96 SCIENTIFIC AMERICAN March 1995

The stark Greek landscape charmseveryone who sees it. For thou-sands of years, Greeks and visi-

tors alike have sung the praises of thissmall country, famous for its BronzeAge civilizations and the culturalachievements of its people in classicaltimes. But is the countryside one seestoday the result of climate acting alone,or have humans played a part by clear-ing forests and causing soil erosion?

Recent archaeological work is chang-ing a long-standing view of the impactof agriculture on the land in Greece.The evidence mounts for episodes of de-forestation and catastrophic soil ero-sion over the past 8,000 years. Manyscholars believe they resulted from along history of human land use andabuse. This new perspective on humansettlement and its impact on the natu-ral environment stands in sharp con-trast to the views of the 19th-centuryRomantics, who saw the ancient Greeksas careful stewards of a land they heldto be Þlled with gods.

Indeed, it remains a widely espousedopinion today that the destructive eco-logical practices of modern civilizationare a new development. The popularpress frequently carries reports of peo-ple who advocate returning to the bal-

anced and reverential regard they sup-pose our ancestors had for the naturalworld. The Garden of Eden is a primalmyth of Western civilization, and it waspreceded in classical antiquity by thebelief in the Golden AgeÑa time, alasnow lost, when human beings were saidto have lived in innocent harmony withtheir natural environment.

Sir Peter B. Medawar (in The Limits of

Science, a book of his essays publishedin 1984) described this kind of thinkingas ÒArcadian.Ó He compared the con-cepts of Utopia and Arcadia and con-cluded that Arcadia is closer to the ide-al of a Golden Age than are the Utopiasof Thomas More and Francis Bacon.Their Utopias were places where sci-ence-based technology was employedfor the melioration of society, whereasArcadia Òis the conception farthest re-moved from Utopia, for one of its prin-cipal virtues is to be pastoral, prescien-tiÞc and pretechnological. In Arcadia,mankind lives in happiness, ignoranceand innocence, free from the diseasesand psychic disquiet that civilizationbrings with itÑliving indeed in thatstate of inner spiritual tranquillity whichcomes today only from having a sub-stantial private income derived fromtrustee securities.Ó

The eÝects of Arcadian thinking canbe seen in the debate about the degreeto which the ancient Greeks were re-sponsible for the deforestation and ero-sion that have reduced much of Greeceto a barren, stonyÑif picturesqueÑwasteland. To be sure, environmentaldegradation (if not its causes) was not-ed in ancient times; many references toit appear in the writings of the ancients,particularly Plato and Aristotle. Theygive accurate and apparently eyewit-ness accounts of deforestation and soilerosion in the fourth century B.C. Ar-chaeology can now conÞrm that de-spoilment of the natural environment

took place in antiquity, but the evidenceallows one also to place the responsibil-ity for this destruction primarily on theinhabitants of that time.

A New Kind of Archaeology

Over the past two decades, the in-troduction of multidisciplinary and

intensive regional survey techniques inGreece has revolutionized Greek archae-ology. One aspect of this revolution is ashift away from the investigation of sin-gle sites to studies of the natural andcultural history of entire regions. Thisapproach is achieved in part by usingnew methods of Þeldwork, particularlysearching large areas (typically morethan 100 square kilometers) with teamsof people who record every artifact, fea-ture and site that can be detected. An-other innovation is the large-scale use ofremote sensing, ranging from ground-penetrating radar to satellite images, toassist the archaeologist in detectingpast human activity. Further aid hascome from the new discipline of geoar-chaeology, which combines the tech-niques of geology and archaeology toimprove the interpretation of the natu-ral contexts of ancient cultures.

An example of the collaboration be-tween archaeologists and geoarchaeol-ogists can be seen in the study of land-scape change in southern Greece. Twomajor projects in which I have partici-pated since 1979 in the Argolid (thenortheastern part of the Peloponnesepeninsula) combined walking tours de-signed to identify archaeological siteswith geologic surveys that attempted toreconstruct the history of the landscapeduring the Pleistocene and Holoceneepochs of the past 50,000 years.

One of these projects targeted theArgive plain. According to Aristotle, ithad undergone considerable alterationin the Bronze Age, from about 3000 to

CURTIS N. RUNNELS earned degreesin archaeology from the University ofKansas and Indiana University and hastaught at Stanford University. He is asso-ciate professor of archaeology at BostonUniversity. Runnels has worked for 20years in Greece, specializing in the studyof early prehistory and the relation be-tween human settlement and landscapethrough time. His most recent book is AGreek Countryside: The Southern Argolidfrom Prehistory to the Present Day, writ-ten with Michael H. Jameson and TjeerdH. van Andel.

Environmental Degradationin Ancient Greece

Contrary to the view that the ancients lived in harmonywith their environment, archaeological and geologic

evidence shows that they often abused the land

by Curtis N. Runnels

Copyright 1995 Scientific American, Inc.

1000 B.C. To investigate ancient landuse, I joined archaeologist Berit Wells,director of the Swedish Institute in Ath-ens, and Eberhard A. W. Zangger, a geo-archaeologist now at the University ofHeidelberg. Our goal was to survey twovalleys, Berbati and Limnes, which lie onthe northern edge of the Argive plainand have a combined area of 60 squarekilometers.

This project, sponsored by the Swed-ish Institute and headed by Wells, hadtwo parts. One was a geoarchaeologicalstudy of the two valleys by Zangger,who had done a similar study of theArgive plain while working on his doc-toral dissertation at Stanford Universi-ty. The other was an intensive archaeo-logical survey.

The survey brought to light evidenceof human activity from the Middle Pa-leolithic period, some 50,000 years ago,and showed that agriculturists enteredthe area about 7,000 years ago, settling

on the edges of the Berbati Valley nearthe best soils and supplies of water. Af-ter this initial colonization, the size ofthe original Neolithic settlements beganto grow, and new ones were founded.By the Late Neolithic period (between4000 and 3000 B.C.), more than 20 set-tlements and smaller sites were scat-tered over the region. This expansioncontinued for the Þrst few centuries ofthe third millenniumÑthe Early BronzeAge in cultural terms.

Ancient Land Use

From the larger settlements came ev-idence of agriculture and many arti-

facts such as pottery, querns for grind-ing grain and simple stone tools of ßintand obsidian. Some of the smaller sitesdoubtless served special purposes (ashuts for shepherds and storehouses fortools, for example) and were not neces-sarily inhabited permanently. Neverthe-

less, they demonstrate the more inten-sive use of the landscape at this time be-cause they are located on steep slopes,at high elevations and in areas whereonly soils of marginal productivity everexisted.

In addition to sites where past hu-man activity left concentrations of arti-facts, our searching technique revealedtens of thousands of individual Neo-lithic items scattered over the country-side. These thin scatters of artifacts arefound in areas now uninhabited and insome cases uninhabitable because theyare devoid of soil and vegetation.

The pattern of settlements and arti-facts reveals the structure of what ar-chaeologists call fossil cultural land-scapes. Particularly interesting are thenumerous Þnds of ground and polishedhard-stone axes and ßint blades, whichhave edges coated with silica depositedwhen the blades were used to cut grass-es and other plants. These common agri-

SCIENTIFIC AMERICAN March 1995 97

GOLDEN AGE of Greece is represented in Claude LorrainÕspainting The Judgment of Paris (1646), in which the goddessof strife throws down a golden apple inscribed Òto the fair-estÓ and says that Paris must award it to one of three god-

desses. The persistent view of the ancient Greeks as carefulstewards of a land that they saw as guided by gods and god-desses is belied by recent archaeological evidence of soil ero-sion caused by human land abuse.

CO

UR

TE

SY

OF

TH

E N

AT

ION

AL

GA

LLE

RY

OF

AR

T, A

ILS

A M

ELL

ON

BR

UC

E F

UN

D

Copyright 1995 Scientific American, Inc.

cultural tools belong to the Neolithicperiod and the Early Bronze Age andmark regions where agricultural Þeldsonce existed. We believe grazing andfarming were the principal activities onthe slopes of the Berbati and Limnesvalleys from 4000 to 3000 B.C. and thatthey resulted in one or more episodesof catastrophic soil erosion, which leftalluvial fans in the valley bottoms anddeep layers of sediments to mantle theArgive plain.

One episode occurred at the end ofthe Neolithic period, well after the ex-pansion of settlement in the Berbati andLimnes valleys. This erosion buried atleast one Neolithic site in the Argiveplain, which was found in one of Zang-gerÕs cores that was drilled through thethick alluvium. What appears to havebeen a second similar event came in thethird millennium, at the end of the Ear-ly Bronze Age. The erosional depositsof this period spread across the Argiveplain to make up most of its present

surface. Nor were these the last eventsto be recorded. Flooding eclipsed partsof the town of Tiryns at the end of theLate Bronze Age (around 1200 B.C. ) .After this natural disaster, the slopesof Berbati were abandoned and for themost part never inhabited again.

ConÞrmation of the picture oÝeredby the archaeological record comesfrom pollen taken from a seven-metercore obtained from the now vanishedLake Lerna in the Argive plain and ana-lyzed by Susanne Jahns of the Universi-ty of G�ttingen. It shows that deciduousoak trees were common in the Argolidin the Holocene epoch. By the middleof the fourth millennium B.C., however,the oak pollen had dropped oÝ sharp-ly and was replaced by pollen of horn-beam, pine, scrub oak and heather.These plants are endemic to clearedand disturbed land. Today the slopessurrounding the Argive plain consist ofbare, rocky land covered with a thinscrub vegetation of low bushes and

herbaceous plants, with only isolatedpockets of treesÑchießy pines.

Other projects in Greece have foundsigns of soil erosion. The episodes donot correlate in time with the periods oferosion in the Argolid. If climate werethe sole cause, most of Greece wouldhave simultaneously been aÝected. Thevariation in timing therefore points toa human role. Clearing natural vegeta-tion from the slopes, which was neces-sary to bring them under cultivationand to open them for grazing by sheepand goats, made the soil unstable andthereby triggered extensive and perma-nent erosion.

The Soil Tells a Story

Our second project produced furtherevidence of regional land abuse.

This study was carried out in the south-ern Argolid, a remote part of the Pelo-ponnese peninsula. With historian Mi-chael H. Jameson of Stanford and geo-

98 SCIENTIFIC AMERICAN March 1995

LAND ABUSE by ancient Greeks followed a consistent pattern(left ). At first (a ), a mixed forest held the soil in place on aslope. When the forest was cleared for farming (b ), the soilstayed in place for a time, but eventually erosion carried itoÝ the slope, depositing alluvium in the valley bottom (c). As a

result, a typical geologic proÞle (right ) from the southern Ar-golid region of Greece shows a sequence of erosional depos-its and intervening soils (with their approximate ages). In eachperiod the soil that gradually forms above the deposits is thin-ner and less developed and is less able to support vegetation.

JAN

A B

RE

NN

ING

17TH TO 20THCENTURY

MIDDLE BYZANTINE

300–50 B.C.

2500–2000 B.C.

45,000–32,000 B.P.

60,000 B.P.

250,000YEARS

BEFOREPRESENT

a

b

c

Copyright 1995 Scientific American, Inc.

archaeologist Tjeerd H. van Andel ofthe University of Cambridge, I conduct-ed an archaeological reconnaissance ofsome 250 square kilometers. The surveyidentiÞed more than 350 sites of hu-man activity spanning a period of some50,000 years.

An interesting pattern of settlementemerged when we plotted the numbersof sites by archaeological period. Itturned out that the sites are not dis-tributed evenly through time but in-stead exhibit a series of abrupt peaksand valleys, suggesting that periods ofexpansion of sites over the land werefollowed by periods of abandonment.Somewhat to our surprise, we foundthat this unexpected pattern of settle-ment could be correlated with the geo-logic history of the region.

This correlation resulted from an anal-ysis of the erosional history of the areamade by van Andel and Kevin O. Pope,also at Stanford. An innovative featureof their work was the study of soils.Soils can be distinguished from the de-posits of erosion because they form onthe deposits as a result of the chemicalweathering and transformation of un-derlying sedimentsÑa process that cantake thousands of years in the semiaridparts of Greece. Van Andel and PopeidentiÞed at least seven cycles of soilerosion that were followed by times ofstability when thick proÞles of soil hadthe opportunity to accumulate. The age-related features of soil proÞles make itpossible to correlate the diÝerent ero-sional deposits and soil proÞles with ar-chaeological Þndings.

Three of the erosional events thatleft marks in the geologic record tookplace during the last ice ageÑat about272,000, 52,000 and 33,000 years ago.They can be attributed to global climatechanges. But four episodes belong to thepast 5,000 years. Each episodeÑatabout 2500 B.C., 350 to 50 B.C., A.D. 950to 1450 and in recent timesÑwas fol-lowed by a period of stability when sub-stantial soil proÞles formed. Althoughsmall-scale climate changes may partlyexplain this pattern, we place the chiefblame on the activities of the local in-habitants. Our evidence for this claim istwofold: the correlation between the pe-riods of erosion and the periods of in-tense human settlement, and the for-mation of soil during the periods whenthe human impact was minimal.

Our dating technique began with asearch for archaeological sites and arti-facts that are covered by alluvium,buried in it or sitting on it. By combin-ing archaeological and geologic data,we were able to date the times of ero-sion closely. An important element inthis work was the recognition that lay-

ers of soils could be used to determineperiods of surface stability that con-trasted with the erosional periods.

From these data we concluded thatthe clearing of land during times of in-tensive human settlement gave rise tosoil erosion, which in turn caused thepeople to abandon their settlements orat least to scale back their activities.And the reduced human activity per-mitted erosional deposits to stabilizeand soils to form.

Soil erosion on a similar scale has beenreported from other parts of GreeceÑthe northern provinces of Macedoniaand Thessaly and the islands of Euboeain the center of the country and ofCrete in the south. The episodes datefrom as early as the sixth millenniumB.C. and continue through virtually ev-ery historical era to the present day.

The archaeological study of ancientenvironmental catastrophes has onlyjust begun, and it is perhaps too earlyto say that such analyses can alwaysdistinguish the eÝects of human habita-tion from those of climate. Nevertheless,it is clear from the variety and differenttiming of these events in neighboringregions that a climatic explanation isunlikely to account for all the data. Wehave learned enough to maintain thatthe changes in the natural landscape in

antiquity are at least partly the resultof shortsighted human activity. The les-son to be drawn from our work is thatpeople have a long history of misusingthe land: environmental catastrophesare not an innovation of the modernworld.

SCIENTIFIC AMERICAN March 1995 99

“In the time of the Trojan wars the Argive land was marshy and could onlysupport a small population, whereas the land of Mycenae was in good condi-tion (and for this reason Mycenae was the superior). But now the opposite isthe case. . . the land of Mycenae has become.. .dry and barren, while the Ar-give land that was formerly barren owing to the water has now become fruit-ful. Now the same process that has taken place in this small district must besupposed to be going on over whole countries and on a large scale.”

— Aristotle, from Meteorologica, Book 1, Chapter 14

A typical landscape in the Argolid region of Greece today

FURTHER READING

FIVE THOUSAND YEARS OF LAND USEAND ABUSE IN THE SOUTHERN ARGOLID,GREECE. Tjeerd H. van Andel, Curtis N.Runnels and Kevin O. Pope in Hesperia,Vol. 55, No. 1, pages 103Ð128; Jan-uaryÐMarch 1986.

LAND USE AND SOIL EROSION IN PREHIS-TORIC AND HISTORICAL GREECE. TjeerdH. van Andel, Eberhard Zangger andAnne Demitrack in Journal of Field Ar-chaeology, Vol. 17, No. 4, pages 379Ð 396; Winter 1990.

THE BERBATI-LIMNES ARCHAEOLOGICALSURVEY: THE 1988 SEASON. Berit Wells,Curtis Runnels and Eberhard Zanggerin Opuscula Atheniensia, Vol. 18, No.15, pages 207Ð238; 1990.

NEOLITHIC TO PRESENT SOIL EROSION INGREECE. Eberhard Zangger in Past andPresent Soil Erosion. Edited by MartinBell and John Boardman. Oxbow Books,1992.

THE GEOARCHAEOLOGY OF THE ARGOLID.Eberhard Zangger. Mann Verlag, 1993.

Paradise Lost

EU

GE

NE

G. S

CH

ULZ

Copyright 1995 Scientific American, Inc.