Population Changes in Ottoman Anatolia during the 16th and 17th Centuries: The "Demographic Crisis" Reconsidered Author(s):

Oktay zel Source: International Journal of Middle East Studies, Vol. 36, No. 2 (May, 2004), pp. 183-205 Published by: Cambridge University Press Stable URL: http://www.jstor.org/stable/3880031 Accessed: 09/02/2010 12:00
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=cup. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. 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 support@jstor.org.

Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access to International Journal of Middle East Studies.

http://www.jstor.org

Int. J. Middle East Stud.36 (2004), 183-205. Printed in the UnitedStates of America DOI: 10.1017.S0020743804362021

OktayOzel CHANGES IN OTTOMAN POPULATION ANATOLIA DURING THE 16TH AND 17TH CRISIS" THE "DEMOGRAPHIC CENTURIES: RECONSIDERED

The historiographyof the past two decades of the demographichistory of 16th- and OttomanAnatoliahas seen diverseand often conflictingargumentsamong 17th-century historians.Whetherthe OttomanEmpire witnessed "populationpressure"in the 16th century,and whetherthis was followed in the 17th centuryby a serious "demographic have constitutedthe central crisis,"considered by some historiansas a "catastrophe," theme of the debate. The roots of these issues can be traced as far back as the early works of OmerLiitfiBarkanin the 1940s and 1950s.1It appearsthat the disagreements not only arose as a resultof the differentmodels of historicaldemographydevelopedby diverse schools of thought,but that they also owed much to the highly disputednature of the sources that providethe bulk of quantitativedata for the demographichistory of the OttomanEmpire.2 When looking at the sources, one immediately realizes that the central part of the debate falls into the realm of what is known as "defterology,"3 sub-field of Ottoman a covering works based on the series of Ottomantax registers,mainly of historiography the 15th and 16th centuries (tahrir defters). Barkan was the first historian to present these sources to the world of Ottomanists,in the 1940s.4 In his seminal article "Tarihi results of the he ve DemografiAragtirmalari OsmanliTarihi," presentedthe preliminary painstakingwork of his team in istanbul on a whole series of deftersof the 16thcentury. Also discussing some methodologicalaspects of Ottomandemographichistory and its sources, Barkanpointed in that article to the main trends of populationmovements in the OttomanEmpirein that century. However,Barkan'spioneeringworks on Ottomandemographichistory were not followed until the late 1960s,5 when some historiansturnedto the same sources for their works on local history.The new explosion in the use of tahrirregisterscame from the 1970s onward,soon leading to the developmentof a separatefield-defterology-with its sophisticatedmethods,distinctterminology,and, inevitably,growing debatesamong the specialists. Thus, Ottomanhistoricaldemographicstudieswere largely developed as partof local-historyresearchand focused primarilyon the periodbetween the mid-15th andlate 16th centuries.6 Duringthe past two decades, however,the researchand debates
Bilkent Bilkent in of Professor theDepartment History, Ozelis Assistant FEASS, 06800, University, Oktay
Ankara,Turkey;e-mail: oozel@bilkent.edu.tr. ? 2004 Cambridge University Press 0020-7438/04 $12.00

184 OktayOzel have expanded to include the 17th century,basing themselves almost exclusively on little attentionin demographic avartz and cizye registers,which until then had attracted studies.7 Barkan'sarticlesuggestedsubstantial growthin the populationof the OttomanEmpire in the 16th century,and subsequentcase studies of variousdistrictsof the empire have his The generallyconfirmed findings." tahrirregistersof theperiodclearlyshow doubling some cases even more) in the recorded tax-paying population,in urban and rural (in areas, during the century.9In his meticulous work published in 1972, Michael Cook developed the argumentthat, especially in the second half of the 16th centuryin some parts of ruralAnatolia, the populationgrew to the extent that it exceeded the amount of arable land availablefor cultivation.To him, this was an indication of "population This argumentconcurredin a sense with the view of MustafaAkdag, who pressure."'0 earlierhadreferred, years thoughimplicitly,to the populationgrowthof the same period, to him, resultedin the increasein the numberof peasantswithoutland which, according ((iftbozan levends). To Akdag, this was an importantfactor in the eventualbreakdown of the inner balance of the village economy and society, as well in the emergence of the ensuing Celali rebellions and widespreadterrorin the Ottomancountrysideat the turnof the 17thcentury.11 correlationthatAkdagestablishedbetween demographic, The socio-economic factors and political developmentswas later discussed-and to some degree, criticized-by Halil Inalcikand HuricihanIslamoglu-Inan.12 The maincriticism of Akdag's argumentfocused on the point that the early-17th-century phenomenonof the large-scaleabandonment villages could not be explainedsimply by economic and of demographicfactors. Akdag's critics drew attentioninstead to what are called "pull" that they thoughtthe cities would have offered to factors, such as variousopportunities as well as to the peasants' desire to enter the militaryclass, which would at peasants, least guaranteethem a steady income.13 At this point comes the importantquestion: what were the factors triggering the peasant masses to leave their villages at the end of the sixteenth century,becoming the main source of manpowerfor the great Celali rebellionsand the widespreadterror that was to devastatethe Anatolian countrysidethroughoutthe 17th century?'4As an explanation,scholarshave often referredto the increasingtax burdenandthe oppressive attitudesof local officials towardpeasants,both of which appearto have been a general Not phenomenonof this period.'" rejectingthe role of these factors,Akdagdevelopedthe argumentthatthe expandingruralpopulationcould no longerbe absorbedby the village economy, forcing many peasants to search for a living elsewhere. Inalclk, however, while accepting to a certainextent the role of demographicpressure,puts an emphasis on the desperateneed of the Ottomangovernmentfor moresoldiersusing firearms during the long and difficultyears of war at the end of the 16thcentury.Accordingto him, this need resultedin the formationof thesekbanandsarica troops,which would soon turninto Celali brigands.This coincided with the peasants'desireto enjoy the privilegedposition of the militaryclass of thatsame period,even thoughthe socio-economic position of the membersof the militarywas also deteriorating.'6 The final point of debate relates to 17th-centurydevelopments.A central theme is whetheror not one can speak of a "demographic crisis."The main discussion revolves aroundthe effects of the Celali rebellionsand focuses on whatis termed"depopulation," which is generally consideredto be closely linked to these rebellions. The debate over

Ottoman"DemographicCrisis" Reconsidered 185 the extent and natureof the radicaldecrease in the recordedtax-payingpopulationwas furtherdeveloped by Bruce McGowan to the point of a "demographiccatastrophe."17 in McGowan'smethod and his somewhatcontroversialfindings and interpretations his works on the Balkan lands, which were based nearly exclusively on the quantitative evidenceprovidedin avartzandcizyeregisters,were latercriticizedby MariaTodorova.'8 used the same While addressingonce morethe disputednatureof these sources,Todorova with different criteriaand centered her criticism on the misunderstanding and figures of misinterpretation the data offered by these registers;thus, she came to an opposite and no less controversialconclusion. She claimed that one could hardlyspeakeven of a considerabledecreasein Ottomanpopulationin the 17thcentury, alone a demographic let catastrophe. In the following article,I will re-evaluatethe main issues in this debatein the light of recentresearch,arguingthatall werepartof a complexhistoricalphenomenonthatcannot I be explainedby reductionist,single-factorapproachesand unfoundedinterpretations. will also emphasize that, althoughthere are many black holes in Ottomandemographic history,one can still reasonablyspeak of a general demographiccrisis duringthe late 16th and early 17th centuries.
THE 16TH CENTURY: FROM POPULATION "PRESSURE"

TO CELALi REBELLIONS

In her study on the dynamics of agricultural production,populationgrowth, and urban to Anatolia,islamoglu-inan,referring the case developmentin 16thcenturynorth-central of the Tokatand 4orum districts,arguesthatpopulationgrowthin the OttomanEmpire neverreachedthe point of "pressure" was describedby Michael Cook.19 that islamogluInan's view appearsto have found a certaindegree of support,becoming an argument often referredto by other Ottomanists.20 In elaboratingher argument,islamoglu-inan of suggests that the fragmentation reayaCiftliks,which is clearly revealedby the tahrir registers,did not necessarilymean thatthe peasantsbecame landless. She furtherargues thatthepeasantsin questionreactedto the worseningconditionsin termsof the imbalance betweenpopulationgrowthandthe insufficientamountof arablelandby (1) intensifying and cultivation; reclaimingunusedandforestedlandsto cultivation; (3) changingcrop (2) and patterns,or rationalizingagriculture, alteringconsumptionhabits.2'She then claims that the populationgrowth did not reach the extent of eventually forcing the peasants to leave their lands. The great increase in population in this respect is explained by the possibility of internalor westwardmigrationand the sedentarization pastoralist of The increase in the number of recorded caba (landless marriedmen) and nomads.22 miicerred(landless unmarriedmen) similarly is accountedfor by the possibility of an increaseddemandfor wage laborin the face of intense cultivation.23 As seen in this argument, islamoglu-Inansuggests, first, that the peasantmovements in Anatoliain the secondhalf of the 16thcenturywere of a migratorynature;and second, thatthe migrationto cities duringthis periodwas in fact the resultof the "preference" of of peasants,especially youngerones, who, underthe "drudgery work"in the Anatolian as countryside, chose to enter into the service of provincial administrators irregular soldiers or join medreses (theology schools) as students.24 The migrationof peasants thereforeshould not necessarily be seen as evidence of a subsistence crisis or of the

186 OktayOzel inabilityon the partof the village economy to absorban increasingpopulation.25 In other words, accordingto Islamoglu-Inan,we cannot speak here of demographicpressure.In saying this, however, she fails to note that the phenomenonof intensifying cultivation and shifting crop patterns,which was seen duringthe second half of the 16th centuryin many otherpartsof Anatolia,can also be linkedto economic anddemographic pressure, as well as to developing marketsand monetarychanges.26 However,the main argument of her work is not the analysis of certainhistoricalphenomenathat she had previously accepted. Insteadof dwelling on the subsistence crisis, the apparentdrop in per-capita productionvis-ai-visa considerablerise in prices, the fragmentationof peasant farms, andthe increasingnumberof landlesspeasants,27 focuses on how populationgrowth she affected the peasant economy and relationshipsin the Ottoman countryside.28 While the reasonsbehindthe migrationfromruralto urbanareasin Anatolia,she tries analyzing to minimize the extent of demographicfactorsbehindthis movement,thus rejectingthe thesis of populationpressure.In doing this, she seems to overemphasizethe possibilities mentionedearlierinsteadof attempting closer analysisof the evidence providedby the a sources she is using.29 The findings of recent studies of the neighboringnorth-central Anatoliandistrictsof Canik and Amasya, as well as Islamoglu-Inan'sown sources on the regions of Corum and Tokat, appearto supportthe argumentfor considerabledemographicpressure,as for Anatoliaduringthe second half of the suggested by Cook particularly north-central 16th century.30 In that region, for example, the fragmentation peasantfarmsreached of of land recordedin the name of certainpeaslevels, and the ever-shrinking high plots ant households (hane) began increasinglyto be cultivatedby more adult peasants or households.3'In addition,the numberof landless peasanthouseholds (caba [-bennak]) increased,for example in the Amasya districtto nearly40 percentof the total recorded adultmen, who constitute households;moreover,this figuredoes not include unmarried nearlyhalf of the recordedmale population.32 Anotherpoint furtherclarifies the picture. In her study, Islamoglu-Inanwrongly inman,"whereasit terpretsthe term "caba"in the tahrirregistersas "landlessunmarried As of man."33 a consequence,the proportion unmarried clearlyrefersto "landlessmarried men in the total adult male population-for example, in the region of Tokatbetween 1554 and 1576-appears to reach 70 percent,34 while in other parts of Anatolia in the same period it variesbetween 20 percentand 40 percent.35 This high percentage,which is difficult to explain, drops to about 45 percent when the term caba is taken in its correct meaning as clearly defined in the law codes (kanunname)of the province in This still significantrise in the numberof unmarried men is paralleledby a question.36 similarlevel of decreasein the numberof landless marriedmen in the very same district duringthe same period. In other words,the proportionof marriedmen in the total adult male populationin the Tokatcountrysidein 1574 shows a decreaseof nearly30 percent men comparedwith the situationtwenty years earlier,while the numberof unmarried increased even more in the same period.37 How can this be interpreted? One possible explanationcould be that,duringthis period,young adultmen found it increasinglyhard to get marriedunderthe worseningeconomic conditions,thus expandingthe unmarried adultmale population. The remarkable increase in the proportionof both landless and unmarried adultmen in the central lands of the province of Rum in Anatolia during the second half of the

Ottoman"DemographicCrisis" Reconsidered 187 16th century is also observablein the Amasya and Canik districts.38 According to the tahrir registers for these districts, the proportionof miicerredsto the total adult male populationin 1576 was 45.8 percent in Canik and 44.8 percent in Amasya. Similarly, the proportionof the landless marriedmen (caba) to the same total again in 1576 was 35 percentin Canikand31.7 percentin Amasya.In otherwords,the combinedproportion of unmarried landless marriedmen amongthe total adultmale populationat the turn and of the last quarterof the 16th century was around80 percent in the Canik region and around76 percent in Amasya.39Given the assumption that the proportionof young people (youngerthanfifteen years) amongthe populationas a whole was from one-third to one-half in pre-industrial societies,40these proportionsof unmarriedmen in northcentral Anatolia may be seen as not significantly abnormal.But when taken together with the numberof landless marriedman, this obviously points to a serious imbalance between the populationand the economy. This in turnalso lends supportto the notion, first suggested by Mustafa Akdag and later cautiously mentioned as a possibility by Cook along with Leyla Erderand SuraiyaFaroqhi,of serious difficulties in marriage in conditions(late marriageor non-marriage) the Anatoliancountryside.41 Having said this, one observes in some cases a different picture of the changing of Anatolia.In the western proportions differentsectors of ruralsociety in 16th-century Anatoliandistrictof Lazlkiyye (Denizli) between the 1520s and the 1570s, for example, we see an extraordinary increase (159.59%) in the numberof households holding the minimumamountof land(a bennak,or less thanhalf a farmstead), while the proportion of or those holding a full farmstead half a farmsteaddecreasedsignificantly(to 51.10% and this 30.05%,respectively).Interestingly, was accompaniedby a drasticfall in the number of unmarried adult men (75.77%).42 In this case, it seems that the observed population men increasinglyleft their growthfollowed a differentpath.While the young unmarried villages for brigandageor to fill the medreses as "students"(suhte) by mid-century43 (which meant that they went unrecordedin their villages), the increasing number of peasanthouseholdswho stayedin theirvillages foundless andless landto cultivate.Such fluctuationsin the compositionof the ruralpopulationof Anatolia in the second half of Behind all these the 16th centuryindicate a situationthat cannot be seen as "normal." developments,one clearly observes demographicpressure,although its consequences variedfrom region to region. There is furtherevidence that points to such pressure. Leaving aside the general populationgrowth that is evident particularlyfrom the second quarterof the century onward, one observes signs of dense settlement particularyin the lowlands and on high plateaus suitable for cultivation. Some plots of land hithertouninhabitedor unarableland for peasantsof used, the mezraas,were eitherreactivatedas supplementary nearbyvillages or were increasinglyturnedinto permanentsettlementsduringthe 16th One century.44 can add to this the increasingcases of lands newly opened to cultivation eitherfrom marginallands or throughthe clearanceof woodland.45 Parallelto this, there were instances of semi-nomadicTurkomangroups establishingpermanentsettlements (etrakiyevillages) in the mountainfringes, where they appearto have engaged in smallscale agricultureand animal husbandry.46 Despite the silence of the registers as to the cause of such cases, this clearly shows that arableland was expanding,probablyat the expense of pastureland, which was essential to the pastorallife and economy. It seems that,in the Amasyaregion, for example,the density of ruralsettlementobservablein the

188 OktayOzel 16th century was never to be reached again, even by the turn of the 20th century.47 In addition,the urbanpopulationof this period witnessed a considerableincrease. There are signs that big cities as regional centers, such as Tokat, received migrantsof rural origin, most of whom are likely to have been the landless and unmarried peasantsfrom the countryside mentioned earlier.It is highly probable that such cities continued to attractthese people throughoutthe second half of the 16th century,48 despite the efforts of the centralgovernmentto preventsuch populationmovements with strict rules and boundaries"of the social and regulationsdeveloped to maintainthe "pre-determined I economic order in both ruraland urbanareas.49 think all this points to the fact that Anatolia-at least, in the north-central parts-was underpressurefromrapidpopulation in the second half of the 16th century.It also indicates an apparentsubsistence growth crisis in the Anatoliancountryside.The demographicpressurethereforeappearsto be a historicalrealityin 16th-century Anatolia;it cannotsimply be ruledout as a hypothetical claim. It seems to have been a phenomenonthathad diverse effects throughoutsociety, includingon urbandwellers andnomads,at least in some partsof Anatoliain the second half of the 16th century.50 In this context, it is not unreasonable view these demographicchanges as a signifito cant factor in the spreadof the great Celali rebellions, and especially in the continuous terrorin the Ottomancountrysidethatbegan in the late 16thcenturyandescalatedin the early 17th century.It also seems more than a coincidence that the humansource of this general devastationwas largely generatedby the changing conditions in the Ottoman countrysidein the late 16thcentury.Populationpressurein this respectshould seriously be considered.This importantsubjectof discussion deserves a separatestudy.However, it should be pointed out here that the "pull"factors suggested by Islamoglu-Inanand offeredby cities to the villagers in difficulty,the urgent Inalcik,such as the opportunities need of the Ottomangovernmentfor more soldiers using firearms,and the employment of already rootless peasants to this end, no doubt possess a certaindegree of validity. It is evident that the government'scrucial decision to resortto this destabilizedhuman element as a short-termsolution to its militaryneeds led to the dangerousmobilization of this "floating mass" in the Anatolian countryside at the turn of the 17th century. However, at this point it is perhapsmore importantto emphasize the very presence of such a peasantmass in itself. Many of these peasants-landless, unmarried, living and at the limits of survivalwhile searchingfor a betterlife elsewhere-were open, despite of restrictions,to the attractiveness outside factors.5" Finally, it is also evident that this mass of peasants,the "surpluspopulation,"52 who had alreadybegun to leave their villages in large numbersmore visibly from the 1580s onward,were not only attracted such "pull"factors;they also resortedto "other" ways by A of life, includingillegal activities such as brigandage.53 cursorylook at the increasing records of such cases in miihimmeregisters of the period bears witness to this. It is highly likely thatthe "tiifenkendaz" groups (those who used firearms)thatthe Ottoman governmentemployed were these levends of peasant origin, whose numbersappearto have been constantlyincreasing in the Anatolian countrysidein the last quarterof the century,or even earlier,ratherthanbeing peasantswho, despite all difficulties,stayedin theirvillages to continuetheirmodestlife. We do not yet know, however,the real extent in of the crucial phenomenon of what can be termed "levendization" rural Anatolia, which seems to have developed moretowardindependentbrigandageor employmentas

Ottoman"DemographicCrisis" Reconsidered 189 ratherthan intermittent sekbanand sarica in the retinuesof provincialadministrators,54 as mercenariesby the government.It is thereforehighly unlikely that the employment peasants' leaving their villages (giftbozanhk),which had intensified prior to the great Celali devastation,can be fully explained by the "pull"factors without knowing the real extent of this levendizationand without knowing how many of these groups were It employed by the governmentas mercenarytroops and how often.55 is also important in this context to keep in mind the critical difference between the peasants' hopes and searchfor a betterlife in cities and the despairthat hopelessly scatteredthem in search of otheroptions such as brigandage.It can even be suggested that,comparedwith other in opportunities cities, brigandageper se was a more attractiveoption for them.
THE 17TH CENTURY: A "DEMOGRAPHIC CRISIS"?

While the rapidpopulationgrowthof the 16th centuryseems well established,research on various parts of the empire, including Anatolia, the Balkans, and Syria, points to an opposite phenomenon from the turn of the 17th century onward:a serious fall in population.56 Signs of the changein this directionareobservedfromthe late 16thcentury The main argumentamong scholars onward,becoming markedin the 17th century.57 on with the subjecthas focussedprimarily the extentof the decreasein population. dealing Historiansworkingon this periodreferagainto the disputednatureof the sources,on the on one hand,andthe problemof interpretation, the other.How reliablearethe sourcesof the 17thcentury-namely, the avariz andcizye registers,which provideonly quantitative datafor demographicdevelopments?How can the picturerevealedby these sources be Some go furtherto ask whetherthere was any real decrease in population, interpreted? while others present the decrease as an obvious historical fact, speaking of a serious "crisis"or even a "catastrophe." As mentionedearlier,McGowan developed the thesis of "demographic catastrophe" on the basis of his examinationof these registers58 belonging to the Balkanprovinces.He startsby observing a dramaticdropin the taxable populationrecordedin these registers and concludes thatthis was a manifestationof a serious demographiccrisis thatin some cases reached catastrophiclevels. According to McGowan, this was mainly the result of (1) the long wars and chaotic events of the.period; and (2) the dispossession of the peasantryunder an increasingtax burdenand exploitation. However, he does not rule out the possible effects of other factors that may well have contributedto this result, such as famine, typhus or plague epidemics, or the climatic change in Europe which is generally called the "LittleIce Age." Some historiansclaim that this climate change manifesteditself in the OttomanEmpireas increasingrainfallandunseasonablefreezing and occurrenceof heavy snow.59 Criticizingthe approachesthat tend to analyze the issue within the disputedcontext of the "17th century crisis," Todorova,maintains that the changes that took place in the demographicstructureof the OttomanEmpire during the 17th century cannot be understoodin such a framework.60 She arguesthat demographicphenomenahave their own distinctrules and chronologyof developmentand thatthey should not be evaluated in terms of conjuncturaleconomic and political developments.61 Therefore, it would be erroneous to link the population growth of the 16th century necessarily to social progress, and adverse development to the so-called crisis. Referring to McGowan's

190 OktayOzel argument,Todorovaraises a question: leaving aside the methodological problem of whether the populationdecrease can be considered a sign of demographiccrisis, did such a population fall in fact occur in the OttomanEmpire in the 17th century? She then goes on to question the extent to which the drop in population that is observed in the available sources representeda real loss. To Todorova,this drop can well be accountedfor by certainhistorical developmentsof the period, such as migrationand of re-nomadization,large-scale abandonment villages by peasants, or their evasion of the apparentfall she refers to in the non-Muslim populationof registration.Similarly, the Balkansin this context may be seen to be a false decrease.62 The first point to be emphasized in this part of the debate is that the problem of of interpretation the relevant data, contained in the sources used by both McGowan and Todorova,is valid for other similar material,including the tahrir registers.There is no doubt that every single piece of research requires the utmost attention in this respect. It should be remembered,however, that the collections of sources employed in this discussion belong to two periods-the 1530s and the 1700s-neither of which includes any part of the 17th century.The degree to which the nearly 170-year-gap between these dates allows us to analyze the long-termdemographicdevelopmentsis this highly questionable.Furthermore, line of argumentclearly says nothing aboutthe short-termfluctuationsthat took place in the OttomanEmpirein the late 16th and first half of the 17th century.To develop a more meaningfuland sound argument,therefore, one should make use of the same kind of sources for these periods or search for other sources availablein the Ottomanarchives. Recent researchhas revealedthe importanceof a new series of archivalsources.The most significant perhaps are the detailed avariz registers, which appearto have been compiled for the first time for various parts of the empire in the first quarterof the 17th centuryand continuedduringthe rest of the century.These are differentfrom the summary-typeavariz-hane registers used by McGowan. Preparedin the same way as the tahrir registersof the previous century,the detailed avariz registers enumeratethe entire tax-payingpopulationas "nefer"(adult men, marriedand unmarried) various in categories,as well as the membersof the rulingclass (askeri)who in one way or another held possessions liable to avariz taxes or extraordinary levies, which were turnedinto regularannualpaymentssometime aroundthe turnof the 17th century.63 The few studiesundertaken these sourcesin comparisonwith the tahrirregistersof on the late 16thcenturypoint to a radicaldecreaseof around80 percentin the recordedtaxAnatoliandistrictsof Amasya, Canik,and Bozok paying populationof the north-central in the firsthalf of the 17th century,with a correspondingfigureof around70 percentin the districtof Tokat(See Table 1).64 In the case of Amasya,30-40 percentof the villages that existed in the 1570s appear by the 1640s to have been abandonedor ruined. A similarpattern,thoughless dramatic,is observablein the neighboringdistrictsof Canik, Bozok, and Tokat (See Table 2)."6 A significant portion of the villages in the district of Amasya, some of which seem to have disappeared,were those that emerged in the either periodof the 16th-century expansionwith relativelysmall numbersof inhabitants on fertile plains or high plateaus.66 This was accompaniedby the disappearance the of etrakiyevillages of the mountainfringes. Similarly,thereis evidence thatthe Turkomans of the Bozok region of centralAnatolia,who had graduallyadopteda sedentarylifestyle to duringthe 16thcentury,had largely returned nomadiclife by the mid-17thcentury.67

Ottoman "DemographicCrisis" Reconsidered 191

TABLE1

Changes in tax-payingpopulation between the 1560s and the 1640s (in nefer)a
1560-70s 1640s %

Urban Tokat Amasya Merzifon Gtimtii Lddik Samsun Gedegrab Harput Rural Amasya (kaza) Samsun(sancak)c Bozok (sancak) Harput(kaza)

3,868 (1,258) 2,835 (1,069) 1,783 (770) 1,176 (524) 833 (248) 520 (229) 97 (42) 1,965 (403) 28,449 (12,923) 39,609 (18,063) 41,484 (22,780) 15,379 (4,147)

3,858 1,736 957 (33) 317 (30) 260 134 (58) 739 348 6,068 (833) 6,617 (1,181) 4,621 (252) 1,476 (615)

+0.3 -38.8 -46.3 -73.1 -68.8 -74.2 +66.1 -82.3 -78.7 -83.3 -88.9 -90.4

aFigures in parenthesesindicate the numbers of unmarriedadult men already included in the totals. To make the comparisonmeaningful, I have excluded a number of askeris recordedin the 1642 register.Therefore,the figuresin both dates presenttax-payingreaya only. bTheexceptionalincreasein the populationof the town of Gedegrais apparentlydue to its top-hill location. With its naturalprotection,it must have served as a perfectrefuge for the displacedpopulace from nearbysettlementson the low plains. kazas of Unye andTerme,which do not appearin the 1640s registers,are not included CThe in these totals. Also note that the kaza of Arim in the 1640s correspondsto roughly half of its area in 1570. The other parts of the kaza were divided in the 1640s into new kazas, which do not appearin the registers.This is also the case for the figures given in Table 2.

It should not be forgotten that this was a period with a number of extraordinary historicaldevelopments,mainly connected with the Celali depredations.It is the period in which the sources increasinglyspeak of frequent"Celali invasions"and of members class (ehl-i irf) roamingthe countrysidewith of the provincialmilitary-administrative their retinues of hundredsof horsemen under the pretext of inspection. At the mercy of the Celali bands and these brigandofficials, the peasants dispersed ("perakendeve
TABLE

2 Decrease in the number of villages between the 1570s and the 1640sa
1570s 372 509 629 1640s 228 452 548 % -38.70 -11.19 -12.87

District Amasya (kaza) Canik(sancak) Bozok (sancak)

aNote that the numbers for the 1640s include the "new" villages appearingonly in the survey of this date, although some of them may have been the old settlements with new names.

192 OktayOzel perigan olub"), leaving their villages en masse ("celdy-i vatan idiib"). City dwellers were not immune to such attacks,either. Contemporary sources unanimouslyrefer to the famines frequentlywitnessed in the countrysideand to the enormousdamagethey causedto the statetreasury "memlekete kitlik,devlethazinesinekiillizarargelmegle").68 ( the combined effects of these events on ruralstructure and village life in Furthermore, the Anatoliancountrysideare likely to have had an adverse effect on the birthrate, the real extent of which may neverbe known because of the shortcomingsof the available sources. To this should be addedthe increase in the deathrate underconditions of constant and widespreadCelali terrorand wars, which would have affected not only adult men, but also women, children,and elderly people-that is, those who were most vulnerableto humanandnatural calamities.69 of these takentogetherwith the possibility All of the phenomenonof late marriageturninginto one of temporarynon-marriage point to extraordinary historicalcircumstances.Comparedwith the generalconditions of the 16thcenturythatallowed, mainlythroughmilitaryexpansion,the growingpopulationto integrateinto an expandingsystem, the 17th centurywas a period of shrinkingmilitary and economic resourcesthatcreatedthe conditionsfor a generalcrisis and depredation. Contraryto Todorova'sargument,therefore, it is not mere speculationto speak of a general demographiccrisis-at least, for OttomanAnatolia in the first half of the 17th century. Whethersuch a crisis was a generalphenomenonin the entireempirein this periodand, if it was, whether there was any degree of recovery during and after the time of Kiprtillis in the later part of the century-can be shown only throughfurthercase studies.70 The question of the extent of the Celali terrorthat appearsto have continued throughoutthe 17th century in different parts of the empire should be kept in mind when examining the problem.Particularly importantin this respect is the extent of the terror'sdestructive effects on ruralstructure," given the facts thatthe ruraleconomy,both agriculturaland pastoral,was the main source of wealth for the imperialtreasuryand that the complex relationshipsof revenue distribution,which constitutedthe backbone of the whole militaryand administrative structure the empire, were based mainly on of the stabilityof both rurallife and the economy. Also crucial is the frequencyof natural disasterssuch as famine,epidemics,drought,earthquakes, floods, andheavy snow in the OttomanEmpireduringthe 17th century.72 It should immediately be pointed out, however, that the apparentdecrease in the recorded tax-paying population in the early-17th-centuryregisters employed in this study does not necessarily imply that 70-80 percent of the rural population simply died as a result of wars or naturalor human-madedisasters. A significant proportion of this "loss" in populationmay well be accountedfor by many peasants' formingthe human source of the hundredsof Celali bands that were still active in the Anatolian countrysideat the time of the surveys in the 1640s. Alternatively,some peasantsmay simply have evaded registration,thus going unrecordedin the registers.One can only speculate about this point. Nevertheless, the early-17th-centuryloss of populationas reflectedin the contemporary in surveyregistersand interpreted this study is too high to be explained only by such possibilities. Even if these are taken into account, it is more thanlikely thatthe picturepresentedby these registersstill remainsthe most significant evidence for a seriousdemographic fluctuation OttomanAnatoliaatthe turnof the 17th in century.73

Ottoman "DemographicCrisis" Reconsidered 193

THE OTTOMON CASE IN COMPARATIVE PERSPECTIVE

Let us turn at this point to the largercontext of the natureof these demographicdevelopments. Inalclk considers the case of late-16th-centuryoverpopulationin Anatoliaor, as Cook puts it, the apparentimbalance between economic resources and the increasing population-to be an overall "populationcrisis" with social and economic complications.74 ConsideringCarlo Cipolla's assertion that, in pre-industrialagrarian societies, fluctuationssuch as sudden and drasticfalls in populationcould be expected when populationgrowthexceeded certainlimits,7 it seems quitereasonableto approach the extraordinary demographicmovements,whetherrapidgrowthor drasticfall, as two phases of a general crisis.76Approachedfrom this perspective,the populationpressure thatCook suggestsfor the secondhalf of the 16thcenturycan also be seen as an indication of such a crisis in OttomanAnatolia. In the light of the findings of recent research, the period from the mid-16th to mid-17th century,with its up-and-downswings, may therefore be considered a period of general crisis in the demographichistory of the OttomanEmpire-a crisis whose first stage manifesteditself in the form of "pressure" and (or overpopulation), the second stage in the form of "implosion"(or depopulation). If true,does this take us back to the neo-Malthusian"population cycle," which has long constitutedthe centraltheme of scholarlydebates in demographicstudies?77 of The scope of the present study is limited to the re-interpretation old evidence in the light of new evidence concerning the 16th- and 17th-centurypopulation changes in Ottoman Anatolia in the hope that it will contributeto the revival of the debate among specialists. Although takingthe presentexaminationbeyond this point deserves a separatestudy, it is not totally without benefit to make some brief remarkson these questions to place the Ottomancase in the wider theoreticalcontext of the worldwide populationmovementsin the early modernperiod. Therole of populationchangesin historyhas been a subjectfor bothdemographers and historianssince the publicationof the classic worksof T. R. MalthusandDavidRicardo.78 Based on their argumentsaboutthe natureof populationmovementsin history and the relationshipsbetween populationand the economy, which have often been regardedas too mechanicalto comprehend complex natureof historicaldevelopmentandexplain the its diversity,there emerged in the 20th centurymany revisionist attemptsto modify or refinethe MalthusianandRicardiandemographic"laws"or to refutethem categorically. The resultantdebates among scholars have thus evolved aroundwhat is termed the "neo-Malthusian" approach,among whose principaldefenderswere historianssuch as EmmanuelLe Roy LadurieandM. M. Postan.79 was mainly on theirworksconcerning It late medieval and early modernFrance and England that Robert Brennerlaunched in the late-1970s a counter-argument rejecting the primaryrole of demographicchanges in the rise of Europeancapitalism in general and in income distributionin particular. However, he never categorically denied the importance of what he referred to as "demo-economic" trends in long-term historical developments.80What he sharply criticized was the mechanistic application to history of demographicmodels, which have almost been exclusively associatedwith Malthusvia Le Roy Laduriein particular. With the participationof other specialists, the "Brennerdebate" led to a productive discussion among historiansthatwas to have a stronginfluence on laterhistoriography. It greatly contributedto the worldwide shift in historiographyfrom the emphasis on

194 OktayOzel demographic-economic processes as the main factors in historical change towarda level, resultingfinallyin bringingthe "state" greaterfocus on the political-distributional back into historicalanalysis in the 1980s and 1990s.81 Concurrently-or, perhaps,as a reactionto this tendency-some scholars, the most prominentof whom was Jack Goldstone,returnedto the primaryrole of demographicecological changes in the developmentof history.82 Accordingto Goldstone,population in principlemovedindependently reasonsexogenous even to the economy andplayed for a centralrole particularly the political crises of early modernsocieties.83 Goldstone's in "post-Malthusian" approachonce more brought attentionto the role of demographic factors in history on the widest scale across time and space, covering areas stretching from Europeto Chinaand in the periodfrom the late medievalages to the 20th century. All of these debates have found echoes in Ottomanhistoriography. Islamoglu-Inan, who wrote in the 1980s mainly about the agrarianeconomy of Anatolia, also touched on populationchanges in Anatolia. She closely followed the currentdiscussionsrevolving aroundWallerstein's"capitalistworld system" approachalong with the Brenner debate,with certainreservationstowardboth based to some extent on the worksof Ester Boserup.84I have already discussed islamoglu-Inan's argument,which places heavy emphasison the determiningpower of the state and its role in socio-political anddistributionalprocesses in the OttomanEmpire. Although I agree with her in rejectingany deterministicmechanicaland reductionist single-factor approachin history, her somewhat eclectic theoretical approachunderestimates the precariousbalance between population and resources that were in fact societies. I also agreewith closely connectedin late medievalandearly modem agrarian her that the roots of populationchanges are not necessarily internalto the agricultural economy. But this does not mean that demographicchanges-rises or falls-have no negative effects or do not put strain on the economy in general and state finance in Indeed,Goldstone'sentireworkconvincinglyshows how populationchanges particular. that occurredoften syncronously across the world during the 16th and 17th centuries led to eventualstatebreakdowns,following strikinglysimilarpatterns,albeitin different forms. Goldstonehimself includedthe OttomanEmpirein his comprehensivestudyof "state breakdowns"in the early modernperiod. His post-Malthusiandemographicapproach, which in the main arguesthat "revolutionsare the result of multiple problems,arising from long-term shifts in the balance of population and resources,"85 deserves closer attentionbecause of its direct relevance to the central argumentof the present study. Goldstone develops the argumentthat the more or less simultaneousstate breakdowns during the late 16th and 17th centuries in Europe, the Ottoman Empire, and China were the best examples of the recurrent waves of similar events in history.All of these originated mainly from a periodic, cyclic imbalance between population growth and inflexible economic and political systems.86In this respect, Goldstone's treatmentof the Ottomancase places the price increases of the late 16th century,the crisis in state finance,and the widespreadCelali rebellionsinto this worldwidecontext. In doing this, he refersto old evidence concerningthe socio-economic and political manifestations of this periodof crisis, such as the increasein the numberand overall proportion young of unmarried men in the population,the fragmentation peasantfarms, and the increase of in landlessness. He then emphasizes the peasants' ensuing search for other means of

Ottoman "DemographicCrisis" Reconsidered 195 livelihood outside their villages. All of this eventuallycontributeddirectlyto the Celali uprisings. The new findingspresentedin this study once more confirmand consolidatethis picis ture.However,a moreimportant pointin Goldstone'sargument thatpopulationgrowth has a non-linearor disproportionate effect particularly marginalgroups-in our case, on the unmarried men andthe landless.87 in Using Goldstone'sown words,"[I]ncreases total populationgenerallyproducea muchlargerincrease"in these marginalpopulations"... thanin the populationas a whole.""88 Leaving aside Goldstone's other arguments,which obviously open new horizons for futurecomparativestudies in Ottomanhistory,this point alone is particularly important for the argumentof the presentstudy.If his argumentis correct,the figurespresentedin this study become more significantbecause they show a substantialincrease in the size of the sectors of ruralsociety that were gradually"marginalized" underthe conditions of populationgrowth.If so, one might expect even further increasesin these populations in the last quarterof the 16th century-increases that cannot be observed because we have no surveys availablefor this period. When the landless and unmarried young men are taken togetherwith the discontentedtimarholders, who had also lost much of their income underthe inflationary trendthatwent handin handwith populationgrowth,it is not a coincidence thatthese were the very groupsthatformedthe main source of Celali bands and rebel armiesin the 17th century.One can furtherassume that the large-scale destructioncaused in the Anatoliancountrysideby the Celali terror,coupled with wars, of resulted in what Brennerdescribes as the "disruption productionleading to further demographicdecline, ratherthan a returnto equilibrium."89 This last point, which is perhapsBrenner'sonly contributionto the neo-Malthusian to debate, although he developed it as a counter-argument the theory of Malthusian relatesto the very point at which we startedto evaluatethe Ottomancase in adjustment, a widerhistoricalcontext.Implicitin my line of argument this throughout studyis thatthe in Malthusianapproachstill has meritin populationstudiesandoffers much,particularly termsof the natureof demographicchanges in essentially agrarian societies. One should also rememberthe remarksof anotherprominenthistorian,Guy Bois, thatknowledge of the demographicchangesis essentialto understanding developmentof societies in which small-scalefamilyproduction thebasic economic unitandin which "reproduction is takes As process."90 such a society, place on thatscale accordingto an economic/demographic the Ottomanpeasantrywas vulnerableto demographicchanges, and the developments in 16th- and 17th-century ruralAnatoliacan be re-interpreted this context.91 in Does this take us to demographicand economic determinism?Certainlynot. No reasonablemind can suggest such a deterministicapproachafterthe decades-long debatesover the complex natureof historicaldevelopment.What this may mean instead is that demographic analysis can be furtherdeveloped andrefinedto widen our perspective,as impressively exemplifiedby the works of scholarssuch as Cipolla, Goldstone, and many others.
CONCLUSION

Whateverthe fruits of discussing the problemat such a theoreticallevel, in the case of Anatolia it is perhapsmore importantto bear in mind the geographicaldimension of the populationchanges in the late-16th- and early-17th-century OttomanEmpire. The

196 OktayOzel crucial question is how representativethe cases of demographicpressurein Anatolia one described here were as far as the whole empire was concerned.92 Furthermore, ask the same question for Anatolia only, considering the fact that in some parts may of Anatoliathe populationseems to have remainedwithin reasonablelimits,93 although the substantial growthin the 16thcenturywas a generalphenomenonthroughout Empire. It is thereforeimperativeto pay attentionto voices that emphasizeregional differences in terms of demographicchanges-differences that depended largely on the quality and quantityof the land, climatic conditions, economic opportunities,and, as Karen Barkeyrightly suggests, the patron-clientrelationsat the local level and in the empire in general.94 It is also clear that population growth does not necessarily or automaticallymean What this study shows in this respect is that one can speak of such pressure "pressure." in at least some partsof the empire-in this case, the north-central Anatolianprovince of Rum. Whetherthe apparent rise in populationresultedin similar pressureelsewhere in Anatolia or throughoutthe empire towardthe end of the centuryremainsa question. Nevertheless,this study has also pointed out that the Celali rebellions and widespread terrorin the Anatoliancountrysidewere closely related to the demographicgrowthof the 16th century. of At this point, it is importantto returnto the sources, the natureand interpretation which constitutean significantpartof the debate.Thereis no doubtthatthe comprehensive series of imperialtahrir,avariz, and cizye registersof variouskinds (separateevkaf datafor demographicstudies, tahrirregistersincluded),which offer the only quantitative have been, and still are, the principlesources. But it has increasinglybecome apparent in thatthe qualitativeinformationthat these sources provideis equally important terms, for example, of settlementpatterns,and abandonedor lost settlements.Miihimmeand sicil collections availablefor the periodin question, as well as other archivalmaterials such as the account books of certain foundations(vakifs),95 often provide useful and sometimesextremelyimportant insights into the complex historicaldevelopmentsof the time. Only throughcross-examination one make a reasonablyconvincingevaluation can of demographicchanges in general,andof the degreeof reliabilityof the figuresgiven in the sourcesin particular. Nevertheless,the varyingroles of factorsaffectingthe birth-todeathratioremainan important issue thatis unlikely,perhapsimpossible,to clarifygiven the shortcomingsof the present sources.96 However discouragingrepeatedmentionof such methodologicalproblems and the questioningof the reliability and shortcomings of the source materialmay be, there seem to be no easy solutions to the problemsof Ottomandemographichistoryof the periodin question.
NOTES Author's note: The initial version of this article was presented at the Eighth International Congress of Economic and Social History of Turkey,Bursa, 18-22 June 1998. The authorthanks Halil Inalcik, Rifa'at Ali Abou El-Haj, Paul Latimer,Mehmet Oz, FikretYilmaz, and anonymousIJMESreviewers,as well as its editors, for theirvaluablesuggestions. 1See esp. 0. L. Barkan, "Ttirkiye'de ImparatorlukDevirlerinin Biiytik Ntifus ve Arazi Tahrirleri ve Hakana Mahsus Istatistik Defterleri," istanbul iniversitesi iktisat Fakiiltesi Mecmuasl 2, 1 (1940): 20-59: ibid., 2, 2 (1941): 214-47; idem, "TarihiDemografi Aragtlrmalanr OsmanhlTarihi,"Tiirkiyat ve Mecmuast 10 (1951-53): 1-27; idem, "Essai sur les donnees statistiquesdes registres de recensementdans

Ottoman "DemographicCrisis" Reconsidered 197

l'Empire Ottoman aux XVe et XVIe siecles," Journal of the Social and Economic History of the Orient 1, 1 (1958): 9-36. See also his later works on the subject: "Researchon the OttomanFiscal Surveys,"in Studies in the EconomicHistory of the MiddleEast, ed. Michael A. Cook (London:OxfordUniversityPress, 1970), 163-71; and "894 (1488/1489) Yili Cizyesinin TahsilatlnaAit Muhasebe Bilangolari," Belgeler 1, 1 (1964): 1-117. 2For a detailed re-evaluationof the related literaturewithin the larger frameworkof the Braudelian Mediterranean world, see Halil Inalcik, "The Impact of the Annales School on OttomanStudies and New Findings,"Review 1, 3-4 (1978): 69-96. 3Thehistorianwho firstused the term"defterology" HeathLowry,himself being a prominentdefterolwas works,as well as his discussion of some methodologicalproblemsinvolved ogist. Forhis majormonographical in the use of these defters,see his Trabzon 1461-1583 (Istanbul: ?ehrininIslamlaymave Tiirklegmesi, Bogaziqi UniversitesiYaylnlan, 1981); idem, Studiesin Defterology,Ottoman Societyin the Fifteenthand SixteenthCenturies (Istanbul: Yaylnevi, 1992). For a criticalevaluationof this field, see esp. Colin Heywood, "Between Isis HistoricalMyth and Mythohistory:The Limits of OttomanHistory,"Byzantineand Modern Greek Studies 12 (1988): 315-45; and for a more recent critique,see FatmaAcun, "OsmanhTarihininGenigleyenSinirlan: Defteroloji,"TiirkKiiltiiriiincelemeleri Dergisi 1 (2000): 319-32. For anotherwork that deals well with the CanikSancagi, (Ankara: TTK majorproblemsof defterologicalstudies, see MehmetOz, XV-XVI.Yiizyillarda Basimevi, 1999). The numberof monographicalstudies in local history for which these defters constitute the principalsources has increasedsubstantiallyin the past two decades. These works have also contributed significantlyto the developmentof complicatedterminologyandtheproblemsof Ottomandemographic history. For the most important,see Michael A. Cook, PopulationPressure in RuralAnatolia, 1450-1600 (London: Oxford University Press, 1972); Leyla Erder,"The Measurementof Pre-industrial PopulationChanges:The OttomanEmpirefrom the 15thto 17thCentury," MiddleEasternStudies 11 (1979): 284-301; Leila Erderand Rise and Fall in Anatolia, 1550-1620," Middle East Studies 15 (1979): 328-45; SuraiyaFaroqhi,"Population BekirKemalAtaman,"Ottoman Journal DemographicHistory(14th-17th Centuries).Some Considerations," ofEconomic and Social Historyof the Orient35, 2 (1992): 187-98; MariaTodorovaandNikolai Todorov,"The HistoricalDemographyof the OttomanEmpire:Problemsand Tasks,"in Scholar Patriot,Mentor:Historical Essays in Honor of DimitrijeDjordjevic,ed. RichardB. Spence and Linda L. Nelson (Boulder,Colo.: East EuropeanMonographs,1992), 151-72. For a bibliographicalessay on populationmovementsin the Ottoman Empire, see Daniel Panzac, "La Populationde l'Empire Ottomanet de ses Marges du XVe au XIXe Siecle: Revuede l'accidentMusulmanet de la MWditerrande (198 1): 31 Bibliographie(1941-80) et Bilan Provisoire," 119-37. 4See Barkan,"Ttirkiye'de Devirlerinin." imparatorluk 5The only exception to this in Turkeywas Halil Inalcik's publicationof the earliest extantregisterin the Ottomanarchive, relating to Albanian lands. See Halil Inalclk, Hicri 835 TarihliSuret-i Defter-i Sancak-i Arvanid(Ankara:TTK Basimevi, 1954). MardinTarihi(Istanbul: U. EdebiyatFaktiltesiYayini, 1969;repr.Ankara: XVI. Yiizyilda i. 6NejatGbytinq's TarihKurumu,1991) deserves special mentionhere in thatit was the first example of this kind of study Tiurk in modernTurkey. 7Bruce McGowan, Economic Life in OttomanEurope: Taxation,Trade,and Strugglefor Land, 16001800 (Cambridge:CambridgeUniversity Press, 1981); Linda Darling, Revenue Raising and Legitimacy: Tax Collection and Finance Administrationin the OttomanEmpire, 1560-1660, (Leiden: E. J. Brill, 1996); Oktay Ozel, "Changesin Settlement Patterns,Population and Society in Rural Anatolia: A Case Study of Ph.D. thesis, Universityof Manchester,1993); idem, "17. Yiizyll Osmanh Amasya, 1576-1642" (unpublished Demografive IskanTarihiIgin Onemli bir Kaynak:'Mufassal'AvarizDefterleri,"XII. TiirkTarihKongresi, TTK Yaylni, 1999), 3:735-44. Here Machiel Ankara,12-16 Eylfil 1994, KongreyeSunulanBildiriler (Ankara: Kiel deserves special mention for his works, each of which are among the most significantcontributionsto in the field particularly terms of the discussion of the problematicnatureand utility of these sources. See his "Remarkson the Administration the Poll Tax (Cizye) in the OttomanBalkans and the Value of Poll Tax of Registers (Cizye Defterleri) for DemographicResearch,"Etudes Balkaniques4 (1990): 70-104; Ayni Yazar, "AnatoliaTransplanted? Patternsof Demographic,Religious and Ethnic Changes in the District of Tozluk The (N. E. Bulgaria), 1479-1873," Anatolica 17 (1991): 1-27; idem, "Hrazgrad-Hezargrad-Razgrad: Vicissitudes of a TurkishTown in Bulgaria,"Turcica21-23 (1991). 8While the main objective of this study is not to discuss well-known but still little appreciatedaspects of defterological studies, because of the natureof the sources and the question of the reliability of the data

198 OktayOzel
they contain, I believe that it is imperativeto remindthe readerof the fact that all the argumentsdeveloped and discussed here are based on the records of the tax-payingpopulationonly, both ruraland urban,whose status was well defined by law and regularlyand systematicallyrecordedwith the utmost care in the survey registers. Other sectors of the society at large, including marginalgroups such as gypsies, generally went unrecorded.Similarly, a certain portion of the peasantry might have not been recorded because of their particularservices to the government,although we know that in most cases they were also included in the a registerswith a mention of their special statuseven if they were tax-exempt.Furthermore, large portionof urbansociety-members of the military class, for example-were not subject to systematic survey.Despite all this, a regularlyand systematicallyrecordedportionof Ottomansociety constitutesin itself an important databasefor historicaldemographicinquiry,allowing us to clearly follow the main populationtrendsas well as certainaspects of demographicchange. What follows is an example of this kind, of study, and like other such studies, it should be read with these limitationsin mind. For a discussion of the subject, see Mehmet DefterlerininOsmanliTarihiAra\unhboxgtirmalarmnda Oz, "Tahrir KullanllmasiHakkindaBazi Diiptinceler," Vakiflar Dergisi 22 (1991): 509-37; FikretYilmaz, "16. Yiizyllda EdremitKazasl"(unpublishedPh.D. thesis, Ege University, Izmir, 1995), 192-205. On avariz and cizye registers, see Oktay Ozel, "Avarizve Cizye in DE, 2000), Defterleri," OsmanliDevleti'ndeBilgi ve Istatistik,ed. Halil inalclk andSevket Pamuk(Ankara: 33-50. 9Since it is unnecessaryand practicallyimpossible to give here a complete list of defterological studies that do not deal totally with populationchanges in the OttomanEmpireduringthe 16th century,I will refer only to those mentionedin n. 3. A relatively recent publicationthat discusses the relevantfindings of these studies is Oz, Canik.See also Kemal (iqek, "Tahrir DefterlerininKullamminda GtiriflenBazi Problemlerve Metod Araylglarn," DiinyastArattirmalart97 (1995): 93-111; inalclk, "Impactof the Annales School." Tiirk 10Cook,PopulationPressure. "1Mustafa Akdag, Celali isyanlari, 1550-1603 (Ankara:AnkaraUniversitesiYaylm, 1963). (For a later, extended version, see Celali Isyanlart. TiirkHalklninDirlik ve DiizenlikKavgasi (Istanbul:Bilgi Yayinevi, Tekin Yayinevi, 1971). Iktisadive 4&timai Tarihi,2 vols. (Istanbul: 1975). Idem, Tiirkiye'nin in 12See Halil inalclk, "Militaryand Fiscal Transformation the OttomanEmpire, 1600-1700," Archivum Ottomanicum (1980): 283-337; idem, "Impactof the Annales School," 80-83; Huricihanislamoglu-Inan, 6 State and Peasant in the OttomanEmpire:AgrarianPowerRelationsand RegionalEconomic Developmentin OttomanAnatolia during the SixteenthCentury(Leiden:E. J. Brill, 1994). 13inalclk,"Militaryand Fiscal Transformation"; islamoglu-inan, State and Peasant, 185. See also Suraiya in Faroqhi, "Political Tension in the Anatolian Countrysidearound 1600: An Attempt at Interpretation," et Miszellen, RobertAnhegger Festschrift,Armagani,Melanges, ed. J. L. Bacque-Grammont al. Tiirkischhe (Istanbul,1987), 117-30. 14Forthe details of the first stage of the uprisingsand the natureof the Celali rebellions in general, see William J. Griswold, The GreatAnatolian Rebellion, 1000-1020/1591-1611 (Berlin: Klaus Schwarz, 1983); Karen Barkey,Bandits and Bureaucrats:The OttomanRoute to State Centralization(Ithaca,N.Y.: Cornell UniversityPress, 1994). 1 have alreadyreferredto M. Akdag's classic Celali isyanlart. S5See, for example, McGowan,Economic Life in OttomanEurope.See also Halil inalclk, "Adaletnameler," Belgeler 2, 3-4 (1965): 49-145; idem, "The Ottoman Decline and Its Effects upon the Reaya,"in Aspects of the Balkans, Continuityand Change: Contributionsto the InternationalBalkan Conference,University of California, Los Angeles 1969, ed. H. Birnbaumand S. Vryonis (The Hague: Mouton, 1972), 338-54. SuraiyaFaroqhi,however,emphasizesthe "political"natureof peasants'exodus from the villages undersuch on conditions,thusexpressingherdoubt,apparently the basis of the worksof islamoglu-inan,aboutthe role of a demographicpressure.See SuraiyaFaroqhi,"PoliticalActivity among OttomanTaxpayersand the Problem of Sultanic Legitimation(1570-1650)," Journal of Economic and Social History of the Orient 35 (1992): 38-39. 16inalcik,"Militaryand Fiscal Transformation." 17McGowan, EconomicLife in OttomanEmpire. 1SMaria Todorova,"WasTherea DemographicCrisis in the OttomanEmpirein the SeventeenthCentury?" EtudesBalkaniques2 (1988): 55-63. State and Peasant, esp. chap. 4. 19islamoglu-inan, 2(See, for example, Faroqhi,"PoliticalActivity";idem, "Crisisand Change, 1590-1699," in An Economic and Social History of the OttomanEmpire, 1300-1914, ed. Halil Inalclkand Donald Quataert(Cambridge: CambridgeUniversityPress, 1994), 433-38; idem, "Seeking Wisdom in China:An Attemptto Make Sense

Ottoman "DemographicCrisis" Reconsidered 199

of the Celali Rebellions,"in Zafar Name: Memorial Volumeof Felix Tauer,ed. Rudolf Veselly and Eduard Gombar(Prague:EnigmaCorporation,1996), esp. 104. 21islamoglu-inan,State and Peasant, 149-54, 156. 22Ibid., 143, 146-48. Frequentmovements of pastoral nomads and peasants from the less secure eastern provinces to the western parts of the empire during the course of Ottoman history seem to be a historical fact. See Halil Inalclk, "Introduction: Empire and Population,"in Inalclk and Quataert,Economic and Social History, 31 ff. However, we have no clear evidence of any significant migration taking place in the region in question in the period concerned. Furthermore, recent research has revealed that the high level of growth in population was the case not only in the western cities but also across Anatolia, in both urban and rural areas. See, for example, Ismet Miroglu, XVI. YiizylldaBayburd Sancagt (Istanbul, 1975); idem, Kemahve Erzincan Kazasi (1520-1566) (Ankara:TTK Basimevi, 1990); Mehmet Ali Unal, XVI. YiizylldaHarput Sancagt (1518-1566) (Ankara:TTK Basimevi, 1989); Orhan Khliq,XVI. Ve XVII, Van(1548-1648) (Van: Van Belediye Balkanllgi Ktilttirve Sosyal Iller Miidtirltigti Yaylnlarl, Yiizytllarda 1997). State and Peasant, 143, 154. 23islamoglu-Inan, 24Ibid., 156. There are many problems in this argument.First, it is highly questionableto assume that for enteringthe askeri class was a matterof "preference" Ottomanpeasants, given the strictrules delimiting such movements. Second, it seems to be chronologicallypremature speak of the existence of such retinues to composed of irregularsoldiers within the Ottomanprovincial-military organizationunder the timar system during the first three quartersof the 16th century,althoughCiftbozanlevend groups existed in the Ottoman countrysidewell before that century.The presentlevel of our knowledge of such retinuessuggests thatit was instead a phenomenonof the years prior to or during the Celali movements at the turn of the 17th century. of Third,speaking aboutthe "drudgery work"as a factor behind Anatolianpeasants' leaving the land while economic and demographic constraintsof the periodis not convincing.I will touch on these rejectingapparent issues later.Cf. Barkey,Banditsand Bureaucrats,150 ff. 25Islamoglu-inan,State and Peasant, 156. YunusKog, "XVI. Yiizylhn Ikinci Yarisinda 26See, for example, Oz, "OsmanhKlasikDiinemindeTarim"; Sorunu:Bursa Kazasi Ol6eginde Bir Araltirma," unpublishedpaperpresentedat the Kiylerin Parqalanmasi Eighth TurkishCongressof History,Ankara,4-8 October 1999. I thankDr. Koqfor permittingme to use this paper. 27See Huricihanislamoglu-Inan, "M. A. Cook's PopulationPressurein Rural Anatolia, 1450-1600: A Critique of the Present Paradigmin OttomanHistory,"Review of Middle East Studies 3 (1978?): 120-35. Islamoglu-Inandeals with the price rise in another article: see HuricihanIslamoglu and Suraiya Faroqhi, Anatolia,"Review 2, 3 (1978). For "CropPatternsand AgriculturalProductionTrendsin Sixteenth-Century the problem of price increases in connection with population growth, see also Mustafa Akdag, "Osmanh ve Imparatorlugu'nun Kurulu?u InkigafiDevrinde Tiirkiye'ninIktisadiVaziyeti,"Belleten 13 (1949): 497A 571. See also OmerLiutfi Pointin the Economic Barkan,"PriceRevolutionof the SixteenthCentury: Turning Journalof MiddleEast Studies6, 1 (1975): 8-15. Cf. Inalcik,"Impact History of the NearEast,"International of theAnnales School,"83 ff. The latest contribution the discussionis from SevketPamuk,who re-evaluates to the findings of Barkanand his interpretation price movementsin the OttomanEmpire:See Sevket Pamuk, of "The Price Revolutionin the OttomanEmpire Reconsidered," InternationalJournal of Middle East Studies 33 (2001): 69-89. State and Peasant, 149. 28Islamoglu-Inan, 29As far as the laterhistoriography concerned,it was KarenBarkeywho developed a systematiccritique is of Islamoglu-Inan'sargument.Barkeyarguesthat Islamoglu-Inan,along with inalclk, has overstatedthe role of "pullfactors"in peasants'leaving theirlands and emphasizesthe impact of declining economic conditions and rapid growthin population-in the landless and unmarried population,in particular-in Anatolia during the 16th century:see Barkey,Banditsand Bureaucrats,148 ff. 30See Oz, Canik;Ozel, "Changes." 31See Cook, PopulationPressure,25; Oz, "Tahrir Defterlerinin," 436. Cf. FeridunEmecen,XVI.Asirda 433, Manisa Kazast (Ankara:TTK Basimevi, 1989), 232-33. 32Theproportionof landless peasanthouseholdsto total householdsreachedmore than50 percentin some nahiyes of the kaza of Amasya (Ozel, Changes,75-76, 78). Note thatthese figureswere reachedvia a detailed examinationof the tax registerof the region dated 1576 (TD 26, Kuyud-1 KadimeArchive,GeneralDirectorate of Deeds andSurveys,Ankara)andincludeneitherthose peasantsrecordedin the registersas "caba"(landless)

200

OktayOzel

who in fact were co-cultivatinglands belonging to others, usually relatives,nor those using plots of land not yet allocated with "tapu"(the title deed) to a peasantfamily, the standards being simply recordedas zemin (land). 33For earlierreferencesto this misinterpretation its implications,see OktayOzel, "XV-XVI.Yiizylllarda and Kirsal (Zirai) Organizasyon:Koyliiler ve Kiiyler" (unpublishedM.A. thesis, Osmanh Imparatorlugu'nda Devlet ve K6ylti Hacettepe University, Ankara, 1986), 80-81; Mehmet Oz, "Osmanh Imparatorlugu'nda IligkileriHakkindaBir Kitap,"Tiirkiye Giinliigii 16 (1991): 151-56. For a similarcase, see ErderandFaroqhi, Rise and Fall." "Population 34Seeislamoglu-inan, State and Peasant, 209, TableVI/5. 81. ve 350zel, "Kibyltiler Kbiyler," See also Oz, Canik,48. OsmanliKanunnameleri HukukiTahlilleri(Istanbul: ve 36Yediylldiz,OrduKazasi, 150; Ahmed Akgiinduiz, Fey VakfiYayini, 1993), 6:203; ibid. (1994), 8:431. 37Cf. Erderand Faroqhi,"PopulationRise and Fall";MiibahatS. Kiittikoglu,XV ve XVI.Asirlardaizmir KazasininSosyal ve iktisadi Yapisi(Izmir: Biiytik Sehir Belediyesi Yayini, 2000), 67-116. Speaking of the of surveyor of an unrecorded change "possibility" the surveyors'overly meticulousattitudesin this particular in the ways of recordingof unmarried men who were not even liable to taxationin some partsof this province (see Yediylildiz,OrduKazast, 150; Oz, Canik, 44-45) seems to me to be an assumptionof an unnecessarily speculativeor overcautiousmind. See Cook, PopulationPressure, 13; cf. Oz, Canik,52. For a more detailed discussion of this assumption,see Ozel, "Changes,"148-49. 38AlthoughI havelimitedmy examinationto north-central Anatolia,similardevelopmentswere seen in other partsof Anatolia. For an example, see SuraiyaFaroqhi,"Peasantsof Saideli in the Late Sixteenth Century," ArchivumOttomanicum (1983): 215-50, in which Faroqhidevelops the argumentthat, despite the fact that 8 land was still abundant the region,it was not always easy for the peasantsto take new land undercultivation in mainly because of economic or technological constraintsand because of the geographicaldistance of such lands from theirvillages (see esp. p. 224). Cf. Barkey,Banditsand Bureaucrats,148-50. 39SeeOz, Canik,44-45, 49, 51-52; Ozel, "Changes," 69-70, 75-76, 146. The drasticincreasein the number of recordedunmarried all adultmenbetweenthe 1530s andthe 1570s seems to havebeen a generalphenomenon over Anatolia.Foranotherexampleof this, see Yilmaz, "Edremit," (thetotalgiven at the end of TableVIII). 160 Accordingto the table, this sector of the populationin ruralareasbetween the 1520s and the 1570s increased Anatoliandistrictof Edremit.It would have been interesting threefold,from429 to 1,289, in the northwestern to see the changes in the same period in the proportionof landless marriedmen in the overall total male does not provide the breakdownof these categories. populationin the same district.Yilmaz unfortunately 40CarloM. Cipolla, TheEconomicHistory of World Population,5th ed. (Baltimore:PenguinBooks, 1970), 83. Rise and Fall,"337. Cf. TodorovaandTodorov,"HistoricalDemog41See Erderand Faroqhi,"Population raphy,"161. TTK Basimevi, 2000), 343. 42See TuranG6kqe,XVI-XVII.Yiizytllarda (Denizli) Kazadst (Ankara: Ldzlktyye 43Ibid.,36-46, 179. 44See MargaretL. Venzke, "The Question of Declining Cereals Productionin the Sixteenth Century:A Journal of TurkishStudies 8 (1984): Soundingon the Problem-solvingCapacityof the OttomanCadastres," of esp. 261-64. The fragmentation villages and the formationof new villages by inhabitantsof existing ones can be seen as anotherdimensionof the same phenomenon. Koq,"XVI.Yiizyllin Ikinci Yarisinda See Kiylerin Sorunu." also Kiittikoglu,izmir, 58-59. See Pargalanmasi 45Despite the fact that the limits of the arable land were reached both in area and in productivityin the districtsof Amasya and Canik, it is puzzling to see the presence in 1576 of a few plots of land recordedin the registersas "unallocated" (mevkufzeminha).See, for example, TD 26; cf. Ozel, "Changes,"76-78. It is difficultto clarify whetherthese pieces of land were in realityleft uncultivated. presenceof such plotsmay The perhapsbe accountedfor by the possibility of factorsFaroqhirefersto in her "Peasantsof Saideli"(see n. 38). For a lengthy discussion of the problem,see Oz, Canik, 190-93. Nevertheless,consideringthe very sporadic natureof such cases in the region underexamination,I do not thinkthis makes any considerabledifferencein the overall economic and demographicconditions describedin these pages. aux 460n these developments,see Xavier de Planhol,De La Plaine Pamphylienne Lacs Pisidiens, Nomadisme et Vie Paysanne (Paris, 1958); idem, "Geography,Politics and Nomadism in Anatolia,"International Social Science Journal 11 (1959): 525-31; Wolf-DieterHitteroth, LandlischeSiedlungen im SiidlischenInneranatolienin den LetztenVierhundert Jahren(Gittingen, 1969); Halil inalclk, "Settlements," Inalclkand in

Ottoman "DemographicCrisis" Reconsidered 201

Quataert,Economic and Social History, 158-71; idem, s.v. "Mazra'a," Encyclopaedia of Islam, 2nd ed. We follow the earlierstagesof the processof gradualsedentarization nomadsin the increasingrecordsof villages of foundedby groupsreferredto as "etrakiye," "cemaat"or "taife"in the tahrirregistersof the 16thcentury.See, for example,KayseriandAmasyadefterspreservedin Ankarain the TapuKadastro Genel Miidfirltigti Kuyud-1 KadimeArgivi, TD 26, 44, 136. Cf. Islamoglu-Inan,State and Peasant, 177-78; FikretAdamr,"Mezraa:Zu Einem Problem der Siedlungs- und AgrargeschichteStidosteuropas Ausgehenden Mittelalterund in der im FriihenNeuzeit," in Deutschlandund Europa in der Neuzeit, Festschriftfiir Karl OtmarFreiherrvon aretin ed. zum65. Geburtstag, RalphMelville, Claus Scharf,MartinVoght,andUlrichWengenroth (Stuttgart,1988), 186 193-204; Emecen, Manisa, 235-36; Ktittikoglu, Izmir,97; Yilmaz, "Edremit," ff. Cf. 47See Ozel, "Changes." Faroqhi,"PoliticalTension,"129. On a similarphenomenonof high population densityin some regions in 14th-century Europe-density thatwas not to be reachedagain until the early 20th century-see CarloM. Cipolla,BeforetheIndustrialRevolution:EuropeanSociety and Economy,1000-1700, 3rd ed. (London:Routledge, 1993), 130-31. 48Cf.Kiittikoglu,izmir, 114. 49See Inalcik, "Introduction," 29-41; idem, "Impactof Annales School"; Faroqhi, "Political Tension"; Mehmet Oz, "Bozok SancagindaIskan ve Niifus (1539-1642)," XII. Tiirk TarihKongresi,Ankara, 12-16 It Eyliil 1994, KongreyeSunulanBildiriler(Ankara:TTK Yayini, 1999), 3:787-94. See also Ozel, "Changes." shouldbe emphasizedhere, however,thatapartfrom some generaland highly superficialremarkson sporadic cases, the real extent, with an empire-wide chronological map of such migratorymovements in Ottoman history,has not yet been documented. 156-57. 50Cf.Todorovaand Todorov,"HistoricalDemography," 51Also to be pointed out here is the pressurecoming from the ehl-i 5rf, who tended to raise, often through illegal means, the amountof the tax burdenof peasantsto compensatefor the shrinkingvalue of their dirliks or holdings. See Faroqhi,"PoliticalActivity."Cf. Amy Singer, "Peasant Migration:Law and Practicein Early OttomanPalestine,"New Perspectives on Turkey8 (1992): 49-65. In addition, compared with most other on partsof the empire, in the province of Rum the peasants paid twice as much tithe (&6r) crops produced because of the malikane-divanisystem that was widely applied in this part of Anatolia (see Oz, Canik, 124). 52Inalcik,"Impactof the Annales School,"73; Faroqhi,"PoliticalTensions,"119. 53As mentioned earlier, the other development, that appears to have been directly linked to the worsening economic conditions and demographicpressure was the tendency among young people to become students in medreses, where they thought they would secure a certain degree of self-subsistence, and as medrese graduatesthey hoped to enter the askeri class. See Akdag, Celali isyanlari. Cf. Giike, Ldztktyye, 36 ff. For a critical assessment of these options for young peasants, see Barkey,Bandits and Bureaucrats, 156 ff. with Akdag's Celali isyanlari, MustafaCezar'swork remainsthe most comprehensivestudy of 54Together this phenomenon.See MustafaCezar,Osmanli Tarihinde Levendler(Istanbul,1965). soldiers"played an important in the Celali rebellionsof the 55 part inalclk is of the opinionthatthese "brigand while CarolineFinkel, who undertookdetailedresearch period(see his "MilitaryandFiscal Transformation"), on the administration the Ottomancampaignsin Hungaryat the most crucial moment of these rebellions, of sekbans recruitedfrom Anatolia for these campaignswas quite points out that the numberof "tiifenkendaz" of insignificant.See C. Finkel, The Administration Warfare:The OttomanMilitary Campaignsin Hungary, 1593-1606 (Vienna, 1988), 40-46. On this point, one should also pay attentionto Barkey's comment that the Celali bands in Anatoliawere the "resultof a process of low-level militarization" that had alreadybegun in the countrysideby the late 16th century.According to Barkey,this militarizationwas the outcome of the It government'sresponseto the existing "crisisof the Ottomanpeasants." is in this context thatBarkey,having prioritizedthe importanceof economic and demographicpressure,acknowledges the effect of employment emphasis opportunitieswithin the militaryand access to arms-phenomena on which Inalclk lays particular (see Barkey,Bandits and Bureaucrats,154-55). 56Fora generalevaluationof these phenomena,see Faroqhi,"Crisisand Change,"438-47. Rise and Fall." 57See Erderand Faroqhi,"Population 58The "icmal"avartzregistersthatMcGowanused are,as the title suggests, the summaryregistersprepared usually on the basis of judicial districts (kaza), which give only the totals of the tax units for avariz levies (avartz-hanes),unitsconsisting of varyingnumbersof actualhouseholds.As is well known,unless it is openly statedin the registersit is impossible to know how many actualhouseholdsmake up an "avartz-hane." Going

202

OktayOzel

by these records,it is no more than mere speculationto make any populationestimate (on these problems, see Ozel, "Avariz Cizye Defterleri").The "mufassal" ve (detailed)avartzregisters,which arereferredto later, however, are of a differentkind of recording, on the basis of comprehensiveprovincial surveys, the actual householdsor hanes being recordedin the same way as the tahrirregisters(see n. 59). 59McGowan,Economic Life, 80-104. For the most recent work on the possible effects on the Ottoman lands of climatic changes, see W. J. Griswold, "ClimaticChange:A Possible Factorin the Social Unrest of SeventeenthCenturyAnatolia,"in Humanistand Scholar: Essays in Honor of Andreas Tietze,ed. HeathW. Isis Lowry and Donald Quataert(Istanbul: Press, 1993), 36-57. "WasThere?"58. 60Todorova, 61AlthoughI generally agree with Todorova'sreservationsabout the directand somewhat superficialcomparabilityof the demographicdevelopmentsin the OttomanEmpire in this period with the so-called 'crisis' in Europe, it does seem problematic,to think of demographicphenomenabeing completely independentof conjuncturalsocial and economic conditions. As seen in this survey,the Ottomancase of the 16th and 17th centuriesis, I think, clear proof of such complex relationships. 156-57. "WasThere?"61. See also TodorovaandTodorov,"HistoricalDemography," 62Todorova, 63See Linda Darling, "AvartzTahriri:Seventeenthand Eighteenth Century Survey Registers,"Turkish Studies Association Bulletin 10 (1986): 23-26; idem, Revenue and Legitimacy; Ozel, "17. Yiizyll Cf. Demografive IskanTarihi iqin Onemlibir Kaynak." Giike, Lzlkiyye, 12-14; OzerErgenq,OsmanliKlasik Ankarave Konya (Ankara:AnkaraEnstitiistiVakfi, 1995), Dinemi Kent TarihgiligineKatki.XVI. Yiizytlda 53. 64Cf.R. J. Jennings,"Urban A Populationin Anatoliain the SixteenthCentury: Studyof Kayseri,Karaman, and International Journalof MiddleEast Studies7 (1976): 21-57; Kiel, "Anatolia Amasya,Trabzon Erzurum," For Transplanted?" the figures in the table, see M. Ali Unal, "1056/1646 TarihliAvariz Defterine Gore 17. Belleten 51 (1987): 119-127; Ozel, "Changes,"143-54; Mehmet Oz, "XVII. Harput," Ytizyll Ortalarmnda Yiizyll OrtasinaDogru CanikSancagi,"in Prof Dr BayramKodaman'aArmagan,ed. M. Ali Unal (Samsun, iskanve Niifus (1539-1642)." On the Tokatregion,see Ali Aqikel, 1993), 193-206; idem, "BozokSancagmnda ed. "TokatOrnegindeXVII. Ytizyiln ilk Yarisinda OsmanliSosyal Yapisindaki Buhran," HasanCelal Giizel et al., Tiirkler (Ankara:YeniTtirkiyeYayinlari,2002), 10:348-58. Aqikelappearsto have carriedout research very similar to the researchI undertookfor the Amasya region. Although I have not seen his original study, Aqlkelpointsin his articleto a 67.98 percentdropin the recordedtax-payingpopulationin the Tokatcountryside duringthe same period. Because his figures are presentedin a differentform, I have not includedTokatin the table. 650n Amasya, see TD 26 and TD 776. Cf. Ozel, "Changes,"136-43. For the figures for the districtsof Fall in SeventeenthCenturyAnatolia:Some Findingsfor the Canik and Bozok, see Mehmet Oz, "Population Districts of Canik and Bozok,"ArchivumOttomanicum (forthcoming).I thankMehmet Oz for allowing me to use his findings. See also Oz, "XVII. Ytizyll OrtasinaDogru Canik Sancagl";idem, "Bozok Sancaginda Iskan ve Ntifus."Aqikel finds that 16.55 percentof the villages in the Tokat district (forty-eight villages) disappearedbetween 1574 and 1942, and the numberof villages temporarilyabandonedduringthe period was much higher:Aqikel, "TokatOrneginde,"150-51. On the question of lost settlementsduringthis period, see Suraiya Faroqhi,"Anadolu'nuniskani ve Terkedilmil Kdyler Sorunu,"in Tiirkiye'deToplumsalBilim (Ankara,1976), 289-302. See also inalclk, "Mezra'a" EI2;idem,An Arattrmalarinda Yaklalm ve Yiintemler Economicand Social History, 165-66. 138-39. Similarchanges areobservedin otherpartsof Anatolia.See ErderandFaroqhi, 66Ozel,"Changes," Sorunu." Rise and Fall,"332; Koq, "XVI. Yiizyllin Ikinci YarisindaKoylerinParqalanmasi "Population 67MehmetOz of Hacettepe University and I have been working on the extant detailed avariz registersin the OttomanArchives and hope to publish in the near futurethe results regardingthese developments.For the time being, see Ozel, "Changes"; idem, "17. Ytizyll Demografive IskanTarihilqin Onemli bir Kaynak"; Oz, "XVII. Yiizyll OrtasinaDogru Canik Sancagl";idem, "Bozok Sancagindaiskan ve Niifus". Cf. Yunus Bir Bakanllgi, 1989). See also Koq,XVI. Yiizydlda OsmanliSancainiln iskainve Niifus Yapisi(Ankara:Kiultuir M. Ali Unal, "1056/1646 TarihliAvariz Defterine Gore 17. Ytizyll Ortalarinda Unal publishedthe Harput." Defteri,"Tarih register he used in this work. See M. Ali Unal, "1646 (1056) TarihliHarputKazisl Avrilz incelemeleri Dergisi 8 (1997): 9-73. 680n the extentof these developmentsthroughout 17thcentury,see Halil inalclk, "Adaletnameler"; the idem, "TheOttomanDecline andIts Effects upon the Reaya,"in Aspects of the Balkans:Continuity and Change,ed. H. Birnbaumand S. Vryonis (The Hague, 1972), 338-54. Cf. Ozel, "Changes," chap. 4.

Ottoman"DemographicCrisis" Reconsidered 203

69Detailedavariz registersof the mid-17th centurytend to record a not insignificantnumberof widowed women (dul havdtin)undera separatecategory (see, for example, TD 776 of Amasya). This may be seen as a social manifestationof the increasingdeathratein the male populationas well as of the position of the women who lost their husbandsduringthe Celali period. One should, however,considerthe possibility thatthis may reflectan ordinarysituationin which widowed women, too, were carefully recordedon the basis of property thatmade them liable to avariz taxes as the heads of households. 70A limited numberof works on this period point to a situationthat was no better than that in the early 17thcentury:see G6kGe, the 44, 46. Furthermore, populationof the city of Lazlklyye (Denizli), for L_zlklyye, example, decreasedsignificantly(36.22%) between 1678 and 1699: ibid., 95-96. 71Theextent of the Celali movementsand the ensuing destructionin Anatoliaappearstill to be a matterof doubtamong some historianson the assumptionthat this was a point highly exaggeratedin the contemporary nasihatnameliterature.This somewhat ad hoc claim may say something about the natureof this particular source. However,it is clear that it does not take into account the ample evidence that many other sources of since Akdag clearly attest. diversenaturepresentand that the increasingnumberof case studies undertaken We are now in a far strongerposition than Akdag was decades ago in evaluating,and even quantifyingto a certaindegree, the extent of ruraldestructionthat the Anatolian countrysideunderwentduring this period. Faroqhi,who has spent perhapsmore time and effort than anyone else on the natureof the Celali rebellions, points out in a recent articlethe apparent neglect of historiansof the extent of ruraldestructioncaused by the Celalis: see, Faroqhi,"SeekingWisdom in China,"111-12. 720n famine--not an infrequentoccurrencein Anatolia during this period-see Ltitfi Giiqer,XVI-XVII. Asirlarda Osmanliimparatorlugu'ndaHububatMeselesi ve HububattanAlinan Vergiler(Istanbul:Istanbul UniversitesiIktisatFaktiltesiYayln, 1964). For an example of the negativeeffects of famine and othernatural disasterson agricultural in the Manisadistrictin the mid-16thcentury,see Emecen,Manisa, 243-44. For a life Anatolia,see R. J. Jennings,"Plaguein TrabzonandReactionsto studyof some cases of plaguein 17th-century It accordingto Local JudicialRegisters," Humanistand Scholar,27-35. On laterperiods,see Daniel Panzac, in Osmanliimparatorlugu'ndaVeba(1700-1850), trans.SerapYilmaz (Istanbul: TarihVakfiYurtYaymi, 1997). The destructiveeffects of droughtand famine on crop patternsand daily lives of peasants in the 16th- and OttomanEmpireare often referredto in local studies. For an example, see Yllmaz, "Edremit," 17th-century in 88 ff. For a detailed chronologicalhistory of earthquakes the OttomanEmpire, see N. N. Ambraseysand C. F. Finkel, The Seismicityof Turkey AdjacentAreas: A Historical Review, 1500-1800 (Istanbul:Eren and Yaylm, 1995). On the dramaticconsequences of famine and epidemics in the same periods in Europe, see Cipolla,Before the IndustrialRevolution,128-34. a 73For lengthierdiscussionof this point, see my "Changes," 203-204; idem, "17. Yiizyll Demografive Iskan TarihiIqin Onemli Bir Kaynak",741-42; idem, "Banditry, State and Economy: On the Financial Impact of the Celali Movementin OttomanAnatolia,"paperpresentedat the Ninth International Congressof Economic and Social History of Turkey,Dubrovnik,Croatia,20-23 August 2001 (in press). 30. 74Inalclk,"Introduction," 75Cipolla,EconomicHistory of WorldPopulation,82, 88-91, 97. 76Cipolla'sargumentmay sound mechanistic and old-fashioned,remindingus of a certain philosophical of understanding socio-historicalevents, but as Todorovahas rightly emphasized (sometimes contradicting herselfin her own argument), movementshave theirown workingmechanisms,which have close demographic andcomplex relationshipswith social andeconomic phenomena.I will discuss this point furtherin the context of neo-Malthusianapproaches. 77See EmmanuelLe Roy Ladurie,"Peasants," The CambridgeModernHistory,XIII CompanionVolume, in ed. PeterBurke (Cambridge: CambridgeUniversityPress, 1979); idem, ThePeasants ofLanguedoc (Urbana, 159. Ill., 1974). Cf. Inalclk, "Settlements," 78Forthe later editions of these works, see T. R. Malthus,An Essay on the Principle of Population,ed. D. Winch (Cambridge:CambridgeUniversity Press, 1992); David Ricardo, The Worksand Correspondenceof David Ricardo,ed. P. Sraffa(Cambridge: CambridgeUniversityPress, 1951-73). 79Fora selection from their works, see M. M. Postan, "Some AgrarianEvidence of Declining Population in the Later Middle Ages," Economic History Review, 2nd series 2 (1949-50), repr.in Essays on Medieval CambridgeUniversityPress, 1973), Agricultureand GeneralProblemsof the MedievalEconomy(Cambridge: in EconomicHistoryofEurope, 186-213; idem, "Medieval AgrarianSociety in its Prime:England," Cambridge immobile," vol. 1, 1966;EmmanuelLe Roy Ladurie,"L'histoire Annales 29 (1974); see also idem, Les Paysans du Languedoc,2 vols (Paris:S.E.V.P.E.N.,1966).

204

OktayOzel

80RobertBrenner,"Agrarian Class Structureand Economic Developmentin Pre-Industrial Europe,"Past and Present 70 (1976): 30-75; idem, "The AgrarianRoots of EuropeanCapitalism,"Past and Present 97 in (1987): 16-113, both reprinted TheBrennerDebate: Agrarian Class Structureand EconomicDevelopment in Pre-Industrial Europe,ed. T. H. Aston andC. H. E. Philpin(Cambridge: CambridgeUniversityPress, 1985). 81SeePeterB. Evanset al., ed., Bringing theState BackIn (Cambridge: CambridgeUniversityPress, 1985). 82See Jack A. Goldstone, Revolutionand Rebellion in the Early Modem World(Berkeley:Universityof CaliforniaPress, 1991). See also his earlierwork,"Eastand West in the SeventeenthCentury:Political Crises in StuartEngland,OttomanTurkey,and Ming China,"Comparative Studiesin Society and History30 (1988): 103-41. 83Goldstone,Revolution28-29. 84Ester Boserupdealt with populationpressurein agrariansocieties andproducedher majorworkin the late 1960s. According to Boserup, these societies were not technologically stagnant;therefore,relative scarcity of arable land did not necessarily imply serious limitations on agriculturalproduction.She concluded that economies of pre-modernperiods. See populationgrowth did not have dramaticeffects on the agricultural Ester Boserup, The Conditionsof AgriculturalGrowth:The Economics of Agrarian Change and Population Pressure (Chicago:Aldine Publishing, 1965). 85Goldstone, Revolution,xxiv. 86Ibid.,496. 87Goldstonealso drawsattentionto a close connectionbetween price increases and populationgrowthand effects of both on groups with fixed salaries-the timar holders, in the points to similar disproportionate Ottomancase. 88Ibid.,32-37, 73. "TheAgrarian in Roots of European 89Brenner, Capitalism," TheBrennerDebate, 224. I thinkthatBrenner's point makes no essential change in the neo-Malthusiantheory of the two-phase demographiccycle. His emphasis on human-madecatastrophe,which in fact was the case in OttomanAnatolia in the early 17th century,may at best only extend the second phase of the Malthusianadjustmentprocess. in 90GuyBois, "Againstthe Neo-MalthusianOrthodoxy," TheBrennerDebate, 117. 91Itmay be illuminatingat this point to considerCipolla's commenton populationgrowthin medievaland early modernsocieties in Europe.He points out thateven relativelylow growthrates over a long periodmight result in explosive situations,thus showing the vulnerabilityof peasant societies to changes in population. He also arguesthat, even at the beginningof the 14th centurybefore the Black Death, many areasof Europe were alreadyoverpopulated relationto prevailinglevels of productionand technology: see Cipolla, Before in the IndustrialRevolution, 130. This is a good example of differentlevels at which one can identify a case of underdiverse circumstances.One should note that neitherGoldstone nor Cipolla sees pre"overpopulation" industrialsocieties as stagnantand inflexible in responseto changingconditions.Whattheyrightlyemphasize insteadis thatthe choices thatsuch societies hadwere seriously limited by the availabilityof natural resources and technology. 92Forsome similar developmentsin the Balkans,see esp. McGowan,Economic Life in OttomanEurope. See also FikretAdanir,"Tradition and RuralChange in SoutheasternEuropeduring OttomanRule,"in The Origins of Backwardnessin Eastern Europe: Economics and Politics from the Middle Ages until the Early TwentiethCentury,ed. Daniel Chirot (Berkeley:University of CaliforniaPress, 1989), 131-76; idem, "The OttomanPeasantries,c. 1360-c. 1860,"in The Peasantries of Europefrom the Fourteenthto the Eighteenth Centuries,ed. Tom Scott (London:Longman, 1998), 269-310. who "PoliticalTensions,"esp. 127-30. Cf. Singer, "PeasantMigration," 62-63. Wolf Hiitteroth, 93Faroqhi, has carriedout researchin the historical geographyof the region, points out that the steady increase in the numberof tax-payersrecordedin the 16thcenturyin northernSyria and southeastern Anatoliamay not in fact indicate real growth in populationbecause of the incomplete natureof the earliest register.He suggests that such an increasemay at best show the expansionof ruralsettlementsinto the hithertouncultivated arablelands: Wolf Htitteroth,"High PopulationIncrease in the 16th Century?"unpublishedpaperpresentedat the Ninth International Congress of Economic and Social History of Turkey,Dubrovnik,20-23 August 2001. I agree with Hiitteroththat the apparentincompletenessof most of the earliest registersof differentregions should be taken into account when interpretingthe figures given in these registers. But this still does not account for the high level of increase both in the recordedtax-payersand, to a lesser extent, in the numberof village settlements observed in the later surveys. I argue that the demographicchanges that took place especially duringthe mid- to late 16thcenturycannotsimply be explainedby the expansionof settlementswithouttaking

Ottoman "DemographicCrisis" Reconsidered 205

into considerationthe size anddensity of arableland as well as the socio-economic compositionof the society. I have alreadypointed out that such an expansion may well be seen as yet anotherindicationof demographic Anatolia. pressurein the north-central Bandits and Bureaucrats,148. 94Barkey, 95Asan example,see SuraiyaFaroqhi,"AGreatFoundationin Difficulties:Or Some evidence on Economic Contractionin the Ottoman Empire of the Mid-Seventeen Century," Revue D'Histoire Magrebine 47-48 Crisis and the Art of Flute-Playing:The WordlyAffairs of the Mevlevi (1987): 109-21; idem, "Agricultural Dervishes (1595-1652)," Turcica20 (1988): 43-70. See also KayhanOrbay,"The FinancialAdministration of an Imperial Waqf in an Age of Crisis: A Case Study of Bayezid II's Waqf in Amasya (1594-1657)," (unpublishedM.A. thesis, Departmentof History,Bilkent University,Ankara,2001). Cf. Ergenq,Ankara,54. 960n the centralimportance mortalityin demographicchanges,see Goldstone,Revolution,27-29. On the of factorsthatdistinguish"catastrophic" mortalityfrom "normal" mortality,such as war,famines,andepidemics, see Cipolla, Before the IndustrialRevolution,3-5, 127 ff. As the readermay have noticed, I have occasionally touched on the importanceof the change in fertility and mortalityrates in drasticpopulationmovements in generalbut have not elaboratedthe point in the context of Anatoliaduringthe period underexamination.One may justifiablythink thatthe dramaticevents of this period, such as long and exhaustingwars, destructionof ruralstructure,and negativechanges in climate, could have resultedin such a change in fertilityand mortality rates, with mortalitysignificantlyovertakingthe fertility.Under such conditions, which may be comparable to those of the HundredYears'Warand the ThirtyYears' Warin Europe,one might expect a developmentin Anatolia similarto thatdescribedby Cipolla: "a peak of catastrophic mortalitywould cancel out the previous demographicgains andthe cycle would startall over again.In this way the frequencyand severityof the peaks of catastrophicmortalitydeterminedthe populationtrend."(ibid., 132-33).