The plague of Justinian and other scourges : an analysis of the anomalies in the development of the iron age population in Finland

(1)

The plague of Justinian and other scourges : an analysis of the anomalies in the development of the iron age population in Finland

Seger, Tapio

Fornvännen 77, 184-197

http://kulturarvsdata.se/raa/fornvannen/html/1982_184

Ingår i: samla.raa.se

(2)

Scourges

An analysis of the Anomalies in the Development of the Iron Age population in Finland

By Tapio Seger

Seger, T. 1982. T h e Plague of Justinian and Olher Scourges. An Analysis of the Anomalies in the Development of the Iron Age Population in Finland. (Den justinianska pesten och andra hcmsökelscr. En analys av anomalierna i befolk- ningstillväxt under järnåldern i Finland.) Fornvännen 77, Stockholm.

In this paper the corpus of excavated and dated Iron Age burial grounds in Finland is quantitatively analyzed with various statislical methods in order to isolate and define the anomalies of burial ground development which probably reflect actual disturbanccs in population growth. Various potential explanations thereof are discussed, and epidemics recorded in historical sources from other parts of Europé — especially the great pandemic of bubonic plague known from the reigns of Justinian I and his surcessors — suggested as probable causes of some of the anomalies.

Tapio Seger, Department of Archaeology, University of Helsinki, P.O. Box 913, SF-OOIOI, Helsinki 10, Finland.

Mors a n a avidos oris hiatus pandit et omnes explicat ålas;

quique capaci turbida cumba flumina servat durus senio navita crudo, vix assiduo bracchia conto lassata refert, fessus turbana vectare novam.

Scncca, Oidipus.

T h e present p a p e r is partly based on an earlier study, where the corpus of excavated and dated Iron Age burial grounds in Fin- land was quantitatively analyzed (Seger

1982). Burial grounds were choscn as indicators of Iron Age population development, because only very few dwelling sites of the period are so far known in Finland ( M e i n a n -

Fornvännen 77 (1982)

der 1980 pp. 7 — 8 ) . T h e material, its limita- tions and results of the statistical calculations were presented in detail in the previous study.

It suffices here to state that the numbers of burial grounds are not absolute, but are highlv representative of the known total. T h e most severc limitation is the fact that the different burial grounds of various types and

(3)

The plague of Justinian 185

periods are not directly comparable quantitatively, and to create a system of weighting them is all but impossible due to tbc vaguc character of the cremation burial grounds, the restrictcd scope of most excavations and the consequent difficulties in cstablishing or even cstirnating the numbers of graves (cf. Meinan- der 1977 pp. 35—36). However, these limita- tions are taken into account in the discussion, and no attempt is made to cstimate absolute population size. T h e main aim of the present paper is to locate the most obvious anomalies by period and province in the quantitative development of numbers of burial grounds, to make them comparable to some degree and to suggest probable explanations for them.

By and large the methods used here rcsemble those applied by Renfrew (1972, Chapter 14) to the Neolithie and Bronze Age dwelling sites in the Acgean in order to detect various patterns of population growth.

T h e convcntional period division of the Iron Age in Finland is followed here for convenience (for discussion on various period transitions see Meinander 1969 p. 66; Sarvas 1972 pp. 4 9 — 5 3 ; Cleve 1978 pp. 194—195, 239; Seger 1979); the Pre-Roman Iron Age is omitted from the analysis because of the lack of burial grounds due to burial customs at that time (cf. Meinander 1969):

Pre-Roman Iron Age Early Roman Period Late Roman Period Migration Period Merovingian Period Viking Period Period of the Crusade

T h e spatial division (Fig. 1) follows the old provincial {maakunta) division which seems to be somewhat better c o r r d a t e d with the Iron Age settlement pattern as rcflectcd in burial grounds (cf. Kivikoski 1959) than the modern division into provinces (lääni).

Åland is omitted here, as being best described as an extension of Central Sweden in the Iron Age (Meinander 1980 p. 7 ) , as is eastern Finland, where burial grounds only began to cmergc in the Viking Period.

Fig. 1 The spatial division of the area under study, following the old division into provinces {maakunta). FP = Finland Proper, S = Satakunta, T = Tavastia, O = Ostrobothnia, N = Nyland. — Den geografiska fördelningen enligt de gamla provinserna (maakunta). FP = Egentliga Finland, S = Satakunta, T = Tavastland, O = Österbotten, N = Nyland.

c. 500 B.C. — 0) c. 0 — 200 A.D.

c. 200 A.D.— 400 A.D.

c. 400 A.D.— 550 A.D.

c. 550 A.D.— 800 A.D.

c. 800 A.D.—1050 A.D.

c. 1050 A.D.—1150 A.D./

300 A.D. in eastern Finland.

The Analysis

First, the number of burial grounds by- period in the whole country (excepting the omissions mentioned above), the percentages by period of the total number and the coefficients of growth between periods are

presented in Table 1.

T h e number of burial grounds increases stceply during the earlier Iron Age. After the Merovingian Period growth is retarded, and there is a sharp d e d i n e from tbc Viking Age to the Period of the Grusades. T h e peak of growth consequently clearly falls in the Merovingian Period.

Second, the periodical distribution of the numbers of burial grounds by province (percentages according to period of the total

(4)

Table 1. Numbers of burial grounds in use by period; percentages of the total number according to period and coefficients of growth between periods in the whole country. — Antalet gravfält i använd- ning per period, procenttalen per period av det totala antalet och tillväxtkocfficienter mellan perioder i hela landet.

Total number of burial grounds = 323

Burial grounds Coefficients Period in use % of total of growth

Early Roman Late Roman Migration Merovingian Viking Crusade

26 55 119 151 133 47

8.0 17.0 36.8 46.8 41.2 14.6

2.1 2.2 1.3 0.9 0.4

Table 2. Coefficients of growth in the numbers of burial grounds according to province from the Early Roman to the Merovingian Period and from the Merovingian to the Period of the Grusades. — Till- växtkoefficienter av antalet gravfält per provins från den äldre romerska järnåldern till merovingertiden och från merovingertiden till korstågstiden.

Province Finland Pftjper Satakunta Tavastia Ostrobothnia Nyland

Eearly Roman —* Merovingian 3.5

13.3 22.0 5.5 2.3

Merovin gian —* Crusade 0.4

0.3 0.6

—.

—

Table 3. T h e provinces arranged in rang order by period according to density of burial grounds (burial grounds/1000 km2, the numbers of burial grounds by period being compared to the total area of parishes with known Iron Age burial grounds in a province). — Provinserna arrangerade i rangordning per period enligt gravfältsfrekvensen (antalet gravfält/1000 km², då antalet gravfält jämförs med den totala area- len av socknar med kända järnäldersgravfält i en provins).

Period ER

LR

Mi Me

V C

1 Finland Proper Finland Proper Finland Proper Finland Proper Finland Proper Finland Proper

2 Nyland

Nyland Ostro- bothnia Sata- kunta Sata- kunta Sata- kunta

3 Ostro- bothnia Ostro- bothnia Sata- kunta Ostro- bothnia Tavastia Tavastia

4 Sata- kunta Sata- kunta Nyland

Nyland

— •

5 Tavastia

Tavastia

Tavastia Tavastia

Ostro- bothnia

—

^ k ^ k ^ m m

(5)

The plague of Justinian 187

Fig. 2. The periodical distribution of the burial grounds according to province (percentages by period of the total number in a province). ER = Early Roman Period, LR = Late Roman Period, Mi = Migration Period, Me = Merovingian Period, V = Viking Period, C = Crusade Period. The scale on the y axis is logarithmic. — Fördelningen per period av antalet gravfält i de olika provinserna (procenttalen per period av det totala antalet i en provins). ER = äldre romersk järnälder, LR = yngre romersk järnålder, Mi = folkvandringstid, Me - merovingertid, V = vikingatid, C • kortågstid.

Y-axcln har logaritmisk skala.

number of burial grounds in a province) is presented as graphic curves in Fig. 2.

T h e growth is steady but fairly slow in the province of Finland Proper (Egentliga Fin- land, Varsinais-Suomi), with a rather high percentage of burial grounds already in the Roman Iron Age and the eulmination in the Viking Period. In the other provinces the growth is much stronger, starting with low percentages in the Early Roman Period and reaching its peak in the Migration Period in Ostrobothnia (Österbotten, P o h j a n m a a ) , in the Merovingian Period in Satakunta and in

the Viking Period in Tavastia (Tavastland, H ä m e ) . T h e development in Nyland (Uusi- maa) is clearly anomalous with the peak already in the Late Roman Period, a sharp temporary decrease in the Migration Period and a further increase in Merovingian times.

Judging by evidence from lhe burial grounds, settlement in Ostrobothnia and Nyland seems to have ceased in the course of the Viking Period.

Third, the development according to province from the Early Roman to the Mero- vingian Period on the one hand, and from the Merovingian to the Period of the Grusa- des on the other, is presented graphically in Fig. 3. T h e method is liable clearly to show both the long-term trends and the relative order of the provinces in the beginning, the middle and the end of tbc period under study. T h e corresponding coefficients of growth are collectcd in Table 2.

Furthcrmore these methods bring out lhe strong increase in the number of burial grounds in Tavastia and Satakunta during the earlier Iron Age, while in Finland Proper the growth is much slowcr. T h e absolute number of burial grounds in the Mero- vingian Period is highest in Satakunta, while Finland Proper comes a close second, the latter being the first in the beginning and at the end of the period under study. T h e development in Ostrobothnia and Nyland is retarded by the anomalous decrease from the Migration to the Merovingian Period, and tbc sharp temporary d e d i n e in the Migration Period, respectively. T h e growth is negative in all provinces between tbc Merovingian and the Period of the Grusades. T h e decrease is smallest in Tavastia, because of the stronger growth from the Merovingian to the Viking Period.

Fourth, lhe provinces are arranged in chronological order according to burial ground density in Table 3. This density here ineans the ratio burial grounds/1000 km'-*, when the number of burial grounds by period is compared with the total area of the parishes in the province with known Iron Age burial grounds.

(6)

burial grounds burial grounds burial grounds continuity

In) In) (n)

new burial grounds %

90 BO 70 60 50 40

10 9 8 7 6 5

Fig. 3. The development of the numbers of burial grounds according to province from the Early Roman to the Merovingian Period, and from the Merovingian to the Period of the Grusades. The scale on the y axis is logarithmic. — Utvecklingen av antalet gravfält per provins från den äldre romerska järnåldern till merovingertiden och frän merovingcr- till korstågstiden. Y-axeln har logaritmisk skala.

90 60

i S ^ f i O P 62.2 ' . i . t .

20.6

i u E R - » L R - » M I - » M l -»V - > C M

30

20 25.0 55.6

44 0 62.3

21.3 E R - » L R - » M I - » M i - » V - » C 10

9 8 7 6 5

100.0 75.0

U9 45.5 18.8 , ER-» LR-»MI-» Me-» V-» C

3

2

75.0 „ , 42.5

t . 5 n.Q E R - » L R - » M I - » M l - » V - » C 1

r

5

^

60.0 3JU.

0.0

FP

S

T

0

N

64.0 "»•3 6?.2

41,7 <."f

LR Mi M i V C

88.9

29.8 37.5

LR Mi M l V C

75.0 76.9 68.2 68.8 53.8

LR Ml M l V C

72.7 82.5 22.7

—1 0.0 0.0 LR Ml M l V C

90.0 80.0 166.7

40.0 0.0 E R - » L R - » M i - » M i - » V — » C LR Mi M l V C

Fig. 4. Percentages of continuity of burial grounds between periods and percentages of new burial grounds by period, according to province, presented as histograms. — Procenttalen av gravfälts- kontinuitet mellan perioder och procenttalen av- nya gravfält per period, enligt provins, framställda i form av staplar.

Finland Proper rctains tbc leading posi- tion throughout the period in question. T h e province of Satakunta steadily asccnds in the order of prcccdencc in the course of time, and the descent of Ostrobothnia after the Migration Period is evident, but the development in the other provinces is not as clear in the light of this method.

Fifth, the principal figures of burial ground continuity (ratio of burial grounds in use in bolh of two succecding periods to those in use in the former period) and the percentage of new burial grounds in the latter period, according to province, are presented as histograms in Fig. 4.

T h e most striking common feature of the histograms is the low continuity from the Fornvännen 77 (1982)

Viking to the Period of tbc Grusades. T h e burial ground continuity is exceptionally low also in Satakunta and Nyland from the Early to the Late Roman Period, in Nyland from the Late Roman to the Migration Period and in Ostrobothnia from the Merovingian to the Viking Period. T h e comparativdy low continuity from the Migration to the Mero- vingian Period, pcrceptible above all in Fin- land Proper (a province of generally high continuity), in Satakunta and in Ostro- bothnia is another prominent feature. T h e percentages of new burial grounds are higher than average in both Tavastia throughout the Iron Age and in Satakunta, Ostrobothnia and Nyland in the earlier Iron Age.

Sixth, the figures of burial ground conti-

(7)

T h e plague of Justinian 189

Table 4. Ratios by province obtained by dividing the actual number ot burial grounds in use in both of two succeeding periods by the figure of expected random continuity (x=1.7). — Relationstalen per provins erhålls genom division av det verkliga antalet gravfält i användning i båda av tvä successiva perioder med väntad tillfällig kontinuitet (x=1.7).

Province E R - > L R L R ^ M i Mi ^ Me Me -* V V -> C Finland Proper

Satakunta Tavastia Ostrobothnia Nyland

2.1 0.7 2.5 2.0 0.7

2.0 1.5 2.0 1.6 0.6

1.7 1.3 1.7 2.4 1.9

2.4 2.6 1.5 2.1 1.9

1.5 1.7 1.3

—

nuity are tested by comparison with the expected random continuity (the number of burial grounds to be expected to have been in use in both of two succeeding periods, when the continuity is assumed as random and all the burial grounds in use in either period are assumed to equally likely to have been in use in both the former and in the latter period).

T h e figures of expected random continuity between periods are obtained from the following formula:

N a x N b 2{Na + Nb)

where A'a = the number of burial grounds in use in the former, and Nb = the number of burial grounds those in use in the latter period; the mathematical proof of the formula is presented by Renfrew (1972 pp.

247—248, footnote 1). Meaningful here are the ratios obtained by dividing the number of burial grounds actually in use in both of two succeeding periods by the figure of expected random continuity. If the ratio is

LO, the actual continuity cquals random continuity. If the ratio > 1.0, the actual continuity exceeds random continuity and vice versa. T h e betwecn-periods ratios are collected in Table 4.

At least the ratios exceeding 2.0 can be regarded as indicators of a continuity decisi—

v d y higher than the random. Ratios < 1.0 appear only in the continuity in Satakunta and Nyland from the Early to the Late Roman Period and in Nyland from the Late Roman to the Migration Period. Uncritical interpretation of the results of the test may,

however, be easily mislcading, cf. the continuity in Ostrobothnia from the Merovingian to the Viking Period (burial grounds 22—>• 1).

T h e test is dependent on the numbers of burial grounds in both periods, and since the only Viking burial ground shows continuity from the Merovingian Period, the ratio will be surprisingly and deceptivdy high.

Discussion

O n the basis of the analysis, three main patterns of burial ground development can bc distinguished:

A. Steady but slow growth with high absolute numbers of burial grounds, high burial ground density and continuity together with rather low percentages of new burial grounds throughout the period under study.

This pattern is discemible in Finland Proper.

B. Pattern defined by strong and rapid growth, generally fairly low burial ground continuity (the high figures of continuity in Tavastia in the early Iron Age depend on the paucity of burial grounds, and cannot be considered significant) and high percentages of new burial grounds coupled with lower burial ground density than in Finland Proper.

This pattern can be seen in the provinces of Tavastia and Satakunta.

C. Pattern of clearly anomalous development with sharp decreases in the number of burial grounds during periods of more or less continous growth in the other provinces. T h e development in Ostrobothnia and Nyland belongs to this catcgory.

Strictly spcaking, the general interpretation Fornvännen 77 (1982)

(8)

of the patterns lies outside the scope of the present paper, but tbc discernment and de- finition of the relevant anomalies, call for a short general discussion.

Burial ground development cannot be considered as dircctly describing population growth. However, in statistical terms it certainly reflects the main trends, and can serve as a basis for analysis of population development, particularly if some economic and social factors are taken into account.

In the area of pattern A, i.e. Finland Proper, the settlement was evidently the densest, comparatlvdy stable and stationary throughout the Iron Age. T h e obvious advan- tages of the region were its suitability for agriculture, favourable climatc and advantageous location in terms of maritime connections. O n the other hand, it is comparativdy small (the total area of the parishes with known Iron Age burial grounds is about 3500 k m2, when the corresponding figure in e.g. Tavastia approaches 10,700 k m2) . These factors apparently resulted in rather dense settlement with limited possibilities for ex- pansion already at an early stage.

T h e settlement in the regions of pattern B, the provinces of Tavastia and Satakunta, seems to have been much more expansive, labile and mobile with virgin areas for slash- and-burn cultivation and for hunting and fishing available in abundance throughout the period under study. It has been questioned, whether the families practicing non-arable farming actually moved also the houses (and burial grounds) to new cultivation areas after an interval of perhaps a few decades, or whether only the sites of cultivation circulat- ed, while the dwellings remained stationary

(Meinander 1980 p. 12). However, periodical removal of at least many of the farms in the regions where this was possible seems to bc the most plausible explanation of the fairly low figures for burial ground continuity and the corresponding high percentages of new burial grounds in the areas of pattern B.

Social factors, e.g. a system of inheritancc dictatcd by economic circumstances and therefore different from that prevalent in the region of pattern A, and the consequent Fornvännen 77 (1982)

differences in readiness to split up farms might also be worth consideration.

T h e retarded growth in the Viking Period together with the sharp decrease in the number of burial grounds connected with low burial ground continuity from the Viking to the Period of the Grusades probably do not indicate a corresponding stagnation of population growth. T h e former development can more naturally bc seen as a result of the transition from single houscholds practicing slash-and-burn farming to village settlements and field cultivation at least in SW. Finland, a process which evidently began in the Sth century and was completed in the course of the Viking Period (Meinander 1980 pp.

12—13). T h e latter development would seem to bc due to the overall change in burial customs in the beginning of tbc Period of the Grusades; cremation burial grounds on levd ground gave way to inhumation graves, which are more difficult to detect. T h e change in burial customs probably also caused burial grounds of the new type to be removed to new sites. Additional causes may be early local influence of Christianity ( H u u r r e 1979 pp. 133, 161) together with the short dura- tion of the period in western Finland.

T h e anomalies of burial ground development which presumably indicate actual anomalies in population growth can now be distinguished:

1) T h e abandonment of the Early Roman tarand type burial grounds in the provinces of Finland Proper, Satakunta and Nyland (e.g. H u u r r e 1979, m a p X I V ) during the transition to the Late Roman Period (Salo 1968 pp. 2 0 1 - 2 0 3 ; 1970 pp. 160, 192; cf.

Hirviluoto 1968 p. 32), detectable in the low burial ground continuitics between the periods in question (there is, however, some doubt whether this occurrence rcally is relevant in the present context; it may simply bc a question of a change in buriaal customs: see below).

2) T h e temporary decrease of burial grounds in the Migration Period in the province of Nyland.

3) T h e decline of the number of burial grounds in the Merovingian Period in Ostro-

^ k ^ k m m

(9)

The plague of Justinian 191 bothnia, a development probably rcflected in

the low burial ground continuities also elsewhere between the Migration and Mero- vingian Periods, as well as the end of settlement in the province at the start of the Viking Period.

4) T h e evident stagnation of settlement in the province of Nyland during the Viking Period.

T h e problem most discussed hitherto is undoubtedly the retardation and disappearance of the settlement in Ostrobothnia (e.g.

Meinander 1950; 1952; 1977; Virrankoski 1978; Ahtcla 1981). Meinander (1977 p. 43) regards the development as a long process caused by the combined effect of unidentified economic and social factors, leading to the inability of the weakencd community to take the decisive step towards village settlements and economy at a time when it apparently proved necessary elsewhere. Virrankoski

(1978; cf. Meinander 1950 pp. 147, 158;

1977 p. 43) suggests that the inhabitants of Tavastia or U p p e r Satakunta destroyed the Ostrobothnian settlement and population as a hindrance to their emergent tax expeditions to the north. Ahtela (1981) mainly reviews the earlier theories.

T h e theory proposed by Meinander is reasonable and most probably valid, but does not, after all, explain very much. T h e sugges- tion by Virrankoski is dramatic enough but lacks concrete proof. Furthermore, it does not explain, or take into account the decrease of population before the beginning of the Viking Period. In any case, the disappcarance of the Ostrobothnian settlement is a fact, the archeological argument being supported by palynological data from the area (Tolonen et al. 1979).

T h e explanation which will be suggested in the present context owes its origin to the inspiring paper by Bo Gräslund (1973), where the bubonic plague, documented in the rcigns of Justinian I and his successors, is proposed as the cause of the stagnation of the kämpagrav settlement on Gotland at the end of the Migration Period and in the transition to the Vendel (Merovingian) Period (Stenberger 1955 pp. 1144—1154,

1161—1183; Biörnstad 1955 pp. 924—951, 963—964; Gräslund 1973 p. 281). T h e theory is accepted and applied to the approximately contemporaneous cessation of settlement in various parts of Holland and Germany (van Es 1968 pp. 559—566; J a n k u h n 1969 p. 159;

Schmid et al. 1973 pp. 97—105, 108) as well as in SW Norway (Myhre 1974; 1977) by Björn Myhre (1978 pp. 233—234, 239).

T h e epidemic known as the plague of Justinian (e.g. Seibel 1857; Zinsser 1947;

Hirst 1953; Shrcwsbury 1964; 1970; Wells 1964; Rackham 1979) is in every respect comparable to the Black Death, the great plague of the 14th century. T h e pandemie, well known from historical sources (Agathius, Bcdc, Evagrius, Grcgory of Tours, Procopius of Caesarea e t c ; see Seibel 1857; Rackham 1979) from just after 540 to c. 590, swept the whole of the then known world from Far East to the British Isles (Gräslund 1973 p . 280; Rackham 1979 p . 116), with severe demographic, economic and cultural con- scquenecs (Hirst 1953 p. 11; Rackham 1979 p. 118). T h e ravaging epidemic recorded in the Irish Annals from the 660's has been interpreted as a secondary outbreak of the great pandemie ( M a c A r t h u r 1949).

T h e plaque in its pandemie form is tradi- tionally considered (e.g. Hirst 1953; Shrcws- bury 1970; Gräslund 1973; Rackham 1979) attributed to the black rat, Rattus rattus

(Linnacus), tbc principal carrier and transmitter of the primary vector of the discase, the flea Xenopsylla cheopis, höst of the bactcrium Yersinia pestis (or Pastcurclla pestis; cf. Hirst 1955 pp. 106—111), although other rodents and flcas may act as carriers and give rise to isolated occurrences of the disease.

T h e main problem for Gräslund in 1973 was that the earliest proof known to him of the black rat in Europé dcrived from the late Viking Period or early Medieval times (Gräs- lund 1973 pp. 278—279 with references), although he argues convindngly that the rapid spread ovcr Europé of the domestie cat approximately at the time of the plague (Stenberger 1933 pp. 136, 139, 144; Gejvall 1955 pp. 789—790; Boessneck >< al. 1968 pp.

(10)

85—86) must be seen as indicating the presence of the black rat. However, R. rattus has latdy been shown to have been present in Central Europé (Luttschwager 1968) as well as on the British Isles (Rackham 1979) at least from the Roman Period, a fact which lends further support to the theory. I t is highly unlikely that Seandinavia and Finland would have escaped invasion by the species, a typical ship rat easily spread by the lively maritime connections of the period (cf. Gräs- lund 1973 pp. 280—281). In addition, as Gräslund (1973 pp. 279—280) remarks, the flea X. chcopis is able to spread the plague bacillus över long distanecs even without the agency of R. rattus c g . in cloth, grain and hay, the most advantageous temperature being about 10—15 C ° , i.e. conditions prevalent for most of the sailing season in the Baltic. Import of cloth and even grain to Finland in the Iron Age is probable, and hay or straw may have been used as paeking material for various commodities. Gräslund (1973 p. 281) maintains that the disease is as contagious to cattle as to people. However, this is not exactly true. Experiments carried out during modern plague epidemics have given contradictory results (Hirst 1953 pp.

150—151). Neverthdess, it seems that the various farmyard animals sometimes can be susceptible to plague, and this would naturally aggravate the economic consequences of the potential epidemic. A possibility of cattle disease at the time on Gotland was suggested by Stenberger (1955 pp. 1177—1178).

T h e potential pandemie could be expected to appear in the burial material as a decrease in the number of graves from the latter half of the 7th century, and probably the 8th as well; in the relevant Finnish archaeological literature the expression "dating to the 7th century" generally means the early phase of the Merovingian Period, c. 5 5 0 ( 5 7 5 ) — 650(675) (cf. Cleve 1943 p. 172; Meinander

1977 p. 27). This is exactly the situation discemible in the Ostrobothnian burial grounds (cf. Meinander 1977 pp. 28, 29, 30).

Another eventual consequence would be the incidence of mäss graves in the early Mero- vingian Period. T h e two remarkable Ostro- Fornvännen 77 (1982)

bothnian water deposits of h u m a n bones from at least 60 individuals at Käldamäki in Vörå (Vöyri) and Levänluhta in Storkyro (Isokyrö), dated to the early Merovingian Period (Hackman 1913; Meinander 1950 pp.

136—145; 1977 pp. 37—38; the Levänluhta find also contains a few artefacts from the middle phase of the period) are certainly worth consideration in this context. They have been interpreted as sacrifieial sites

(Hackman 1913; Meinander 1977 p. 38) or burial grounds for slaves (Meinander 1950 pp. 145, 146), but they can with justification be regarded as sites, where victims of the highly infectious plague were hastily interred.

Another possibly relevant feature is the emergence of double graves in the early Merovingian Period, four of which are known from Ostrobothnia and a few from southern Finland (Meinander 1977 pp. 36—37, cf. pp.

28, 29; 1950 p. 206, find no. 183; Salmo 1938; find no. 29; Meinander 1950 p. 178, find no. 2, p. 204, finds nos. 171, 172; Salmo

1938 pp. 51—52; the latter grave is somewhat younger than the rest). All are warrior graves, where two males in full armour are buried together. They may represent graves of a chieftain and a slave sent to protect him in the afterlife, but this is contradictcd by the fact that the outfits of both bodies are approximately as "rich". It is likely that they are graves of a father and a son, two brothers, or close friends who died at the same time, very possibly as victims of the plague.

T h e above, together with the comparativdy low burial ground continuity from the Migra- tion to the Merovingian Period in other provinces too, lends strong support to the possibility of the plague of Justinian having spread even to Finland. Moreover the evident local cessation of settlement in the early Merovingian Period in SE Finland Proper, e.g. the Salo region (Seger 1981), should probably be seen in this light. However, Ostrobothnia seems to have been most seve- r d y afflicted, for reasons best undcrstood by epidcmiologists. Neverthdess, the Migration Period settlement in the province was fairly dense and concentrated in a comparativdy restricted area, and connections between the

^ M H

(11)

province and foreign regions were lively at the time; some kind of loose political liaison between Ostrobothnia and parts of Sweden has even been suggested (Meinander 1977 p.

34). Notwithstanding, the graves in the province from the middle phase of the Mero- vingian Period are comparatively few and poor, and many parishes had become totally devoid of inhabitants (Meinander 1977 pp.

28, 34). T h e process of the weakening of the Ostrobothnian settlement outlined by Meinan- der (1977 p. 4 3 ; see above) can best be understood as a result of the great plague with its economic consequences. T h e initial shrinking of the settled area in Ostrobothnia from c. 400 (Meinander 1977 p. 42) should rather be Interpreted as an increased cohesion of the settlement pattern for economic, social and eventually political reasons on the thre- shold of economic boom in the province in the Migration Period; it should be remembcr- ed that the Ostrobothnian settlement in the Late R o m a n Period, although distributed över a fairly wide area, was very thin, the figures of burial ground density being 2.7/

1000 k m2 in the Late R o m a n and 9.8/1000 k m2 in the Migration Period (Seger 1982, Appendix I V ) .

Another problem is the disappcarance of the tarand burial grounds in southern and south-western coastal Finland during the transition from the Early to the Late Roman Period, apparent in the low percentages of burial ground continuity in Satakunta and Nyland. This has been attributed to stagnation of the fur trade in the latter half of the 2nd century, resulting from the dissolution of the Pax R o m a n a (Salo 1968 pp. 201—203;

1970 pp. 160—161, 192—193). This event may well have affected the whole of Europé, but it is difficult to bdieve that it could have been this drastic as far north as Finland. T h e traditional study of the Iron Age in Finland seems to have had an exaggerated and somc- how very modern conception of prehistoric exchange with supposed groups of professio- nal merchants, networks of rcgular trade routes över Europé, commercial colonies etc.

A reasoned criticism of this view was recently presented by Meinander (1977 pp. 33 f).

The plague of Justinian 193 T h e evident import commodities in the preserved Finnish archaeological record are mainly luxuries: status weapons, foreign finery etc. While these certainly were of social significance, the introduction of iron work- ing had made possible the local manufacture of tools and basic weapons (e.g. Nylén 1974), and it is difficult to imagine an import commodity in the Iron Age meaning life or death to the people. Salt was evidently imported, but one does not die for the lack of it. T h e desertion of the tarand burial grounds and tbc corresponding settlements for commercial reasons would only be undcr- standable, if the people buried in them were indeed traders of foreign origin.

/ / the process really rcflccts an actual disturbance of population development (an assumption subject to challengc, as other types of burial ground seem to show continuity) and not simply a change in burial customs, a more plausible explanation would be another epidemic, afflicting most scvcrcly lhe coastal regions in closcst communication with outside areas. A well-documented pesti- lence called the plague of Antoninus (or of G a l e n ) , with a high mortality råte swept the endre Roman Empire; it has been identified as typhus (Castiglioni 1947), a form of smallpox (Zinsscr 1947) or true plague (Patrick 1967), and dated to the end of the 2nd century, the exact time of the cessation of the tarand burial grounds in Finland. T h e Marcomannian Wars (166—180) carried the epidemic into Free Germany, and there to the very areas bordering northern Europé (Gräslund 1973 pp. 281—282).

T h e only point to be made at present with any certainty concerning the stcep decline in the number of burial grounds in tbc province of Nyland during the Migration Period is, that the development was not caused by the plague of Justinian. T h e decrease is clearly due to the very low burial ground continuity from the Late Roman Period, which means that the cause must coincide with the end of the Roman Iron Age, but for the time being cannot be identified.

T h e last of the defincd anomalies is the end of the settlement in westernmost Nyland in

(12)

the course of the Viking Period. There is some doubt whether the disappearance was factual, as pollen analytical data from the area demonstratcs continuous agricultural activity from the late Iron Age to Medieval times (Tolonen ct al. 1979 pp. 16, 22, 58—59). This is usually interpreted as indicating long-distance slash-and-burn cultivation practiced by the Tavastians or the Esto- nians (Tolonen ct al. 1979 pp. 58—59;

Meinander 1980 p. 12), but there remains tbc possibility of continuous local habitation, for one reason or another out of the rcach of archaeological observation (Kerkkonen 1945, 152; Honkanen 1981; cf. Gardberg 1970 pp. 28—29).

O n lhe other hand, the cessation has generally been considered real, and potential cxplanations ought to be discussed. O n e of the routes of east-bound Vikings evidently followed the coast of Nyland, and attacks by them have been suggested as the reason (af Hällström 1948 pp. 87—88). However, it is difficult to understand what would have been the gain for Vikings on their way to rich Miklagaard in destroying a rather poor settlement of small farmers so close to their point of departurc, somewhat hazardous but hardly profitable entcrprise. A possibility could be raids by groups of Vikings returning empty- handed from the east, to capture slaves in order to have something to bring home. I t is cqually plausiblc, however, that the cause was another epidemic, one of the many reportedly (Birkdand 1954; Skyum-Nidsen 1967 pp. 3 4 — 4 1 ; Gräslund 1973 p p . 282—

284) brought back by the Vikings from their travels.

Concluding Remarks

T h e evident anomalies of Iron Age population development in Finland as reflected in the burial grounds have been discussed, and epidemics known from historical sources are suggested as explanation for many of them. T h e aim is not to maintain that disease was the sole or even the main reason for every disturbance of population growth, but it represents a very important potential

factor capablc of wide devastation, and hitherto totally neglccted in the study of the Iron Age in Finland. Very few of the diseases caused by bacteria or viruses leave any trace on h u m a n skeletons (Hare 1967 p. 115), and consequently cannot be detected directly in the osteological material, but, in any case, the explanatory potential of historically firmly documentcd and dated epidemics must be considered far greater than that of purely hypothetical wars or economic catastrophes.

As Gräslund (1973 p. 280) says: . . . " R a t t u s rattus och Xenopsylla chcopis kan knappast tilltros fåfängan att ha förlagt sin verksam- het enbart till länder där skrivkunnigt folk var tillstädes" . . . ("It is härd to believc that Rattus rattus and Xenopsylla chcopis were vain enough to restrict their activity only to countries where people capable of writing

were at h a n d " ) , and the same naturally holds good of any other transmitter of tbc various pcstilcnces, particularly in the light of the fact t h a t . . . "an infectious disease such as influenza can break out in pandemie pro- portions in the present century, circle the world and bc reputed to have killed as marn- as one hundred million h u m a n beings, reaching into such isolated populations as the Eskimos of Greenland and L a b r a d o r " (Rack- h a m 1979 pp. 118—119; cf. Elton 1958).

References

Ahtela, E. 1981. Piirleitä Vähänkyrön rautukauti- sesta asutuksesta. Pro gradu-työ arkeologiassa.

(Abstract: Iron Age Settlement in Vähäkyrö, Southern Ostrobothnia, Finland. Graduation paper.). Helsingin yliopiston arkeologian laitos. Moniste n:o 25. Helsinki.

birkeland, H. 1954. Nordens historie i middel- alderen etter arabiske kilder. Oslo.

Biörnslad, A. 1955. Previous investigation of iron age building remains on Gotland. In Sten- berger, M., Vallhagar. A migration period settlement on Gotland/Sweden. In collaboration wilh O. Klindt-Jensen. Copenhagen.

Boessneck, J., v. den Driesch-Karpf, A. & Gejvall, N.-G. 1968. The archaeology of Skedemosse III. Die Knochcnfunde von Säugetiercn und von Menschen. Stockholm.

_ _ g _ _ _ |

(13)

T h e p l a g u e of J u s t i n i a n 195 Castiglioni, A. 1947. A History of Medicine. Lon-

don.

Cleve, N. 1943. Skelettgravfälten på Kjuloholm i Kjulo. I. Den yngre folkvandringstiden. Finska fornminnesföreningens tidskrift X L I V : 1. Hel- singfors.

— 1978. Skelettgravfälten på Kjuloholm i Kjulo.

I I . Vikingatid och korstågstid. Gravfältct C.

Finska fornminnesföreningens tidskrift X L I V , 2. Helsingfors.

Elton, C. S. 1958. The ccology of invasions by animals and plants. London.

Gardberg, J. 1970. Karjaa ennen ja nyt. Ekenäs.

Gejvall, N.-G. 1955. T h e animal remains from Vallhagar. In Stenberger, M., Vallhagar. A migration period settlement on Gotland/Swe- den. In collaboration with O. Klindt-Jensen.

Copenhagen.

Gräslund, B. 1973. Aiing, näring, pest och salt.

(Summary: Farming, nutrition and disease.

O n some negative effeets of the introduction of farming). Tor XV.

Hackman, A. 1913. Ein Opferfund der Völker- wanderungszeit in Finnland. In Opuscula archaeologica Oscari Montelio septuagenario dcdicala. Stockholm.

Hare, R. 1967. T h e Antiquity of Discases Caused by Bactcria and Viruscs, A Review of the Problem from a Bacteriologist's Point of View.

In Brothwell, D. & Sandison, A.D. (eds.) Diseases in Antiquity. A Survey of the Diseases, Injuries and Surgcry of Early Populations.

Springfield, 111.

Hirst, L. F. 1953. The conquest oj plague. A sludy of the evolution of epidcmiology. Oxford.

Hirviluoto, A.-L. 1968. Friiheisenzeitliches Gräbcr- feld auf dem Pikku Linnamäki von Porvoo, Suomen Museo.

Honkanen, P. 1981. Uiidenmaun kansainvacllus- ja merovingiaika. Pro gradu-työ arkeologiassa.

(Abstrart: T h e Migration and Merovingian Periods in the Provinre of Uusimaa/Nyland.

Graduation paper.). Hclsingin yliopiston arkcologian laitos. Moniste n:o 26. Helsinki.

Huurre, M. 1979. 9000 vuottn Suomen csihistoriaii.

Kcuruu.

af Hällström, O. 1948. Karis socken från forntiden till våra dagar. I. Forntiden. Ekcniis,

Jankuhn, H. 1969. Vor- und Fruhgeschichte vom Ncolithikum bis t a t Völkerwanderungszeit.

Dcutsclie A gr ar geschichte I.

Kerkkonen, G. 1945. Västnyländsk kustbebyggelse under medeltiden. Skrifter utgivna av Svenska Litteratursällskapet i Finland 301. Helsingfors.

— 1952. Karis socken från forntiden till våra dagar. I I I . Medeltiden. Ekenäs.

Kivikoski, E. 1959. RauUikauden asutus. Population in the Iron Age. In Jutikkala, E. (cd.) Suomen historien kartasto. Atlas of Finnish History.

Porvoo — Helsinki.

Luttschwagcr, J. 1968. Hamslcr- und Hausratten-

fundc im Mauci werk eines römischen Brunnens in Ladenburg, Landkrcis Mannhcim. Säugc- ticrkdl. Mitteilungen X V I .

Meinander, C. F. 1950. Etclä-Pohjanmann historia I. Esihistoria. Helsinki.

- - 1952. Myntfyndet från Lillkyro. 1. Fyndets arkeologi. In Kivikoski, E. (ed.) Corolla archaeologica in honorem C. A. Nordman.

Helsinki.

—- 1969. Dåvits. En essä om förromersk järn- ålder. Finskt Museum.

— 1977. Svenska Österbottens historia I. För- historia. Vasa.

— 1980. T h e Finnish Society during the Sth—

12th Centuries. In Fenno-ugri et slavi 1978.

T h e University of Helsinki, Department of Archaeology, Stencil no. 22. Helsinki.

Myhre, B. 1974. Reply to the Comments on Iron Age Farms in SW-Norway. Norwegian Archaeological Review 7, no. 1.

- - 1977. Jernåldersgården på Ullandhaug I. H u - set i jernalderen i Ser Norge. Arkeologisk museum i Stavanger. Skrifter.

— 1978. Agrarian Development, Settlement His- tory, and Social Organization in Southwest Norway in the Iron Age. In Kristiansen, K.

& Paludan-Muller, C. (eds.) New Directions in Scandinavian Archaeology. Studies in Scan- dinavian Prehistory and Early History. Volume I. The National Museum of Denmark. Odense.

Nylén, E. 1974. Bronze, Eisen und Gesellschaft.

Die Kunde 25.

Patrick, A. 1967. Disease in Antiquity: Ancient Greece and Rome. In Brothwell, D. & Sandi- son, A.T. (eds.) Diseases in Antiquity. A Survey of the Diseases, Injuries and Surgcry of Early Populations. Springfield, III.

Rackham, J. 1979. Rollus rattus: the introduction of the black rat into Britain. Antiquity 53.

Renfrew, C. 1972. T h e Emergence of Civilisation.

The Cyclades and the Acgcan in the Third Millenium B.C. London.

Salmo, H. 1938. Die Waffen der Merowingerzeit in Finnland. Finska fornminnesföreningens tidskrift X L I I : 1. Helsinki.

Salo, U. 1968. Die friihrömische Zeit in Finnland.

Finska fornminnesföreningens tidskrift 67.

Helsinki.

— 1970. Metallikautinen asutus Kokcmäenjoen suussa I. Muinaisjäännökset ja muinaislöydöt.

(Synthcse in deutscher fjbersetzung: Die me- tallzeitliche Besiedlung an der Miindung des Flusses Kokcmäenjoki). Pori.

Sarvas, P. 1972. Länsi-Suomen ruumishautojen raha-ajoitukset. Helsingin yliopiston arkeologian laitos. Moniste n:o 6. Helsinki.

Schmid, P., Behre, K.-E. & Zimmcrman, W. H.

1973. Die Entwickhmgsgeschichte einer Sied- lungskammer im Elbe-Weser-Dreieck seit dem Neolithikum. Nachrichlcn aus Niedersachsens Vorgeschichte 42.

(14)

Seger, T. 1979. O n the Beginning of the Viking Period. Kontaktstencil 17. Helsinki.

— 1981. Katajamäki och Ketohaka i Salo, Egent- liga Finland. Kontaktstencil 20. Helsinki.

—• 1982. Suomen rautakautinen asutuskehitys kalmistojen valossa — tilastollinen analyysi.

(Abstract: Iron Age Population Trends in Finland in the Light of Cemeteries — A Quantitative Analysis). A paper presented at an archaeological symposium 11—17 May 1981 in Leningrad, to bc published in Russian, Leningrad 1982.

Seibel, V. 1857. Die grosse Pest zur Zeit Justinians I , und zur Zeite gehenden ungewöhnlichen Natur-Erzeignisse. Dilingen.

Shrewsbury, J. F. D. 1964. The Plague of the Philistines and other medical-historical essays.

London.

— 1970. A History of bubonic plague in the British Isles. Cambridge.

Skyum-Nielsen, N. 1967. Vikingerne i Paris. Berct-

ninger frei 9. århundrede oversat og forklaret.

Köbenhavn.

Stenberger, M. 1933. Oland under äldre järnåldern.

Kungl. Vitt. hist. och antikv. akademien.

Stockholm.

— 1955. Vallhagar. A migration period settlement on Gotland/Sweden. In collaboration with O. Klindt-Jensen. Copenhagen.

Tolonen, K., Siiriäinen, A. & Hirviluoto, A.-L.

1979. Iron Age cultivation in SW Finland.

Finskt Museum 1976.

van Es, W. A. 1968. Wijster. A Native Village beyond the Imperial F ronder 150—425 AD.

Palchistoria X I , 1 & 2.

Virrankoski, P. 1978. Kainu — Pohjanmaan rautakautinen kansa. Faravid 2.

Wells, C. 1964. Bones, bodies and disease. Evi- dence of disease and abnormality in early man.

London.

Zinsser, H. 1947. Råts, lice and history. Boston.

(1st edition 1935, London).

Den justinianska pesten och andra hemsökelser. En analys av anoma- lierna i befolkningstillväxt under järnåldern i Finland

I d e n n a a r t i k e l a n a l y s e r a s d e u t g r ä v d a o c h d a t e r a d e j ä r n å l d c r s g r a v f ä l t c n i F i n l a n d k v a n - t i t a t i v t m e d o l i k a s t a t i s t i s k a m e t o d e r för a t t få f r a m o c h d e f i n i e r a d e a n o m a l i e r i d e n k v a n t i t a t i v a t i l l v ä x t e n a v a n t a l e t g r a v f ä l t p e r p e r i o d o c h p r o v i n s , s o m s a n n o l i k t å t e r - s p e g l a r v e r k l i g a r u b b n i n g a r i b e f o l k n i n g s - t i l l v ä x t e n .

T r e t y d l i g a m ö n s t e r a v t i l l v ä x t b l o t t a s g e n o m a n a l y s e n :

A . E t t m ö n s t e r m e d j ä m n m e n l å n g s a m tillväxt m e d k o m p a r a t i v t h ö g a a n t a l a b s o l u t a

g r a v f ä l t s m ä n g d e r , h ö g g r a v f ä l t s f r c k v e n s o c h - k o n t i n u i t e t s a m t g a n s k a l å g a p r o c e n t t a l av- n y a g r a v f ä l t u n d e r h e l a d e n g r a n s k a d e p e r i o - d e n . D e t t a m ö n s t e r k a n i a k t t a g a s i E g e n t l i g a F i n l a n d .

B. E t t m ö n s t e r p r ä g l a t a v s t a r k o c h s n a b b t i l l v ä x t , i regel g a n s k a l å g g r a v f ä l t s k o n t i - n u i t e t o c h h ö g a p r o c e n t t a l a v n y a g r a v f ä l t s a m t l ä g r e g r a v f ä l t s f r e k v e n s e r ä n i E g e n t l i g a F i n l a n d . D e t t a m ö n s t e r u p p t ä c k s i T a v a s t - l a n d o c h S a t a k u n t a .

C . E t t m ö n s t e r m e d k l a r t a n o m a l i s k till- v ä x t m e d t y d l i g a s ä n k n i n g a r i a n t a l e t g r a v -

Fornvänncn 77 (1982)

(15)

The plague of Justinian 197 fält under perioder av mer eller mindre kon-

tinuerlig tillväxt i andra provinser. Tillväxt- mönstren i Österbotten och Nyland hör till denna kategori.

Skillnaden mellan mönstren A och B för- f a r a s här med olikartade ekonomiska förut- sättningar dikterade av territoriella förhål- landen. Varken vikingatidens fördröjda till- växt eller den skarpa sänkningen av antalet gravfält under övergången från vikinga- till korstågstiden återspeglar sannolikt någon verklig tillbakagång av befolkningstillväxt, utan snarare förorsakas de av övergången från svedjebruk och enstaka hemman till bybebyggelse och åkerbruk åtminstone i SW Finland från 700-talet framåt, samt av den generella förändringen i begravningsbruket i början av korstågstiden, respektivt.

De anomalier, som troligtvis reflekterar verkliga rubbningar i befolkningstillväxten, är följande:

1. U p p h ö r a n d e t av " t a r a n d " gravfälten i S och SW Finland under övergången mellan den äldre och yngre romerska järnåldern, uppenbart i de låga gravfältskontinuitets- talen i Satakunta och Nyland. (Det är dock möjligt, att fenomenet helt enkelt härrör av cn förändring i begravningsbruket, då grav- fält av andra typer tycks visa kontinuitet).

2. Den tillfälliga sänkningen i gravfälts- antalet i Nyland under folkvandringstiden.

3. Sänkningen i antalet gravfält under merovingertiden i Österbotten, en utveckling som troligen kan beaktas i den låga gravfälts- kontinuiteten även annanstans, samt upp- hörandet av bebyggelsen i provinsen i början av vikingatiden.

4. Det (sannolika) upphörandet av bebyggelsen i W Nyland under vikingatiden.

Det hittills mest omdiskuterade problemet iir utan vidare tillbakagången och upphöran- det av den österbottniska bebyggelsen. U t - vecklingen förklaras här som cn följd av den justinianska böldpestcn, en i övriga Europa väldokumenterad epidemi som härjade ca 540—590. Det starkt minskade antalet gra- var i Österbotten daterbara till merovingcr- tidens andra fas, d.v.s. efter ca 650, stödjer denna förklaring. De två välkända öster-

bottniska fynden av människoben från åt- minstone 60 individer (Käldamäki i Vörå och Levänluhta i Storkyro), daterade till början av merovingertiden, tolkas här som massgravar för pestens offer. Eventuellt rele- vanta är också några dubbelgravar, som framträder i början av merovingertiden. Pes- ten med dess demografiska och ekonomiska följder ledde uppenbarligen till upphörandet av bebyggelsen i provinsen senare, då över- gången till åkerbruk och större bebyggelse- enheter tycks ha blivit nödvändig liksom i andra provinser.

Om upphörandet av tarand gravfältcn på södra och västra kusten verkligen återspeglar en motsvarande tillbakagång av bebyggelsen, kan fenomenet bäst förklaras som beroende av cn a n n a n svår epidemi med hög dödlighet som härjade det romerska riket under slutet av 2:a århundradet, identifierad antingen som tyfus, en form av smittkoppor eller möj- ligen pest, överförd till det fria Germanien genom de pågående markomanncrkrigen

(166—180).

Sänkningen i antalet gravfält i Nyland under folkvandringstiden förorsakades inte av den justinianska pesten, ty utvecklingen är klart en följd av den mycket låga gravfälts- kontinniteten från den yngre romerska järn- åldern i provinsen. Orsaken, tillsvidare oiden- tifierad, måste sålunda dateras till slutet av romartiden.

U p p h ö r a n d e t av bebyggelsen i västra Ny- land under vikingatiden har tidigare förkla- rats genom anfall av vikingar i österled, ö v e r - tygande motiv är dock inte lätta att hitta, och smittospridningen av någon av de svåra epidemier som vikingar enligt samtida skrift- liga källor medförde från sina resor, är en åtminstone lika trolig förklaring.

Epidemiska sjukdomar är en viktig aspekt, hittills helt försummad i finländska järnålder- studier. Dock h a r de, som historiskt doku- menterade och daterade, större förklarings- potential än helt hypotetiska krig eller ekonomiska kriser. Det är föga troligt, att de flera svåra och vidspridda epidemier kända från det 1 :a årtusendet e. Kr. skulle ha härjat endast i de områden varifrån det finns skrift- liga källor.