Historia Del Pta T2

Complexity in Physical Geography Author(s): ANDREW RICHARDS Source: Geography, Vol. 87, No. 2 (April 2002), pp. 99-107 Published by: Geographical Association Stable URL: http://www.jstor.org/stable/40573664 Accessed: 18-06-2015 15:57 UTC REFERENCES Linked references are available on JSTOR for this article: http://www.jstor.org/stable/40573664?seq=1&cid=pdf-reference#references_tab_contents You may need to log in to JSTOR to access the linked references.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Geographical Association is collaborating with JSTOR to digitize, preserve and extend access to Geography.

http://www.jstor.org

This content downloaded from 168.176.55.24 on Thu, 18 Jun 2015 15:57:36 UTC All use subject to JSTOR Terms and Conditions

in Physical Complexity Geography ANDREW RICHARDS

to ABSTRACT: attempt Physicalgeographers understandthebehaviourofEarthsurface However,withtheadventofhighsystems. and themeansofcollecting poweredcomputers and analysingvastamountsofdata, scientists are realisingthatclassicalprinciplesare often inadequatein explainingtheformsand processespresentin thenaturalworld.Thefield includesa varietyoftopicsthat ofcomplexity havebecomeprominentresearchareas in thegrowthof recentyears.Thisarticlehighlights and within physicalgeography complexity assessestheextentto whichthenewframework newparadigm can be viewedas an evolving, in research physicalgeography for

Introduction

Throughoutthe earlystages of work on GEOGRAPHY werewillingto manyscientists complexsystems, dismiss non-linearbehaviour and emergent VOLUME 87(2) to faulty PAGES as extraneous'noise',attributed 99-107 patterns ofthe equipmentoran inadequateunderstanding system being studied. However, on closer the behaviourand resultsof nonexamination, linear systems is often characterised by have foundthat, unexpectedpatterns.Scientists in orderto fullyunderstand suchsystems, a new whichexpandon classical, Geography©2002 setoflawsis necessary Newtonian principles. is particularly susceptibleto this Geography newtrainofthought. the Geographers synthesise of and human systems understanding physical derivedfromthe pure and humansciencesand anddynamics analyseanddescribetheinteraction The purposeofthisarticleis to ofthesesystems. summarisethe developmentof the complexity itsinfluenceon present paradigm and highlight in physicalgeography. thoughtand methodology A numberof examples will be given,which illustrate theneed fora widerunderstanding and criticalexaminationof the basic tenetsof the developingframework. Finally,the articlewill discuss whether the paradigm is likely to researchand revolutionise physicalgeographical discussthepossibleimpacton physical geography as an academicsubject.

IS ONE of thefastestgrowingand COMPLEXITY pervasivebranchesof general science. It has Whatis complexity? witha new set of rulesthat providedscientists havebothdevelopedand contradicted long-held thatformthebasisofthepuresciences. Manson (2001) recognises three forms of principles The suddeninterestin systemsthatexhibit complexity:'algorithmic','deterministic'and has stemmed 'aggregate' complexity.However, the author non-linear, dynamics unpredictable For would argue that complexitymay be viewed and fromworkin biology, physics chemistry. and Belousov-Zhabotinskii the Landolt the continuumbetween example, simplyas incorporating (B-Z) 'clock'reactionswherecertaincocktailsof 'order'and 'chaos' (Figure1). Atone end of the chemicalsrefuseto settleinto an energetically scale, it is concernedwith apparentlyorderly stable state and periodicallyoscillatebetween systemsthat resultin verycomplex products, differentcolours (see Scott (1994) for an whileat theotheritis connectedwiththestudyof ofthiswork)andworkin the verycomplexsystemsor therandominteraction accessiblesummary latterpartsofthenineteenth which century byLiapunov betweena numberofdeterministic systems and Poincareon chaoswithincelestialmechanics yieldregulatedproducts. (see thegeneraldiscussionin Hall (1991)). While suchworkspansover100years,theterminology Deterministic chaos and methods employed when studying Chaos theoryinvolvesdeterministic systemsthat Whilewe may havebeen definedrelatively randomresults. recently produceapparently complexity thehuman fullyquantifyand understandthe inputsinto hasinfiltrated Thestudyofcomplexity to sciencesand has resultedin the publicationof certaindeterministic theyaresensitive systems, at the and interact with books aimed initial conditions science other, general may best-selling readeronlyin the last few decades (Waldrop, apparentlydeterministic systemsto produce 99 to notethatchaotic 1995; chaoticresults.Itis important 1992;Lewin,1994;Coveneyand Highfield, in When viewed are random. Solee/tf/., patterns apparently 1999).

This content downloaded from 168.176.55.24 on Thu, 18 Jun 2015 15:57:36 UTC All use subject to JSTOR Terms and Conditions

GEOGRAPHY IN COMPLEXITY PHYSICAL GEOGRAPHY

Geography0 2002

ofthecomplexity representation Figure1:A simplified and themajorresearchareas.Self-organised continuum theareawithin or the'edgeofchaos",represents criticality, chaos (cf.Phillips, and deterministic whichself-organisation orderstartto and qualitative 1998)meetandbothquantitative systems. emergeindynamical

or detailthereis oftenan underlying quantitative qualitativepattern,whichmaybe describedby scale-invariant, fractal dimensions. However, while underlying patternsof behaviourcan be orquantitative describedand thesamequalitative thepatterns describe be used to can relationships ofthescaleatwhichtheyareviewed, irrespective discretestatesof to determine itis difficult future, the system accurately.In chaotic systems behaviourat anypointin timemaybe as mucha ofpreviousstatesas a productofpresent function Such systemstypically state or environment. stateand morethanone equilibrium incorporate the systemmayflipbetweenthe equilibriaas a inthesystem'sstate resultofminormodifications variables.This meansthat,forexactlythe same predictthe input,itis impossibleto reproducibly an occurrencewhichmay outputsofthesystem, be unnerving to academicswho are comfortable ina worldthattheyfeelis adequatelydescribedby classicalprinciples.

Self-organisation

100

variablesthat have resulted in the ordered of outputs.As a consequence, understanding process-responserelationshipsis limitedand prediction of future behaviour becomes problematic. A formof self-organisation, self-organised criticality(SOC), occurs where the internal dynamicsof a systemorganisesitselfinto a state(Bak and Chen,1991).Bak pervasivecritical etal. (1987,1988)and Bakand Chen(1991) have used thedevelopmentof a sand-pileas a simple of physicalmodel to explainthe characteristics is on A of sand placed criticality.pile self-organised a circulartable. Grainsof sand are randomly heightuntil droppedontothepilefroma uniform the pile reachesa criticalangle,or the angleof angle,the repose.Astheslope reachesthecritical additionof sand grains triggerslandslidesof varioussizes. The systemmaynot changeat all, or more, theremay be minorrearrangements rarely,a sand grain may cause complete and failurewithinthe sand pile. disorganisation and magnitudeof the 'landslides' The frequency whichresultas thesystemapproachesitscritical state is fractal,and follows a power-law distribution (Turcotte,1993). Thus the average ofa givensize ofavalancheis inversely frequency proportionalto some power of its magnitude. ofevents and frequency themagnitude Although it is practically can be described statistically, impossibleto predictthe preciseeffectsof the sand grain.A fractalor additionof a particular has no inherentlength power-lawdistribution and therelationship scale (i.e. itis scale-invariant) between the frequencyand size of events is expressedby the fractaldimensionD. Physical systemsthatexhibitSOC producesignalsthatare characterised by 'coloured' 1/fnoise,or 'flicker noise' and the variationsin frequencyhave orderratherthanbeing randomor underlying the 'white'noise (Bak et al, 1987). Interestingly, of scientific subjects, including development whenviewedintermsofthevolumeof geography, withina specificframework, written publications size distributions can be describedbypower-law (vanRaan,2000).

occurs when a complex or Self-organisation stochastic system produces an ordered or regulatedproduct.The termwas firstused in associationwithchemicalsystems(Nicolisand Prigogine,1989) followingearlier work by Landolt, Belousov and Zhabotinskiiwhich involvedthe developmentof a systemfroman unpatternedto a patternedstate withoutthe interventionof an external control. The spontaneous 'emergence*of regularpatterns non-linear oftenoccursthrough processesor as a Examplesofcomplexity reactionto the couplingof processeswithina Such inphysicalgeography Sole et al., 1994; 1999). system(Kauffman, behaviour is interestingas it self-organising impliestheactionof 'universallaws'that,as yet, who The followingsection outlines a number of Forthescientist understand. we do notfully of describetheapplicability the productsof complex studieswhichbriefly is tryingto interpret to determinethe complexity to physical geography. These systems,it is very difficult

This content downloaded from 168.176.55.24 on Thu, 18 Jun 2015 15:57:36 UTC All use subject to JSTOR Terms and Conditions

examples also underlinethe fact that while deterministic chaos has been the subject of researchforover50 years,self-organisation and related concepts have infiltrated research in physicalgeography relatively recently, principally inthelastdecade. Deterministic chaos Meteorology One of theearlieststudiesin therealmofchaos andaccessibleexampleof providesa well-studied chaoticpatternsproducedbynaturalsystems.It has long been recognisedthatthe globalwind whichaffect thenorthern and southern patterns are producedbytherotation ofthe hemispheres Earth and as a result of the atmosphere theglobalenergybudget;a surplusof controlling heatenergyneartheEarth'sequatorand a deficit of heat energytowardsthe poles. The styleof energytransportvaries accordingto position aroundtheglobe.Fromtheequatorto thepoles, systems, energyis transferred bylargeconvective knownas Hadleycells. This cellularpatternof involveswarmair risingat the energytransfer towardsthe poles in the equatorand travelling beforedescending upperpartsofthetroposphere at the level of the sub-tropicalanticyclones. inthemid-latitudes is muchmore transfer Energy by complex.Here,thermalenergyis transferred transfer air RossbyWaves.These wavestructures massesacrossthemid-latitudes, pushingwarmer airmassestowardsthepolesandcolderairmasses towardsthe tropics.Often,where there is a betweenthepressureand pronounceddifference of the tropicaland characteristics temperature polar air, steep pressuregradientscause air masses to be moved more quicklyby the ofjetstreamswiththeRossbyWaves. interaction the EdwardLorenz attemptedto transfer lawsofphysicstoa computermodelthatcouldbe used to model energytransferin the midlatitudes, and, as a result, predict the weatherpatterns developmentof mid-latitude such as depressionsand anticyclones.Lorenz discoveredchaoticpatternsbyaccident(Lorenz, 1963). Using his computer model of the Lorenzhad produceda patternof atmosphere, futurestatesfroman initialvariable,0.506127, describing the position, temperatureand ofa bodyofair.His later pressurecharacteristics butroundedup usedthesamefigure, experiment to threedecimalplaces, 0.506. Lorenzdid not on expectsucha minutechangeto haveanyeffect the futurestates of the system,the patterns producedshouldhaveduplicatedthoseproduced

However,the new by the earlierexperiments. patternsquicklydivergedfrom the patterns producedbytheearlierrun.Thesmallchangehad resulted in the productionof a completely different patternof weatherconditions.Lorenz later used a simple analogyto describe this to initialconditions; thebutterfly sensitivity effect. Ifa butterfly flapsitswingsin thetropicalforests ofBrazil,itmayhaveno effect on theatmosphere. this Alternatively, apparentlysmall change in atmosphericconditionsmay resultin a chain reactionofcatastrophic proportions, producinga thunderstorm overChicago.

GEOGRAPHY IN COMPLEXITY PHYSICAL GEOGRAPHY

Geography©2002

Ecology A furtherearly example of chaotic patterns produced by a sequence of relatively simple, withR.M. deterministic equations,is illustrated May's models of population dynamics(May, 1976). Manyorganisms, includingfish,birdsand some pests,havea definite breedingseason.May attempted to model the growth of such populationsfromseason to season throughthe use ofthelogisticequation.

= rX Xnext (1-X) is thenextyear'spopulation;r is the where:xnext growthrate(whichtakesintoaccountmortality rate,etc.)andx is theprevious rate,reproduction year's population.(1-x) keeps the population within bounds, between 0 and 1, with 0 the extinction and 1 representing representing maximum carrying capacityoftheecosystem. Mayfoundthatwhenlow growthratesare used (Figure2) the populationbehaves as we and then would expect,i.e. initially increasing, settlingto an equilibrium,or stable state. However,whenvaluesof r exceed 3, fluctuating high and low values are produced around an equilibrium.When r exceeds 3.5, a chaotic population patterndevelops throughperiod a population doublingbifurcation,i.e. initially showa two-year wouldfirst cycle,thena four-year cycle(andso on). cycle,thenan 8-,16-and32-year Beyonda certainpoint,Mayfoundthattheperiod brokedown;the completely doublingbifurcation systemno longerachievedanyformofperiodicity, and developedapparentlyrandombehaviouror chaos. Self-organisation Geomorphology Many of the most strikingexamples of selfat the occurintheformoflandforms organisation

This content downloaded from 168.176.55.24 on Thu, 18 Jun 2015 15:57:36 UTC All use subject to JSTOR Terms and Conditions

101

GEOGRAPHY IN COMPLEXITY PHYSICAL GEOGRAPHY

0.0 I

1

Geography© 2002

i

i

1

1

1

1

1

1

1.0-1

0.0-1

1

O.OH

1

1

11 21 31 41 51 61 71 81 91 Year

1

1

1

1

1

1

1

1

1

11 21 31 41 51 61 71 81 91 Year

1.0-1

1

1.0-1 - i

'

0.0-1 -

1

i

i

i

i

i

i

i

i

i

11 21 31 41 51 61 71 81 91 Year 1

|-i

^- i

1

1

1

1

1

1

1

1-

' i

1

*

i

11 21 31 41 51 61 71 81 91 Year

1

0.9- (c) 0.8-

¿¿ 18-0.4°-

0.30.20.1-

O.O-I 1

, 11

, 21

, 31

1 41

, 51 Year

1 61

, 71

. 81

1 91

1

an initialpopulationof0.6. usingthelogisticequation(May,1976),employing Figure2: Simulated pestpopulationfluctuations thepopulation, whichdwindlestowardextinction to maintain Whena growth rateof1 is used,therateis insufficient (a). Witha andsettlesto an equilibrium of3 (c) and 3.5 (d), rateof2, thepopulationincreases, fluctuates, (b). Atfigures briefly growth is produced. occursaroundan equilibrium, before,at a growthrateof4 (e), a truechaoticpattern perioddoublingbifurcation

102

Earth'ssurface.Orderproducedbyverycomplex systemscan be foundat a rangeofscales:in the coastlineof westernBritain(Mandelbrot, 1983), landscape evolution (Phillips, 1995; 1998), drainagebasins(Rigonet al, 1994),beach cusps (Wernerand Fink, 1994) and sand bedforms (Hallet, 1990). These featuresall show the emergence of qualitativeand/or quantitative orderthroughtheactionofnon-linear processes or as a reactionto the couplingof processes of an withina systemwithoutthe intervention external control. One of the moststriking examplesof selfin is exhibitedby organisation geomorphology the Algodones Dunes of southernCalifornia (Andersen,1990; and see Figure4). These sand

of dunesinvolvethreesuperimposed generations withwavelengths at different ordersof bedforms, magnitude from kilometres to tens of centimetres. The processesinvolvedin aeolian transportand depositioninvolvethe complex the interaction betweenturbulent fluiddynamics, sporadic injection,collision and ejection of suspended and saltating sand-grains,and tractive andavalancheeventsatthe unpredictable fluid-sediment interface. itwasbelieved Originally, thatthephasechangesweretheresultofdiscrete reductionsin wind-speed.However,it has been establishedthatall threebedformwavelengths with no influenceof formedsimultaneously, externalflux.

This content downloaded from 168.176.55.24 on Thu, 18 Jun 2015 15:57:36 UTC All use subject to JSTOR Terms and Conditions

whichare retainedwhileothersare eliminated. The onlycompetitioninvolvedis one between subsequent states of the same system.Such selectioncan stillbe 'natural'.The stability ofthe as a structure, evolvingecological functioning selection criterion,is purely internalto the no outsideforcesor pressuresare configuration: necessaryto explainthem.In cases likethese,the selectionis inherentin the configuration itself, and an asymmetric transitionfromvaryingto stablemaybe calledself-organisation. Therefore, see 'natural ecologistsinterestedin complexity selection' as involving external, Darwinian selection. selection,and internal, self-organising

Self-organised criticality SOC and natural hazard prediction

GEOGRAPHY IN COMPLEXITY PHYSICAL GEOGRAPHY

Geography© 2002

Earthand environmental scientists havebecome involvedindevisingmeansto reduce increasingly and mitigatethe occurrenceof a varietyof naturalhazards,suchas earthquakes, devastating volcaniceruptions, floods,and landslides.Within 50 years,more than one-thirdof the world's exhibit and volcanically Figure3: TheAlgodonesDunesofsouthernCalifornia populationwilllivein seismically ofbedforms thathaveformed threegenerations active zones. The InternationalCouncil of underthesameenvironmental conditions, spontaneously withUNESCOand the Scientific Unions,together ofaeoliandeposition.©JulianWorker despitethecomplexity endorsedthe 1990sas the WorldBank,therefore International Decade of Natural Disaster Reduction(IDNDR). They planneda varietyof Ecology to addressproblemsrelatedto the within programmes In additionto wideruse of complexity and mitigation of these disasters, predictability Falkner and Lowes, 1999; ecology(i.e. Bellamy in wherethe Third World countries particularly havebeen and are and Falkner, 2000),criticisms most of the are acute. Many computerbeing raised against the Darwinianview of problems evolution. Darwinian theories of biological modelling techniques used to predict the evolution involve the process of 'natural occurrenceof these physicalprocessesinvolve andcomplexity, i.e. dynamics selection'. Darwin believed that ecological theuse ofnon-linear It has or fractal models. cellular automata growth communities developedwhichfittedthe natural thataremostsuited long been recognisedthatthe magnitudeand withorganisms environment, ofmanyformsofdisasterfollowpower and evolvingat to theenvironment frequency proliferating laws(Turcotte, the expense of the less Tit'. The selectionis 1993).Thusthereis a spectrumof event sizes, from small to large, that are naturalin the sense thatthereis no actor or of relatedbythenon-linear dynamics presumably purposivesystemmakingthe selection.It is be models the Therefore A involved. or without may computer process. plan design spontaneous, arenowdevelopingthebelief used to studythe fundamental numberofworkers physicsof the to developmeans thatnaturalselectionmustbe complemented by process,and mostimportantly, in order to explainevolution to predictthe patternsof occurrenceof large self-organisation precursory (Kauffman,1994; Swenson, 1992, 2000). eventsin the modelsand to identify Darwinismsees evolution as the result of eventsthatoftenheraldthe onsetof potentially hazardousphenomena. selection by the environmentacting on a Malamudetal. (1998) developeda computer forresources. oforganisms competing population model thatuses theconceptof SOC and powerof theories of The incorporation self-organisation law relationshipsto address contemporary thatcan be selected or allows a configuration of the presence of geographicalproblems.Theirmodel chartsthe eliminatedindependently 103 a singlesystemcan pass otherconfigurations: frequencyof small and mediumforestfiresin in and the to assist scientists of order some of prediction througha sequence configurations,

This content downloaded from 168.176.55.24 on Thu, 18 Jun 2015 15:57:36 UTC All use subject to JSTOR Terms and Conditions

preventionof large forestfireslike those that

was notto 1877; 1914),however,theframework fullydevelop untilthe 1960s. The charismatic forestfirescomplywithpower-law IN COMPLEXITY relationships, figureof WM. Davis dominatedgeomorphology in the 1950sand revolution should permitthe small untilthequantitative PHYSICAL management strategies fires that occur as of the evolution 1960s 1952;1957;Schumm,1963).The GEOGRAPHY (Strahler, naturally part and regeneration of foreststo take place. Fire process paradigm,still developingduringthe advance of computingtechniques,became a resourcemanagerscan use thisinformation to inputthenumberoffiresthatoccurovera period dominantframeworkin the late 1960s and the 1970s (Yatsu,1966; Carsonand oftimeandcalculatehowmanyfirestheyneed to throughout setinorderto preventlargerones.In addition,an 1972; Kirkby, Young, 1972). Althoughother frameworks were of the law allows better being employedat thistime, Geography© 2002 understanding power of the likelihoodof bigger ones. notablytheFrenchSchoolofClimaticGeography forecasting Management practicesshouldallowthe burning (Tricart and Cailleux, 1972), the process dominateduntiltheholisticprinciples of old growthvegetation, and encouragesmall framework of the systems paradigm were embraced firesabove the naturalnumber,in order to all aspectsof physicalgeographyin throughout preventcatastrophe. the late 1970s and early1980s (Chorley,1965; Similarcomputermodels can be used to ofearthquakes. Chorleyand Kennedy, 1971;Strahler, 1980). explainthetemporaldistribution Otherbranchesof physicalgeographyhave Providedwithan adequate knowledgeof local geologicalstructureand historicalrecordsthe been throughsimilarphases of development. scientist can easilypredictthespatialdistribution Whetherthe futuredevelopmentof the subject and non-linear of earthquakes.Traditionally, have willinvolvecomplexity dynamics seismologists remains to In be seen. reliedon precursorphenomena,such as minor addition,whethersuch a a paradigmshiftis also tremors,animal behaviour and groundwater changewouldconstitute for debate. As involvestheholistic to the of up complexity changes, predict timing large of it be seen as an extensionof study systems, These may have met with earthquakes. techniques framework. as scientists littlesuccess. However,the Gutenberg-Richterthesystems Alternatively, have discovered that law shows thatthe distribution manyphenomenacannotbe of earthquake described deterministic itmaybe using is a principles, and each magnitudes powerlaw, earthquake a true zone organisesitselfintoa criticalstate- much arguedthatsuch a changewould signify revolution. likethe sand grainsin Bak's sandpileavalanche scientific The complexity paradigmenteredthepublic model. In addition,the fracturesand faults in the form of domain, popularsciencebooksand associated with earthquake centres are before the lexicon describing the articles, characterised fractal dimensions. Whilethisis by associated concepts and processes could be of no use in prediction,the phenomenonhas defined.Arguably, thesamecanbe said accurately been employedwithgreatsuccessin mineralore of 'Gaia which sits hypothesis', verycomfortably andJohnston, exploration (McCaffrey 1996). withinthecomplexity framework. The notionthat theearthregulatesconditionsto an optimumto support life is possiblythe most convincing as example of self-organisationin physical Complexity Itisinteresting tonotethattheancient geography. a paradigmfor Greeksappear to have been firstin recognising this link,as the earth goddess Gaia was the physicalgeography ofprimordial ChaosinGreekmythology. daughter As Mansonnotes,it is important thatacademics Figure 5 illustrateshow the developmentof explore the 'ontologicaland epistomological can be viewedas a model that corollariesof complexity' geomorphology (2001, p. 412). The thathas oftenbeen used complies with Kuhn's theories of paradigm breadthofterminology development.Beforethe adventof Darwinism, to describeidenticalaspectsof complexsystems andtheextension ofhisconceptsbyDavis(1899), hascausedmanyproblemsto thedevelopment of Penck(1924) and King(1957), the studyof the theframework withinacademicstudy. Earthwasessentially As has been discussed, Manson (2001) ideological(Ritter, 1822-59). One of Throughthe latterpartsof the nineteenthand recognisesthreeaspects of complexity. 104 into the twentiethcenturywork began on theseaspects,'algorithmic complexity', highlights causeandeffect identifying relationships (Gilbert, another problem. The term describes those GEOGRAPHY devastated Yellowstone NationalParkin 1988.As

This content downloaded from 168.176.55.24 on Thu, 18 Jun 2015 15:57:36 UTC All use subject to JSTOR Terms and Conditions

Complexity

Quantitative Revolution

^

-

i

o^^"

i

1900

/ -/ 4?

A / /

^r

1850

-

Time

►

1950

systemsthat are so intricatethat they are practically impossibleto study.Such a literal translation ofcomplexity cannotformpartofthe studyofcomplexsystemsas suchproblemsarise froman insufficient ofthesystem understanding being studied or inadequate cerebral or computationalpower to model and describe them.Therefore,'algorithmic complexity'is a movementawayfromthe potentially misleading academicdefinition of complexity and its strict, associatedissues. A majorproblemencounteredby scientists in non-linear thatare interested is that dynamics theirworkoftenmeetswithhardenedopposition fromacademicswhoseviewsaredirectly opposed to the new approach.This problemis not only associatedwithqualitativestudies,but is often Newtonian thoughtto be againstdeterministic, principlesemployed withingeneral scientific research. Reasonsforthe lack of developmentand as a viableframework acceptanceof complexity mayalso be associatedwiththe habitualwayin whichscientistsperceiveand studythe world. Barriersto the acceptanceof complexitymay differ,according to whether we discuss deterministic chaosor self-organisation. aversionto chaoticphenomenais, Scientists' more perhaps, easily understoodthan their oppositionto SO or SOC. Since time began, sciencehas been trying to describethe physical worldin termsof formulasand theoriesthat and generalise.However,by reduce, simplify, reductionist thetheoriesare principles employing Itmaybe thecase that,untilthe oftenincomplete. the tools thatscientists revolution, quantitative

2000

'

GEOGRAPHY IN COMPLEXITY PHYSICAL GEOGRAPHY

Vnoie Phase

of Figure5: Aninterpretation paradigm development within geomorphology. is,at Geomorphology dominated present, bythe framework. Whether systems complexity maybe seenas an extension ofthisframework, oras a newparadigm which a revolution mayresultfrom a crisisphase, following remains tobe seen.

used to measure,recordand analysethe vast amountofdatapresentin thenaturalworldwere this verylimited.Withthe adventof computers, reductionism has finally startedto giveway,and more scientistshave begun to studycomplex systems. Scientists are alwaysmorecomfortable with thingsthattheycan explain.The methodsand theoriesthatworkare exercisedmuchmoreand thereforebecome more successfulmodes of enquiry. When studying complex systems, motivationis more likelyto decrease and the scientist'sattentionmoves elsewhere. As a scientist gains more knowledge about a phenomenon,the phenomenonbecomes more However,withcomplexsystemsthe predictable. more knowledgethat is acquired the more thephenomenonbecomes- little unpredictable wonderthenthatthe scientistprefersto study simple systems. Therefore the scientific is farmorelikelyto studywhatever community reduces its short-term discomfort.These are invariablysimplistic,linear and predictable processes, not the complex systems that incorporatelarge amountsof data that easily exceed the limitsof the mind.How can thisbe sensiblewhen the reductionist principlesthat sciencehas preferred cannotalwaysdescribethe waythattheworldworks? As we have stated,scientists and people in withtrulyrandom generalare rarelycomfortable and phenomenaand,inherently, prefer symmetry order.This,in turn,is likelyto haveaffected the of development,widerappeal and applicability and SOC. Evenwhenthepattern self-organisation canonlybe examinedanddescribedbycomputer

This content downloaded from 168.176.55.24 on Thu, 18 Jun 2015 15:57:36 UTC All use subject to JSTOR Terms and Conditions

Geography ©2002

105

techniques Bellamy,J.A.and Lowes, D. (1999) 'Modellingchange in modellingand/orcomplexstatistical (as is thecase withmanychaoticorscale-invariant stateof complexecologicalsystemsin space and time: an application to sustainable grazing management', IN phenomenaor those thatmaybe describedby COMPLEXITY EnvironmentInternational,25, 6-7,pp. 701-12. with to be comfortable tend PHYSICAL powerlaws) people Carson,M.A. and Kirkby, MJ. (1972) Hillslope Form and ofJacksonPollockare a GEOGRAPHY them.The drippaintings Press. Process.Cambridge:CambridgeUniversity perfectexampleofthis.Whilenotto thetasteof Chorley,RJ. (1962) Geomorphologyand General Systems all artlovers,manypeople enjoyand admirehis Theory.US Geological SurveyProfessionalArticleNo. workfroma purelyqualitative 500-B. pointofview.The arewidelyacceptedto factthatPollock'spaintings Chorley,RJ.and Kennedy,B.A.(1971) PhysicalGeography: A systemsapproach. London: Prentice-Hall. be masterpieces and notsimplyrandomexercises P and Highfìeld, R. (1995) Frontiers of Coveney, inself-expression byan unstablealcoholicmaylie Geography©2002 The searchfor order in a chaotic world. Complexity: in thefactthatthepaintings are characterised by London:Faber. dimensions(Taylor ei ai, 1999).Evenmore Davis,WM. (1899) 'The geographicalcycle',Geographical fractal the fractal dimension steadily mysteriously, Journal, 14,pp. 481-504. his Espejo, R. (2000) 'Self-construction increasedforpaintings of desirable social producedthroughout life. systems',Kybemetes,29, 7-8,pp. 949-63. working Otherinsights maybe gainedwithreference Falkner,G. and Falkner,R. (2000) 'Objectivisticviews in biology: an obstacle to our understandingof selfto the literatureon the psychologyand organisation processes in aquatic ecosystems', of Real-life philosophy perception. experience FreshwaterBiology, 44, 3, pp. 553-9. suggests a great communalityamong the Gibson, JJ. (1979) An Ecological Approach to Visual ofdifferent those perceptions people,particularly Perception.Boston:Houghton Mifflin. sharing a common landscape. In addition, Gilbert,G.K. (1914) The Transportationof Debris by becausesensoryinputsareoftennotrichenough Running Water. US Geological Survey Professional ArticleNo. 86. to mediateperception, thevieweradds to them, frommemoryor habit,to completegaps. When Gilbert,G.K. (1887) Reporton the Geologyof theHenry Mountains. US Geographicaland GeologicalSurveyof thephenomenabeingviewedare scale invariant, the RockyMountainRegion. theprocessbecomeseasier,and therefore more Gordon,I.E. (1997) Theoriesof VisualPerception(second comfortable(Gordon, 1997). The necessary edition).Chichester:Wiley. mathematics needed to understandand accept Hall, N. (1991) The New Scientist Guide to Chaos. in the landscapeis an implicitpartof invariance Harmondsworth: PeneuinBooks. our visualperception(Gibson,1979). It maybe Hallet, B. (1990) 'Spatial self-organisation in thecase then,thatthestudyof self-organisation geomorphology:fromperiodicbedformsand patterned ground to scale-invariant topography',Earth Science has been hinderedby a predisposition not to Reviews.29. dd. S7-7S. or examine and question qualitative quantitative S. (1994) AtHome in the Universe:Thesearch Kauffman, becausetheyarepleasingto us emergent patterns . London:Viking. for laws ofcomplexity and the inferentialprocesses involved in natureof hillslopes', King,L.C. (1957) 'The uniformitarian areoftenunconscious. identifying similarity Transactionsof theEdinburghGeological Society,17, GEOGRAPHY

Acknowledgements

The authorwould liketo acknowledgethetwoanonymous refereesof thisarticlefortheirconstructive commentson an earlierdraft.

References

106

Anderson,R.S. (1990) Aeolian ripplesas examplesof selforganization in geomorphological systems', EarthScience Reviews,29, 1-4,pp. 77-96. Bak, R and Chen, K. (1991) 'Self-organisedcriticality', American,264, 1, pp. 46-53. Scientific Bak, P, Tang,C. and Wiesenfeld,K. (1987) 'Self-organized criticality:an explanation of low frequencynoise', PhysicalReviewLetters,59, pp. 381-4. Bak, P, Tang,C. and Wiesenfeld,K. (1988) 'Self-organized criticality', PhysicsReview,38, 1, pp. 364-74.

pp. 81-104. Lewin,R. (1994) Complexity:Lifeon the edge of chaos. London: Phoenixarticlebacks. Lorenz, E.N. (1963) 'Deterministicnon-periodicflow', Sciences,20, pp. 130-41. JournaloftheAtmospheric D.L. (1998) 'Forest Malamud,B.D., Morein,G. and Turcotte, fires:an example of self-organizedcriticalbehavior', Science,281, 5384,pp. 1840-42. Mandelbrot,B. (1983) The Fractal Geometryof Nature. NewYork:WH. Freeman. Manson,S.M. (2001) 'Simplifying complexity:a reviewof complexity theory',Geoforum,32, pp. 405-14. May,R.M. (1976) 'Simple mathematicalmodels withvery complicateddynamics',Nature,261, pp. 159-67. KJ.Wand JohnstonJ.D. (1996) 'Fractalanalysis McCaffrey ofa mineralisedveindeposit:Curraghinalt gold deposit, MineraliumDeposita, 31, 1-2,pp. 52-8. CountyTyrone', Nicolis,G.and Prigogine,I. (1989) ExploringComplexit. NewYork:WH. Freeman. Penck,W (1924) Die Morphologische Analyse:Ein Kapital der PhysikalischenGeologie.Stuttgart: Engelhorn.

This content downloaded from 168.176.55.24 on Thu, 18 Jun 2015 15:57:36 UTC All use subject to JSTOR Terms and Conditions

Phillips,J.D. (1998) Earth Surface Systems:Complexity, orderand scale. Oxford:Blackwell. Phillips,J.D. (1995) 'Self organisationand landscape evolution',Progressin PhysicalGeography,19, 3, pp. 309-21. I. (1994) 'On Rigon,R., Rinaldo,A. and Rodriguez-Irturbe, Journal of Geophysical landscape self-organisation', Research,B99, pp. 11971-93. zur Natur C. (1822-59)Die Erkunde,im Verhältnis Ritter, und zur Geschichtedes Menschen,oder Allgemeine VergleichendeGeographieals sichere Grundlage des Studiums and Unterrichtsin Physikalischen und HistorischenWissenChañen. Berlin:Reimer. Schumm,SA (1963) TheDisparitybetweenPresentRates of Denudation and Orogeny.US Geological Survey ArticleNo. 454-H. Professional wavesand chaos inchemical Scott,S.K. (1994) 'Oscillations, Primers,18,p. 92. kinetics',OxfordChemistry S. and Bak, Sole, R.Y,ManrubiaS.C., Benton,M.,Kauffman, and scalinginevolutionary P (1999) 'Criticality ecology', Trendsin Ecology& Evolution,14,4, pp. 156-60. Strahler,A.N. (1952) 'Dynamicbasis of geomorphology', Bulletinof the Geological SocietyofAmerica,63, pp. 923-38. Strahler,A.N. (1957) 'Quantitativeanalysisof watershed geomorphology', Transactions of the American GeophysicalUnion,38, pp. 913-20. Strahler, A.N. (1980) 'Systems theory in physical geography',PhysicalGeography,1, pp. 1-27. Swenson,R. (1992) Autocatakinetics, yes- autopoiesis,no - steps toward a unified theory of evolutionary ordering',InternationalJournal of General Science, 21, 2, pp. 207-28.

Swenson, R. (2000) 'Spontaneous order,autocatakinetic closure,and the developmentof space-time',Closure: EmergentOrganizationsand theirDynamics,901, pp. 311-19. TaylorR.R,MicolichA.RandJonasD. (1999) 'Fractalanalysis of Pollock'sdrippaintings', Nature,399, 6735, p. 422. Trican,J.and Cailleux,A. (1972) Introductionto Climatic London: Longman. GeomorpholoBV. Turcotte,D.L. (1993) Fractals and Chaos in Geology. London: Longman. van Raan, A.F.J.(2000) 'On growth,ageing, and fractal of science', Scientometrics,47, 2, pp. differentiation 347-362. Theemergingscience at Waldrop,M.M.(1992) Complexity: theedge oforderand chaos. London:Viking. Werner,B.T and Fink,T.M. (1994) 'Beach cusps as selforganisedpatterns',Science. 260, pp. 968-71. Yatsu, E. (1966) Rock Control and Geomorphology. Tokyo:Sozisha. Young,A. (1972) Slopes. Edinburgh:Oliver& Boyd.

GEOGRAPHY IN COMPLEXITY PHYSICAL GEOGRAPHY

Geography0 2002

Dr AndrewRichardslecturesat The School of Earth Science and Geography, Kingston University, Penrhyn Road, Kingston-upon-Thames,Surrey KT1 2EE (E-mail:[email protected]).

107

This content downloaded from 168.176.55.24 on Thu, 18 Jun 2015 15:57:36 UTC All use subject to JSTOR Terms and Conditions