Ant0650332.pdf

8/11/2019 Ant0650332.pdf

1/16

The structuring of prehistoric landscape

MARTIN

KUNA

Some recent archaeological landscape projects

in Czechoslovakia have found a theoretical

background within the concept of community

areas. According to this concept prehistoric

populations are approached as divided into

communities. Each community is supposed

to

have shared a common territory within which

most of its activities were concentrated. The

community area consisted of several sub-areas

(e.g. habitation areas, specific production areas,

funerary areas, etc.) where activities different in

function were performed (Neustupny 1986;

1991). The theory of community areas is not,

however, limited to the identification of com-

munity areas themselves. It is rather a general

approach, based on presumption and identifi-

cation

of

patterns

or

structures underlying the

archaeological record and reflecting structured

human behaviour in the past. Using some con-

cepts of the community area theory, this paper

aims at analysing prehistoric habitation areas in

the territory of Bohemia and articulating some

general hypotheses concerning set tlement pro-

cesses and structures on various levels of

complexity.

The concept

of community areas

The concept of community area implies several

aspects in which the traditional [normative,

pre-processual) archaeological paradigm has

been challenged. Traditional archaeology

usually approached landscape as an empty

space within which some points of particular

interest, sites, are situated. Archaeologists

were usually very little concerned with the

space itself. Most of their attention was paid to

the excavation and description of sites because

only sites can provide archaeologists with the

artefactual record which was thought to be the

main

or

the only source of archaeological

knowledge. Everything beyond sites and arte-

facts [off-site activities, environmental back-

ground, settlement structures) was usually

viewed as a matter of secondary importance.

This attitude can still be found both in tradi

tional archaeological theory and in archaeologi-

cal field practice in Czechoslovakia (only sites

are e.g. protected by law; only positive arte-

factual evidence

is

registered during fieldwork

whereas various kinds of non-artefactual

or

negative evidence, although very often having

contextual significance, are mostly neglected].

111 contrast to the traditional approach, the

concept of community area makes it possible to

see landscape as a continuum of information

because any part of landscape is assumed to

have had a specific function in the past and is

therefore viewed as in principle significant for

archaeologists. Even the empirically empty

space around sites can indicate either a parti.

cular activity (which cannot be identified by our

contemporary field techniques, such as,

hunting-grounds, grazing land, etc.) or an unoc-

cupied area with a specific function or meaning

(buffer zones, etc.).

The second aspect of the community area

concept is the implicit notion that landscape

(although it is continuously inhabited and

exploited) was divided into more or less regular

spatial segments corresponding to basic

economic and social units, that i s, to prehistoric

communities. The presupposition of the seg-

mented character of the cultural landscape is

based upon theoretical generalization about the

prehistoric economy and social system.

The third underlying goal of the community

area approach is to change the diachronic evalu-

ation of prehistoric agricultural populations.

Whereas traditional archaeology understands

the evolution of past societies as basically

discontinuous and divided into separate cul-

tures endowed with their individual ethnic o r

*

A\-rryi, lri.

65

(1991):3 3 2 4 7

Ins t i t u t e of Archaeology, Mala St r an a , L c t e n s k a 4 , 118 01 Iraha 1, Czi:c:hoslovakia

8/11/2019 Ant0650332.pdf

2/16

STRUCTLJRING OF PREHISTORIC LANDSCAPE 3 3 3

social significance, the community area concept

stresses long-term continuity within prehistoric

populations and the r61e of their autochthonous

evolution. It has been predicted that many

community areas or their parts did not change

in extent and inner structure for centuries or

even millennia which, along with other argu-

ments, certainly limits the validity of many

migration theories (Neustupny 1982a).The con-

tinuity of prehistoric settlement systems

cannot, however, become apparent until a

sufficient density of finds is reached within a

part of the landscape.

The archaeological record is never repre-

sentative in reflecting human activities of

different kinds. Each particular type of archaeo-

logical and environmental context leaves at

least some of the particular human activities

almost completely unrecognized

or

unrecogni-

zable. The model of community areas brings

new possibilities for studying past economic

and social systems because it has a strong

predictive value.

It

forces us to define human

activities and their spatial correlates that have

not yet been discovered because they have not

been sought and their existence has not been

presumed. The archaeological record is

approached as comprehending various struc-

tures that can be explained in terms of struc-

tured behaviour of the living society in the past.

The archaeological record must not be treated as

a set of isolated phenomena only because of the

fact that some of the elements and links of the

once-structured entity have disappeared.

I

would like to show that looking for patterns and

defining structures can be useful at various

levels of the archaeological record. Before

detailing this, some comments on the general

characteristics of the archaeological record in

Bohemia and its explanatory value should be

added.

Landscape destruction, archaeological

resources and research

in

Bohemia

The territory of Bohemia has suffered much

damage to its archaeological record during its

recent history. Deep ploughing was practised

intensively in most of the lowland areas since

the last century and culminated in the general

removal of field boundaries during the collecti-

1;ization programme of the 1950s. This led to the

destruction of nearly all off-site activity

remains, burial mounds, enclosures and field

systems. Due

to

the fact that the distribution of

prehistoric habitation areas followed very simi-

lar locational principles to later medieval vil-

lages surviving until today, much of the

archaeological record has been destroyed by

intensive building and industrial activities

within the villages and around them during the

19th and 20th centuries. Large-scale open

mining in northwest Bohemia and the intensive

industrial activities of the 1970s and 1980s

represent a recent chapter in large-scale land-

scape destruction.

It

seems that Czech archaeo-

logy has not always been able to cope with these

problems in an appropriate and effective

manner. Some recently published analyses of

archaeological field activities in Bohemia serve

as a serious warning. The long-term trends in

accumulating new field information revealed a

gradual slow-down during the last decades,

quite in contrast with the increasing personal

and technical potential of archaeological insti-

tutions (KlapStB 1989). Elsewhere it has been

shown that the routine activity of archaeologists

in Bohemia is able to document less than 5% of

the total amount of permanently damaged and

destroyed archaeological contexts, the rest of

the potential information escaping without any

registration (Kuna 1990). The necessity

of

a

more systematic (however selective) approach

to rescue excavations and of deeper changes

within the methodology, organization and

ethics of archaeological research has been

claimed (Kuna KlapStB 1990).

The archaeological record in Bohemia is,

however, characterized by several positive

qualities. The feedback effect of the extensive

damage to archaeological sites is the high

density of information scattered within land-

scape. The preservation

of

habitation sites

(composed mostly of sunken features, i.e. pits of

variable function) is relatively good, and these

sites are usually rich in dist inctive finds. Even

the field-walking evidence alone usually en-

ables us to classify sites of nearly all post-

Mesolithic periods to phases of about 100-300

years. The precision of relative chronological

classification of excavated archaeological

assemblages from certain prehistoric periods

(e.g. Neolithic, Iron Age) is, of course, much

higher. The existing information base of

Bohemian archaeology is the result of long-term

processes that have been only partly influenced

by archaeologists themselves. In many respects,

8/11/2019 Ant0650332.pdf

3/16

3 3 4 MARTIN KUNA

however, overall utilization of the archaeologi-

cal record is very much dependent upon how

effectively existing knowledge is treated and

extended by the archaeologists. If the already

available, scattered information about indi-

vidual sites is revised and computerized it may

become a very rich source of initial hypotheses

for further theoretical research. The dense net-

work of known sites could be very successfully

completed by intensive field-walking projects

within selected regions. The methodological

convenience of air photography within the

specific conditions of intensively cultivated

(ploughed and fert ilized) soils should be tested.

The limited number of wet sites in Bohemia

makes the routine employment of palynological

and other environmental studies more compli-

cated, but this handicap could be overcome by

more environment-oriented projects in future

(cf. Neustupnji 1985; 1987; Kyncl 1967; Smri

1967).

Rapid landscape destruction in the open

mining regions has positively transformed the

theoretical and methodological orientation of

many archaeologists. The necessity of under-

taking a selective and sampling approach to all

the field projects has been admitted and much

attention has turned on total settlement struc-

tures instead of the si te-oriented approach pre-

vailing before. Intensive activities within

selected small geographical units called micro-

regions (usually identified with smaller valleys

of brooks or basins) has become the leading

strategy of fieldwork. Several projects of

complex microregional studies have been

started in the coal-mining regions of northwest

Bohemia during the 1970s and 1980s (Velimskji

1986; Smri. 1986). These projects were mainly

based

on

fast rescue excavations

of

large land-

scape areas (sometimes even stripping of tens of

hectares). Although they exceeded in extent

anything that had been done in the field of

landscape archaeology in Bohemia thus far,

their results partly suffered in two ways. Firstly,

the usually very rapid nature of excavations

often limited the possibility of finding rare or

less obtrusive components and features (shal-

low cremation burials, post-hole structures,

individual pits outside the main habitation

areas, etc. ). Secondly, the main goal of these

projects was usually seen as collecting as much

settlement data in the disappearing archaeo-

logical landscape as possible, while theoretical

questions appeared only later. Other projects

elsewhere in Bohemia were initiated during the

1980s with the aim of further developing micro-

regional studies, especially by means of field-

walking landscape surveys. Along with this an

attempt to articulate a general theory of prehis-

toric settlement processes has been made

within the community area concept (Neus-

tupny 1986).

The methodology of landscape studies in

Bohemia should be understood in the context of

specific qualities of the archaeological record

here (which is very different when compared

e.g. to Britain). Having nearly no evidence of the

features surviving above the surface (field

boundaries, roads, enclosures, etc.), archaeo-

logy in Bohemia is almost completely deprived

of the chance of considering landscape struc-

ture as a whole before investigating its parts. At

the same time, a rich set of chronologically

well-defined details can be studied. Spatial

structures within this record are much more the

result of reconstruction, combining empirical

and theoretical knowledge, than of observation

(however sophisticated). I would like to show

how such reconstruction

of

settlement struc-

tures can operate on several levels of the

archaeological record, namely on the level of

household, of habitation area, of community

area, and of settlement zone.

Identificationof prehistoric households

One of the starting-points concerning a com-

munity area is the quantification of the com-

munity population and the description of the

settlement processes on the local (site) level.

This can be done only by finding such units

of

the archaeological record that reflect concrete

and measurable units

of

the past society.

Funerary remains often provide us with this

kind of data because the number ofthe deceased

and various demographic parameters can be

measured. The information contained in habita-

tion sites is usually more complicated. The

number of occupation relics in habitation areas

depends on at least three variables: the size

of

the population group inhabiting the site area;

the number of features simultaneously used by

one group of inhabitants (family, household.

etc.); and the frequency of rebuilding

of

dwellings and other facilities.

The smallest social unit which can be traced

in the archaeological record is the household

8/11/2019 Ant0650332.pdf

4/16

STRIJCTURING

OF

PREHISTOKIC LANLISCAPE 3 3 5

The household is a dynamic category, conti-

nuous in t ime. We can, however, expect that its

basic activities were usually bound to certain

spatial limits for a certain period of time. This

supposition is what the concept of household

cluster is based on (Winter

1976).

I understand

the household cluster as a group of features

connected to one dwelling place. Each house-

hold could, of course, have produced one or

more household clusters during one generation,

depending on factors such as the life-

expectancy of houses, mobility of households,

etc. Empirical observations allow us to presume

that prehistoric household clusters were hardly

larger than 50 m in diameter.

In Central Europe household clusters were

first defined in the context of Neolithic sites

(Bogucki-Grygiell981)where the connection of

various features with the groundplans of houses

is usually obvious (cf. the concept of building

complex or Hofplatz: Pavlu 1977; Boelicke

1982). In most

of

the later prehistoric periods

the features that can be interpreted

as

houses are

extremely rare. The types of features current in

the post-Neolithic period are storage pits, pro-

duction features of different kinds, and a large

quantity of pits of unknown purpose. The

identification of household clusters on such

sites can rest only upon an observed or statis-

tically proven clustered distribution of features

and the repeated composition of such clusters as

to the different functional categories of features.

A computer program consisting of several

kinds of cluster analysis has been applied

for

the identification of household clusters on some

Bronze Age habitation sites (Turkovti Kuna

1987). In many cases clusters can, however, be

observed even without using any sophisticated

method. The interpretation

of

the clusters as

relics of temporary activities of individual

households seems very probable. In the context

of different sites and different prehistoric

periods the clusters vary in number and in the

kind of features that they consis t of. The clusters

on the Early Bronze Age (2200-1900

BC)

site at

Cakovice consisted e.g. of only one storage pit

and two

or

three functionally indeterminate

features; whereas the clusters on the Final

Bronze Age site (1000-750 BC) of Roztoky were

composed of one large workshop with clay

looms, oven, etc., one or two other production

features, four storage pits and some smaller

indistinct pits (FIGURES-3 . The farmstead at

Bilina (Iron Age, La Tene B 1, about 400

BC)

1

2

Roztoky, District

of

Prague-West.

3

DobromBfice, District

of

Louny.

4

Hadonice , District of Louny.

C

4

60km 5 Radovesice, District of Teplice.

The Vinoi:

brook project , including

Cakovice-Prague

9.

the site of

8/11/2019 Ant0650332.pdf

5/16

6 MARTIN

KUNA

0

CI

0

t

FIGURE. Cakovice,

the Early Bronze Age

3

habitation site.

1 Large, shallow

pits (huts?).

2 Storage pits.

3

Shallow pits of

I

1 a schematic plan of

2

I

unknown

3

function.

- 4 Graves (Kovcifik

1983).

0

I

I

0

0'0

F I G K R E3

Bronze Age habitat ion site .

1 Large pits workshops.

2

Shallower pits with production traces.

3 Storage pils.

4 Pits of unknown function (excavat ed by the

(1u th or).

Roztoky, a sch ema tic plan of the Final

consisted of a post-ho le structure, one 'sunken

hu t' , one storage pit an d about four larger pit:s

with few distinctive elements (Waldhauser

Holodfiak 1984). By contrast, no such clusters

have been observed at th e large Iron Age site at

Radovesice during the later Iron Age (La Tene

C-D, about 250-0

BC:

Waldhauser

1984;

here

FIGURE

A). The 'clusters' of this period wen?

usually formed only by one su nken h ut which

can be understood either as a change in storage

and other habitation activities or

as

a conse-

quence of loss of all the shallower pits durin,g

the rough removal of the top soil. Anyway, the

definition of the cluster composition char-

acteristic for a particular site, region or period of

time

is

an important step in interpretation

of

an17 habitation site.

The inference that similar social units,

namely households of different periods , can be

reflected in a very different way (quantita tively

as well as qualitatively) can help u s to under-

stand those archaeological cultures which lack.

enough habitation sites and features. The lack of

obvious settlement sites and the occurrence

of

only

isolated pits during certain prehistoric

periods has usually been taken as an indication

of a specific subsistence system, settlemerit

8/11/2019 Ant0650332.pdf

6/16

STRUCTURING OF PREHISTORIC LANDSCAPE

7

0 1 0 2

0

0

0 3 m

-

pattern or high population mobility (typical for

some Eneolithic cultures, Middle Bronze Age,

etc.). However, this need not be the case. Some

recent large-scale excavations have made clear

that the relics of those cultures are not

so

much

isolated in space, as usually few in number and

sporadically dispersed within the habitation

areas. Their mutual distances are often about

several tens of metres, which may correspond

with the distances between the household clus-

ters of other prehistoric periods. This is the case

of a Funnel Beaker site at DobromBfice where

about six small pits contrasted with

200

features

of other cultures (Zapotocky Smri. 1980; here

FIGLJRE).

It

is possible that these isolated

features represent something like minimal

household clusters which are, of course, much

less likely to be found by smaller excavations.

Variability of this kind (namely the variability

in number of the pits per cluster) can be caused

by secondary factors (like the frequency of

rebuilding of storage pits) while the main

underlying settlement pattern could have been

about the same as in other prehistoric periods.

There is, of course, no reasonable way to define

household clusters in those archaeological con-

texts where the average number of features per

cluster is supposed as smaller than 1,or where

110 exactly identifiable and quantifiable

archaeological features can be observed.

The number

of

households within a

habitation area

An estimation of the average number of house-

holds occupying a given habitation area can be

FIGURE. DobromCNce,

a

schematic plan

of

the

Funnel Beaker

habitation site.

1 Storage pits.

2 Pits of unknown

purpose (Zapotockq-

Sm r i

1980).

achieved by a simple numerical operation (cf.

Neustupn 1983; 1986). An indispensible pre-

condition of such an operation is, however, the

estimation of other variables, such as the total

number of household clusters within the area,

the mean lifetime of the household clusters and

the total time-span of the occupation.

If

house-

hold clusters cannot be identified we have to

know at least the total number of certain specific

features (e.g. huts , storage pits, e tc.) , the time-

span of occupation, the average frequency of

rebuilding of these features, and the average

number

of

features that were simultaneously

used by one household.

If

a household cluster

(or

a house) correspon-

ded to the activity of one household and if its

average lifetime was between 10 and

25

years

(probably the maximum according to published

estimates, cf. e.g. those for a Neolithic long

house; Neustupny 1983: 108), hen, in most of

the habitation sites we come to a surprising

inference: the number of households that once

lived in the area might have been probably

much smaller than were our intuitive estimates.

It can be assumed that one household could

have produced even more than 50 different

features during one century; that means that

even the empirically very large and densely

occupied sites can bear witness, in fact, for the

activity of a small number of households, per-

haps only one or two (cf. e.g. the

Iron

Age

habitation area of Radovesice which was

completely excavated: Waldhauser 1977; 1984;

Holodiiak 1967; here FIGURE A). These esti

mates fit in well with the results of demographic

8/11/2019 Ant0650332.pdf

7/16

3 3 8 MARI IN KlJNA

analyses concerning prehistoric cemeteries of

various periods. The size of the population

groups using the cemeteries is very often sup-

posed to have been about 15-20 individuals

(Encolithic, Iron Age: Stloukal 1981; Neus-

tupny 1983; Holoddak 1987).

The most complicated problem of these cal-

culations is usually represented by the esti-

mation of total habitation remains (extent of a

habitation area) and their general chronological

position. Having no idea about the whole, we

cannot decide how many households occupied

e.g. the sites of Cakovice or Roztoky (FIGURES

2-3)

because the five or six observed household

clusters coming from the same archaeological

phase

c.

100 years) could be either a product of

five

to

six simultaneously existing households

or the remains of five rebuilding events of the

same household. The fact that many completed

excavation projects in Central Europe very often

lack an explanatory potential of this kind prob-

ably results from the prevailing inductive

approach underlying the fieldwork activities.

Excavations of continuous areas - as large as

possible but almost never covering the total

habitation area or a substantial part of it have

been unambiguously preferred to sampling the

site as a whole (Neustupny 1982b). A con-

tinuously-excavated area very often, of course,

succeeds in discovering clear and intelligible

individual contexts, but it is , from the point of

view of the whole site, usually unrepresen-

tative,

a t

least insofar as concerns most of the

large prehistoric habitation areas (often cover-

ing about 10-20 hectares).

The dynamics

of

habitation areas

The activities of individual households of the

same community can be found in one

or

more

habitation areas if a longer period of time is

taken into acxount. The very fact that we cannot

find the remains of all the cultural periods or

phases within the same habitation areas indi-

cates that

a

kind of mobility of prehistoric

households was common. The observed dis-

continuity of habitation places has been tradi-

tionally interpreted as the result of different

'historic events' like invasions, epidemics or

other catastrophes, etc ., not as a logical part of

settlement processes. One of the first processual

explanations has been given by B. Soudsky

(1966) in his model of periodic shifts of

settlements consequent upon the system of

slash-and-burn agriculture. This model, pre-

supposing periodic shifts in about each gener-

ation, was, however, articulated in the 1960s

and is no longer accepted. One other model

published since that time

is

analogous in prin-

ciple, but different in time scale. This is the

model of rotation of habitation areas based

upon the observed chronological patterns of

some Late Bronze Age sites in northwest

Bohemia (Smri

1987).

Periodic shifts in inter-

vals of about 200 years have been explained as a

necessary consequence of the exhaustion of the

micro-environment of the site (especially con-

cerning e.g. reserves of wood and forage). In

general, it seems that there is, for the time being,

no better processual explanation.

Discussion about the dynamics of habitation

areas in Czech archaeology is still characterized

by one latent presupposition. It is the intuitive

notion that the village-like form of habitation

was the prevailing or the only possibility in the

past. Such an idea needs not always be correct,

being most probably a projection of the contem-

porary picture of the Czech countryside on to

the archaeological record. This approach pre-

supposes the spatial and temporal coordination

of households in arranging their habitation

areas and in their occasional moving from place

to place.

The archaeological record itself is very

ambiguous in this respect. Firstly, there are

more prehistoric periods in which evidence for

a dispersed settlement pattern prevails (Late

Eneolithic, Early and Middle Bronze Age, Early

Iron Age, La Tene period, etc.; this inference is

based upon the preliminary results of the Vinof

brook project,

cf.

below, page 342). Secondly,

even if we can suppose the presence of more

than one contemporaneous household within

the same habitation area, their coordination in

time and space can mostly not be proven.

I t

seems that a random, short-distance mobility of

individual households by rebuilding their

dwellings within the site is most probably the

typical way that habitation areas developed.

Apparently, rebuilding of households occurred

only exceptionally on exactly the same place,

because there are usually no superpositions 1

the features and clusters of the same archaeo-

logical phase. A systemic reason for more

or

less,

regular transfers of household activities must

obviously have existed: hygienic reasons; look-

inn for an unexdoited area from the Doint

of

8/11/2019 Ant0650332.pdf

8/16

STRlJCTURlNG

OF PREHISTORIC

LANDSCAPE 9

view of clay sources; space for storage pits;

profitable use of the exploited area for some

specific act ivity, symbolic and religious reasons

-leaving the house after the death of its builder;

etc. The distance over which the household

usually moved was, however, mostly very short

and probably involved just leaving the range of

refuse and debris of the previous household

(household cluster).

I t

seems that there is often

an almost standard distance (30-50 m) between

two neighbouring clusters of the same or the

immediately following phase (cf. the Early

Bronze Age site at Cakovice: FIGURE) or the La

Tene C-D site at Radovesice: FIGUREA) which

could support such a supposition. This model

is, however, not universal (cf. the cluster A in

FIC~JREA, composed of huts of three conse-

cutive phases).

The continuous relocation of households

over short distances and the avoidance of the

over-exploited parts of the habitation areas can

be understood as two main general principles of

the habitation activities explaining their spatial

patterns. The degree of spatial continuity would

depend upon the time-scale used. We can

illustrate this point at the site of Radovesice

(FIGUREB). Whereas there is a substantial

spatial overlap between any two succeeding

chronological phases, a spatial shift can be

observed if phases more distant in time are

considered. The distribution of the La Tene

A

remains almost fully coincides with the extent

of features from the preceding Hallstatt

D

phase.

The features of the La Tene B phase do appear

within the area of La Tene A period but they are

outside the area of Hallstatt D relics. La Tene

C-D

features partly share their area with La

Tene B activity range but they are situated apart

from the Hallstatt

D

and La Tene

A.

This

phenomenon could be simply described as a

gradual growth of the site or a typical example

of horizontal stratigraphy if it were a cemetery

or a modern town where the once-erected

buildings

or

monuments remain

in

function

until the final abandonment of the site. Such a

cumulative growth cannot be suggested for a

prehistoric habitation site, and it can be sup-

posed that there were no functional obstacles

to

building new dwellings

or

other facilities

within the household area of the preceding

phase (cf. the cluster A in FIGUREA). Neverthe-

less,

the long-term gradual shifts of habitation

areas can be observed in many prehistoric sites

(Kuna Slabina 1987). As further examples of

the gradual shifts of habitation areas many Late

and Final Bronze Age sites in Central Bohemia

(Bronze D-Hallstatt

A

versus Hallstatt B) can be

noted: the relics of both the periods usually

appear within the same habitation areas (sites)

but their distribution is often (at least partly)

disjunctive.

The Radovesice site offers a preliminary

model for the dynamics of prehistoric habita-

tion areas. Such a model may, however, be only

one of the possible explanations and it may be

changed after more data is known. It seems that

most prehistoric habitation areas (omitting

specific habitation areas, such as hillforts, etc.)

were occupied by a few households perhaps

1-3, cemeteries, however, attesting about 3-6

families. There is little evidence for any spatial

coordination of activities of individual house-

holds, but it cannot be excluded in principle (cf.

indications of a differentiation between the

living and storage areas within some sites, etc.).

It seems that activit ies of each household occur-

red within a household range about 40-50 m in

diameter, This household range was probably

avoided both by other contemporary house-

holds and by the next (re)building activity of the

same household. The process of rebuilding

households went on within the very same place

until the area was exploited the results of such

an exploitationmay have left archaeologically

visible traces, such as covering the area with

pits,

or

not). This could go on during one to

three archaeological phases

1

00-300 years, cf.

e.g. the Late Bronze Age site of Radonice, FIGURE

5). The result was a dense covering of the

habitation area with debris and refuse (the

critical density was surely a relative, culture-

dependent variable). At this point the habita-

tion area was either left,

or

its gradual shift or

expansion started, the result of locating new

households already outside the exploited area.

Such a shift appears as spread or growthof the

site. This process continued for some period of

time, its length most probably influenced by the

environmental capacity of the area as a whole. It

seems that there was a certain limit for intensive

habitation activities within habitation areas. It

is, for example, suspicious that Late Hallstatt

period remains are usually not found within the

preceding Late or Final Bronze Age sites. This

could point

to

the fact that these areas were

not

suitable for occupation for the next few cen-

8/11/2019 Ant0650332.pdf

9/16

34 I MAKTIN K U N A

3

m

11

tur ies af ter the period when they were inten-

s ively explo i t ed . We m us t , however , adm i t tha t

there are other explan at ions for this part icular

phen om enon ( the cul tural preference for

another type of micro -environ men t , cf. Kuna in

press) , and that a n exact ana lysis of the spat ial

coincidence of these two archaeological cul-

tu res has ye t no t been mad e.

Various exp lanat ions can be suggested for the

above-ment ioned process , be they his toric

events , man-environm ent relat ions,

or

social

and symbol ic aspects of human act ivi t ies .

According

to

the presen t s t a te of

our

data as wel l

as our theoret ical knowledge, i t would be

premature to g ive o ne unam biguous so lu t ion .

It

seems, how ever, qui te ce rtain that al l of the

factors noted influenced the development of

prehis toric set t leme nt system s s imul tan eously ;

any of them may become the preferred expla-

nat ion when a part icular cul tural t ransi t ion is

quest ioned.

It

i s , for exam ple , very probable tha t

the abandonment o f many habi ta tion areas and

the format ion of l a rge communi ty areas ( see

page 345 below) could be connected wi th the

his toric event

of

invasion or gradual infi l -

t ration of foreign e thnic gro ups at the beginning

of the

R om a n

era

(1st

century AD) . Nevertheless ,

the processual -enviro nme ntal hypothesis abo ut

the mooiii ty

of

habi tat ion areas s t i l l remains

suff ic ien tly convincing (S m ri 1987 , see above,

page

338) for

most of the recorded s i te aban-

donmen ts . After a l l , the preference an d aban-

FIGURE

.

Radonice,

a schematic plan of

the Late Bronze Age

hubitution site.

1 Pits ofthe

transition horizon

MiddlelLute

Bronze

Age).

2

Pits of the Early

Middle Knoviz

culture.

3

Pits

of

the Late

Middle Knoviz

culture Bouzek et

al. 1966).

donm ent of certain types of locat ions could wel l

be connected even wi th some socia l phe-

nom ena a nd w i t h t he expec t ed sym bo l ic m ean -

ings of things a nd places

(e.g.

the per iod ic usage

of natural ly-protected places , hi l ls and prom-

ontories , for habi tat ion which correlates wi th

certain aspects of the development of burial

r i tes : Kuna in p ress) . Progress in this f ield can be

achieved only by analysing more habi tat ion

areas in detai l as wel l as in context of the wh ole

comm uni ty areas .

The identification

of

community areas

The set t lement system was not , of course,

exactly the sam e th roughout the w hole preh i s -

toric era. Differences and changes probably

exis ted in t he degree of household nuclea t ion

and mobi l i ty . Remains of variable set t lement

sys tems are a lways super imposed upon each

other , result ing in a d ens e ne t of archaeological

s i tes . The h abi tat ion areas usual ly co ncentrate

along s t reams ( the cr i t ical dis tance being

between 300 an d 500 m: Rulf 1 983; Kuna

8

Slabina 1987 ).The concep t

of

com m un i t y a reas

is based upon the not ion of a community

shar ing a com mon ter ri to ry an d coopera t ing in

certain economic and social act ivi t ies . The

main prob lem, however , i s to discover this

presumed uni t wi th in the archaeolog ica l record

and to define i ts dimensions. The theoret ical

defini t ion of the c om mu nity area do es not i tsel f

tell

us

anyth ing about the exp ected s ize of the

8/11/2019 Ant0650332.pdf

10/16

I

3

I;I ;URI,; 6.

h

1 Iiuts of the La T e n e B2-C2 period. B Chronological shifts

of

the habitation activities.

2 La

T h e

C1.

3

LaTeneC2.

1

Featu res of the Hallstntt D period.

4 La

Tene

CZ-Dl. 5 La

Tene

D1. 2 La Tene A.

6

La

T h e

D2.

7 Other pits. 3

La

T e n e B .

H u t s

with oven. 9

Huts

with storagt: pits.

4

La Tene

C-D

Waldhauser afaL' pers. comm.).

Radovesice, schematic

plans

of the Iron Age habit ation area

f-Itrbitation structures of the La T ne C-La Tene

D

periods.

8/11/2019 Ant0650332.pdf

11/16

8/11/2019 Ant0650332.pdf

12/16

STRUCTIJRINC; OF PREHISTORIC LANDSCAPE

4

FIGURE. Vinof brook project.

A

Community areas of the Late Ha ht a tt ih rl y La

T h e

period (c. 650-400 RC)

1 Promontory sites.

2 Sites with intensive habitation activities.

3 Low-density find scatters.

B

Community areas

of

the La T h e B-D period (400-0

R C ) ;

low-density find scatters prevail

8/11/2019 Ant0650332.pdf

13/16

344

MARTIN KUNA

0

0

F K I J R E

.

Vinoi: brook project

A Community areas

of

the Early Roman period

0-150

AD)

Sites with intensive habit ation activities.

2 Low-density scatters.

B

2 Cadastral bounda ries.

Community (cadastral) areas of the 19:jOs.

The area of towns and villages.

8/11/2019 Ant0650332.pdf

14/16

STRUCTIJRING O F PREHISTORIC LANDSCAPE

345

suggest that a community area usually repre-

sented a territory of about 4-8 sq. km if it was a

circle around the living places. In fact, the

community areas were probably larger, having

an ovoid or irregular shape with the longer axis

perpendicular

to

a stream. The number of

households inhabiting such a territory is not

quite clear. Regarding the context of the habita-

tion sites as well as cemeteries we can suggest

that the communities were rather small (up to 6

households if there were more than one habita-

tion area within a community area; cf. Neus-

tupny

1983 .

The size of community areas and the general

position of habitation places did not change

much during all of agricultural prehistory. The

first significant change of the settlement system

can be suggested only for the Early Roman

period during the 1st century

BC.

Instead of

about

15-20

smaller community areas 5-6 large

areas have been observed within the same

region (FIGUREA). In addition, the habitation

areas themselves moved into places previously

avoided. Both these phenomena witness a

transformation of the social system and its

spatial correlates. The tendency towards the

enlargement of communities is supported by

the evidence of large cemeteries that appear in

the Roman period which cannot be compared

with the smaller cemeteries of the most part of

the preceding periods.

Large community areas existed throughout

the whole Roman period and possibly also in

the Migration period. It is very interesting that

the settlement distribution of the High Middle

Ages (13th century onwards) which in prin-

ciple survives today goes back to the old

prehistoric pattern both as concerns the density

and the size of community areas and the loca-

tion of habitation areas (FIGURE

B) .

Settlement zones and general population

density

It is very probable that community areas formed

a stable structure during major parts of prehis-

tory. Within this structure the average extent of

community areas and the average number of

households within the communities did not

probably change substantially, the only known

exception being the change at the beginning of

the Roman era. If there was any oscillation of the

total population number this would probably be

reflected not in the increasing

or

decreasing

I

B I

FIGURE0.

settlement zone.

A Early Bronze Age.

B

C

Vinof

brook

project. Extent

of

Early

Iron Age

Hallstatt

C).

Late

Hallstatt

period

Hallstatt

D).

population density within community areas but

by increasing

or

decreasing the total number of

communities (and their community areas)

within a given region. I call the territoTy that

was actually inhabited and used a settlement

zone. In the example of the landscape project of

Vinoi: brook we may observe changes in the

extent of the sett lement zone during the Bronze

and Iron Ages. It seems that some areas were

abandoned during this time and recolonized

later.

It

can be suggested that the settlement

zone formed a coherent cluster of community

8/11/2019 Ant0650332.pdf

15/16

346

MARTIN KLJNA

areas at any moment

of

its existence, ensuring

probably the most effective interaction of

communities. The observed tendency of

population growth and decrease cannot, how-

ever, be generalized before more landscape

studies are carried out in other regions.

References

BENES, J .

D. KOUTECKY.987. Die Erforschung der

Mikroregion Lomsky potok [Probleme und Per-

spektiven], in Cerna (ed.):31-8.

B O E L I C K E , U. 1982. Gruben und Hauser: Unter-

suchungen zur Struktur bandkeramischer Hof-

platze, in J . Pavuk (ed. ), Siedlungen der Kultur

mit Linearkeramik in Europu: 17-28. Nitra: Insti-

tute of Archaeology.

BOG~XKI,P.I. R. GRYGIEL.

981.

The household

cluster at BrzeSC Kujawski

3 :

small site methodo-

logy in the Polish lowlands, World Archaeology

13 1): 59-72.

BOUXK, J., D. KOUTECKY E. NELISTUPNY.

966.

The

Knoviz settlement of northwest Bohemia. Prague:

National Museum. Fontes Archaeologici Pragen-

ses 10.

C E R N A ,

E. (ed.). 1987. Archaologische Rettungst-

atigkeit in den Braunkohlengebieten und die

Problematik der Sied lungs -gesch ich t l i chen For-

schung. Prague: Insti tute of Archaeology.

H O L O D ~ ~ A K .

. 1987. Methodische Probleme bei der

Bestimmung von Populationsgrosse in der Laten-

ezeit, Anthropologie (Brno)

2512: 143-54.

KLArFrE, J .

1989.

Poznamky k nekterym tendencim v

soutasnem vyvoji archeologicke pramenne zak-

ladny [Bemerkungen zu einigen Tendenzen in

der Entwicklung der archaologischen Quellen-

basis], Archeologicke rozhledy

41: 75-84.

KOVARIK,

.

1983. SidliSte uneticke kultury v Praze

9-Cakovicich [Eine Siedlung der Uneticer Kul tu r

in Prag 9-Cakovice], Archaeologica Pragensia 4:

13 7-68.

K L ~ N A . 1990. K problemu registrace poruSenych

archeologickych naleziSf [Towards the regis-

tration of damaged archaeological sites], Studie a

z p r o v y Okresniho muzea Praha-vqchod 10

1988): 21-44.

I n press:Social system of the Iron Age as reflected

on the microregional level, in Fonctionnernent

social de I hge d u fer: operat eurs et hypotheses

p o u r la

France.

K L I N A ,

M.

J .

KLAPSIB.

1990.

Poznamky ke koncepci

terhni archeologicke prace [On the character

and concepts

of

field work in contemporary

Czech archaeology] , Archeologicke rozhledy 42:

435-445.

K L : N A ,

M .

M . SLABINA.987.

Z u r

Problematik der

Acknowledgements. 1would like to thank Dr E. Neustupnq

and Dr N. Venclova

for

their comments on this paper. I am

indebted

to

Dr M . Slabina for his substantial work in

classifying all pottery fragments from the Vinoi. brook

project.

M y

thank s also go

to

Dr J. Waldhauser and V. SalaC,

who enabled me to work with the unpublished materials of

the Radovesice site . All drawings were ma de by A. NBmcova.

Siedlungsareale (in der 13ronzezeit), , in Cerna

(ed.):263-78.

KYNCL,

.

1987. Vztah vegetace a osidleni

v

mikrore-

gionu Lui ickeho potoka na Kadalisku [The rela-

tionship of vegetation and settlement

in

the

microregion of the stream Luiicky potok in the

area of Kadafi], Archeologicke rozhledy 39:

622-8.

N E U S T U P N Y ,

E .

1982a. Prehistoric migrations by infil-

tration, Archeologicke rozhledy 34: 278-93.

1982b.

Optimalizace vyzkumu archeologickeho

regionu [Die Optimalisierung der Erforschung

archaologischer Region], in J . Hrala (ed.),

Metodologicke problemy Ceskoslovenske archeo-

logie:

178-82.

Prague: Insti tute of Archaeology.

1983.

Demografie pravekych pohfebiSt [The demo-

graphy of prehistoric cemeteriesj. Prague: Insti-

tute of Archaeology.

1985. K holocen u Komofanskeho jezera [On the

holocene period in the Komofany Lake area],

Pamatky archeologicke 76: 9-70.

1986.

Sidelni arealy pravekych zemedelcu

[Settlement areas

of

prehistoric farmers],

Pamutky archeologicke 77: 226-34.

1987. Praveka eroze a akumulace v oblasti Luiick-

6ho potoka [Prehistoric erosion and accumu-

lation in the Lui ice brook basin], Archeologicke

rozhledy 39: 629-43.

1991. Community areas of prehistoric farmers in

Bohemia, Antiquity

65: 326-31.

PAVL

8/11/2019 Ant0650332.pdf

16/16

STKIJCTURING O F PREI IISTORIC

LANDSCAPE

347

Kadovesic [Skelette au s de m latenezeit licheri

Graberfeld bei Kad ovesice], Praehistorica 8

(Varia archaeologica 2 , Charles Universi ty ,

Prague]: 221-3.

TURKOVA,. M . K L JN A .987.

Z u r

Mikrostruktur der

b ronzezei tl ichen S ied lu ngen , in Die Urnen-

fe lderku l tu ren Mit te leuropas . Symposium

Liblice 1 985 : 217-29. Prag ue: Ins titu te of

Archaeology.

VEL~MSKY,

.

1986. Archeologicke vyzkum y v oblasti

t e ib y sev e ro cesk y ch h n ed o u h e ln y ch

do l u -

problemy, vysledky a perspektivy [Die archao-

log ischen Ausgrabungen im nordbohmischen

W

ovesice, Bez. Teplice in Nordwe stbohm en (Uber-

s ich t der Ausgrabungen und S tand der

Aus wer tu ng)] , Archeo log ickk rozh ledy 29:

144-76.

984. Mobili tat und Stabil i tat der kelt ischen

Besied lung in Bohmen , in S tud ien

zu

Sied-

lungsfrage n der Latknez eit: 167-86. M arburg:

Veroffentl ichungen des Vorgeschichtl ichen

Seminars Marburg . Sonde rband

3 .

IDHAUSER, J. P. HOLODNAK.984. Keltske sidlig te

a pohfebiSt6

u

Biliny, 0 Teplice [Kelt ische Sied-

lung und Graberfeld bei Ril ina, Bez. Teplice],

Pam dtkv archeoloe icke 75: 181-216.

Braunkohlengeb ie t ] , i n T . Vel imsky ( ed . ) ,

Archeologickk vyzkumy v severozapadnich

Cech ach v letec h 1973-1982: 7-35. Prague: Insti-

tute of Archaeology. Archeologicke studijni

materialy 15.

WALIIHAUSER,

.

1977 . Keltske sidliSt6

u

Radovesic v

severozapadn ich Cechach : Preh led vyzkumu

a

stav h odn oce ni [Die kelt ische S ied lung bei Kad-

WI NT ER, .C . 1976 . The archaeo logical househo ld

cluster in the valley of Oaxaca, in K.V. Flannery

(ed .) , The e arly Me soame rican vil lage: 25-31.

New York (NY):Academic Press.

ZAPOTOCKY,M. Z SMRZ.1980. SidliStB KNP v

Dobromeficich, okr. Louny

[A

Funnel Beaker

sett lement si te in DobromBfice, the Louny

distric t], Arche ologicke r ozhle dy 32: 481-9.