Traditionalbiomechanicalanalysesofhumanmovementaregenerallyderivedfromlinearmathematics.Whilethesemethodscanbeusefulinmanysituations,theydonotdescribebehaviorsinhumansystemsthatarepredominatelynonlinear.Forthisreason,nonlinearanalysismethodsbasedonadynamicalsystemsapproachhavebecomemoreprevalentinrecentliterature.Theseanalysistechniqueshaveprovidednewinsightsintohowsystems(1)maintainpatternstability,(2)transitionintonewstates,and(3)aregovernedbyshort-andlong-term(fractal)correlationalprocessesatdifferentspatio-temporalscales.Thesedifferentaspectsofsystemdynamicsaretypicallyinvestigatedusingconceptsrelatedtovariability,stability,complexity,andadaptability.Thepurposeofthispaperistocompareandcontrastthesedifferentconceptsanddemonstratethat,althoughrelated,thesetermsrepresentfundamentallydifferentaspectsofsystemdynamics.Inparticular,wearguethatvariabilityshouldnotuniformlybeequatedwithstabilityorcomplexityofmovement.Inaddition,currentdynamicstabilitymeasuresbasedonnonlinearanalysismethods(suchasthefinitemaximalLyapunovexponent)canreveallocalinstabilitiesinmovementdynamics,butthedegreetowhichtheselocalinstabilitiesrelatetoglobalposturalandgaitstabilityandtheabilitytoresistexternalperturbationsremainstobeexplored.Finally,systematicstudiesareneededtorelateobservedreductionsincomplexitywithaginganddiseasetotheadaptivecapabilitiesofthemovementsystemandhowcomplexitychangesasafunctionofdifferenttaskconstraints.
