`P3 7p=.207 .$HQ3 .WeightedAverageVital PFlR  p= ף?1.79p= ף? P1.79 zpAlS3 p= ף?1.79 p= ף?  HzG?.27HzG? z.27 zp@lT3 HzG?.27 HzG?  Q?1.82Q? z1.82 zpAlU3 Q?1.82 Q?  333333?.30333333? z.30 XN@lV3 333333?.30 333333?   XЄ.03(.04) 7-4X3  7p=.504 .$lY3 .Length SIZ )\r@1052.74)\r@S1052.74 vD[3 )\r@1052.74 )\r@ Ql@973.54Ql@973.54 uC\3 Ql@973.54 Ql@ q= ף@757.33q= ף@757.33 uC]3 q= ף@757.33 q= ף@ @785.95@785.95 [QC^3 @785.95 @ [295.40(138.46) 7-X`3 7p=.036 2(a3  2#UnequalVariances(p>.95))b    )