On the development of the town of Viljandi in the light of earliest archaeological find complexes/Viljandi linna kujunemisest varaseimate arheoloogiliste leiukomplekside pohjal.
Haak, Arvi ; Russow, Erki
Introduction
Among the nine medieval towns in the territory of the present-day
Estonia, Viljandi was one of the four members of the Hanseatic League.
The inland town of Viljandi was situated next to the castle of the
Livonian Order, one of the mightiest in Livonia, and it has widely been
suggested that the development of the town was heavily dependant on the
castle. However, the traces of the Medieval Period still visible in
Viljandi today are all but numerous. In addition to the castle ruins,
only the church of St. John, and the location of the main streets within
the medieval town date back to the Middle Ages. The medieval town was
fully destroyed during the Livonian War (1558-1583) and the following
Polish-Swedish Wars (1600-1622/23), and town privileges were granted
anew to Viljandi only in 1783. A few remains of buildings dating from
the Middle Ages have been recorded during archaeological investigations,
but it is impossible to reconstruct the town structure solely on that
basis.
Written sources concerning medieval Viljandi are also far from
abundant (see the following chapter). Thus, investigators have turned to
other sources. A long-established tradition for research is the analysis
of town plans (Neumann 1911; for Viljandi: Alttoa 1978). Other sources
for studying the medieval period in Viljandi and especially the
formation of the medieval town are the finds collected during
archaeological excavations. These have also been discussed in regard to
Viljandi (Valk 1993; 1995; 2005; Haak 2005 (1)), but a detailed analysis
of the finds from the early contexts needs another research.
A great share of archaeological finds from Viljandi, although
originating from rescue investigations, has been collected from
stratified contexts, thus making context-based analysis possible. As
usual, the strata that have deposited just upon the virgin soil have
been better preserved than those originating from the periods of
well-established urban life. In the case of Viljandi, strata from the
13th-14th centuries and those connected to the destructions caused by
the Livonian War (from 1560 onwards) have yielded most of the finds. In
the present study, we concentrate on the former contexts, in an attempt
to reconstruct the pattern of town formation in Viljandi.
As mentioned above, the data concerning buildings from that period
are extremely scarce, most likely due to the fact that wood and other
organic materials survive only in certain circumstances in Viljandi,
namely in waterlogged contexts, which are not very numerous. Thus, in
most cases we have little if any clue from where the strata unearthed
during archaeological investigations actually originated. Taking this
into account, our main objective was the reinvestigation of the
collected finds, and the drawing of conclusions on that basis.
Investigation of the genesis of Viljandi on the basis of written
sources
As the medieval town archives of Viljandi have not been preserved,
there exist only limited data about medieval Viljandi in the written
sources. In addition to the beginning of construction works at the
castle of the Livonian Order in 1224 (Wartberge 1863, 4), there exists a
note that the master of the Livonian order, Willekinus de Endorpe, had
confirmed the town privileges and established the borders of the town
mark in 1283 (LGU I, 40). The letters of confirmation of the privileges
from 1481 (Freymann 1918, No. 1) and from 1533 (Freymann 1918, No. 3)
have been preserved. In connection with the former, the destruction of
the town of Viljandi by the Russian troops in 1480 has been mentioned
(Russow 1993, 60 f.; Kelch 2004, 114), as the town privileges were burnt
during that event (Freymann 1918, No. 1). During the Livonian War, the
town was totally demolished, so that only five (Russow 1993, 127) or six
(Renner 1995, 145) houses survived. (2) This was followed by a fire in
1591 (Bergengrun 1897, 37).
There are some sources regarding the appearance of medieval
Viljandi. The most informative of these is an inventory carried out by
the Poles in 1599 (Jakubowski & Kordzikowski 1915; Viljandi linn
1998; Alttoa 1999). In addition to the owners of property at the end of
the 16th century, it also includes the names of ca. 50 citizens of
Viljandi before the Livonian War. The sources also contain the date of
construction of the Franciscan monastery of Viljandi (1466-1472, Alttoa
1995, 164), the church of which is the only medieval building that has
survived within the town and is currently named St. John's--the
church of the town congregation. (3) In addition to that, a will of a
citizen of Tallinn mentions two houses by the market square in Viljandi
(Freymann 1918, No. 5; Alttoa 1977, 28). However, there exists nothing
else, and all of that had been published by the 1920s.
The situation concerning documents from the early periods of town
development is pretty much similar in other towns of Livonia. Only in a
few cases, the documents listing town privileges from the 13th century
have survived (e.g. Angermann 2001; Kala 2002), although the process of
discussing the privileges to be granted to a small town in the territory
of the Livonian Order has been recorded (Johansen 1938). As historian
Juhan Kreem (2002, 39 f.) has pointed out, the privileges of Livonian
towns were renewed in writing before the 16th century only in cases when
the earlier document had been damaged or destroyed. Only from the 16th
century, there are sources showing that the master of the Livonian
branch of the Order had confirmed the privileges of Livonian towns in
writing after his succession to the office (Kreem 2002, 39 ff.).
Until the 16th century, written sources other than town privileges
are minimal about most of Livonian towns except Tallinn, where the town
archives include a remarkable collection of documents from the Middle
Ages. In all other cases, we have to rely on casual documents, and data
from archaeological investigations. It is only during the 16th century
that a full treatment of the Livonian small town, based on written
sources, becomes possible (Poltsam-Jurjo 2009).
The earliest writing on the history of Viljandi (if we exclude
notes from chronicles) has survived from 1784 (Pullat et al. 2006, 309
ff.). It seems that the city council answered to some kind of a
questionnaire. The oldest town plan with street network and plots, drawn
by Johann Christoph Brotze, originates from the same period (Pullat et
al. 2006, 315). Interest toward the glorious past of Viljandi increased
remarkably during the 19th century, and several Baltic German authors
published their studies of the history of the town. The most
comprehensive of these was written by Friedrich Amelung (1898). He was
the first to offer his interpretation regarding the oldest part of the
town. According to him, it was located next to the castle, containing
St. John's and the market place (Amelung 1898, 15); however, he
gave no reasons to his point of view.
The first drawing depicting an author's reconstruction of
medieval Viljandi was made by Karl von Lowis of Menar (1914, first
published in Lowis of Menar 1922). His drawing shows the location of
streets, town church and the Franciscan monastery. As has been pointed
out, the map combines the Swedish map from the 1650s (4) depicting town
fortifications with the street network from Brotze (Alttoa 1978, 48).
Fragments of the town wall had survived at the beginning of the 20th
century, so that the combination of the maps might not have been
difficult at that time. Although the map of Lowis of Menar added details
to the drawing of the main castle, probably his personal recording of
the excavated ruins of the late 19th century, it does not add much to
our knowledge about the medieval town. Despite that, his drawing has
been reprinted several times.
Chronologically the next reconstruction drawing was published by
August Westren-Doll (1929), a pastor of Viljandi, deeply interested in
history. He could use the Polish inventory (Jakubowski &
Kordzikowski 1915), and attempted to locate the plots mentioned in this.
Although it was clear that the author attempted to reconstruct the plan
of medieval Viljandi, he neither gave clearer temporal hints, nor
discussed the development of the situation shown on the plan.
In an article about medieval Livonian towns, American geographer
John Leighly (1939, 275, fig. 14) published his reconstruction of
medieval Viljandi. As the former investigators, he had combined the
Swedish map from the 17th century with the street network from Brotze.
However, he was the first to group Viljandi/Fellin, Valmiera/Wolmar and
Koknese/Kokenhusen on the basis of their location on interfluve spurs
(1939, 271). He stresses that this type of castle and town combination
is typical of Livonia, as glacial and glacifluvial deposits provide
smooth surfaces, where towns could easily be built next to castles, in
contrast with most of Europe, where the landscape is uneven and usually
hilly (1939, 271 f.). In case of Viljandi, he stressed that the location
of the market-place in the centre of the town indicates a commercial
town not too dependent on the castle, as could be the case in Koknese
and Valmiera, where the market place was situated just in front of the
castle (1939, 275 f.).
The comprehensive article on the formation of Viljandi by historian
Paul Johansen (1955) gives his interpretation of the development of the
town, connecting the most influential period with the activities of
Bernard of Lippe and the years between 1217 and 1223 (Johansen 1955, 219
ff.). His treatment includes a reconstruction drawing (Johansen 1955,
plate), where he tried to depict the 99 plots, known from the Polish
inventory. Johansen's drawing has some of the characteristics from
Lowis of Menar and Westren-Doll.
On the question of town formation, the article by Johansen pointed
out several parallels between Viljandi and Lippstadt, which he connected
with the activities of Bernard of Lippe (Johansen 1955, passim). He
distinguished a group of towns, characterized by the location of the
town just in front of the castle, so that the town walls are located on
the extension of the castle walls, and introduced them as a group called
town "on a shield" (Germ. Auf dem Schilde) (Johansen 2005
(1941), 373 f.). From the latter article, the concept was used by
Estonian art historian Armin Tuulse (1942, 55).
Architecture historian Elmo Raadik discussed the formation of
Viljandi town area in his graduation thesis (Raadik 1960). His view was
most likely influenced by the ideological paradigm of the period that
stated that the towns in Estonia could not have been a German innovation
(see Mall & Russow 2004; Lang 2004 for that cause). According to
Raadik, the town was preceded by a settlement of Estonian artisans next
to the market square (Raadik 1960, 56). Simultaneously with the
construction of the castle, German artisans started to settle in front
of the castle, thus almost the whole town area was inhabited. In the
further developments of the town, Raadik was the first to discuss the
location and function of roads starting from or passing by Viljandi
(Raadik 1960, 54 ff.). The argument of a prehistoric settlement in the
territory of the later town, which also determined the location of the
market place, was also given in the general treatment of history of
architecture in Estonia (Arman 1965, 31), but this was stated on purely
hypothetical basis.
In the second half of the 1970s, art historian Kaur Alttoa compiled
a historical overview of the old town of Viljandi (Alttoa 1977). His
main arguments regarding town formation were published as an article
(Alttoa 1978). Alttoa tried to find functions to the medieval streets
(Fig. 1) if there existed a prehistoric settlement, and concluded that
in that case, only Pikk Street and Vaike-Turu (Small Market) Street
(medieval Kauba Street) could have any function, while the main streets
of the medieval town--Lossi and Kauba (medieval Storke) did not lead
anywhere before the medieval town and its market place had formed, and
bridges had been erected across the moats (Alttoa 1978, 53). From that
basis, he produced a hypothesis that a prehistoric settlement site could
have been located near the crossing of Pikk and Vaike-Turu streets
(Alttoa 1978, 53). In his reconstruction drawing (Fig. 1), Alttoa
included only the houses and fortifications, which can be located with
greater certainty on the basis of the Swedish map or the text of the
above-mentioned Polish inventory.
A brief look at the interpretations of the written sources, which
are of interest in studying the development of Viljandi in the 13th
century, results in the following conclusions:
1. The conquest of the prehistoric hill fort in 1223 and the
beginning of the construction of a stone castle in 1224 form the
terminus post quem for any medieval settlement in Viljandi.
2. By 1283, a settlement must have existed, which was granted the
privileges of a town.
3. All the other arguments for any date of town formation, or
earliest settlement areas within the territory of the medieval town are
hypothetical in nature. The attempts to reconstruct the earlier roads
are interesting, but cannot be relied upon without any supporting
evidence from archaeological sources.
4. As for the town plan, there are sources from the late 16th
century (Polish inventory) and 17th century (Swedish map). Even if we
take it for granted that the overall layout had formed by the end of the
13th century (which remains rather doubtful), the alterations during the
following three centuries are beyond doubt. Thus, it is only with the
help of archaeological sources that it might be possible to develop a
reconstruction of the scope and spatial organization in Viljandi in any
period before the end of the 16th century.
[FIGURE 1 OMITTED]
History of archaeological investigations in Viljandi: sites and
issues
First notices concerning the medieval town of Viljandi on the basis
of material remains were made in 1911, as water pipelines were
constructed in the medieval town area. During those works, the remains
of two principal town gates were unearthed. In addition to that, three
older cobblestone pavements were traced below the street level (Freymann
1918, VI). In addition, a few finds from the pipelines have been
preserved in the Museum of Viljandi (VMT 84). At least one of them, a
fragment of a Russian-style field flask, dates from the second half of
the 16th century (cf. Tvauri 2004, 403 f.; Kildyushevskii 2006, 101 ff.,
fig. 5.8: 1-9).
The first archaeological investigations in the town area of
Viljandi took place in 1979. Under the direction of Henn Moora, the
museum archaeologist, and art historian Kaur Alttoa, an attempt was made
to locate the town wall in the block between Oru, Tartu and Lossi
streets at the northern border of the town. In the next two years,
archaeological excavations took place in the territory of the medieval
monastery of the Franciscans, located north of the church of St.
John's, and in 1981 also in the territory of the Riga Gate of the
medieval town (Selirand 1981; 1982b). Both of these works, directed by
Alttoa and archaeologist Urmas Selirand, took the localization of
medieval constructions as their main task. However, some of the
archaeological finds from the area of Riga gate were dated to the
prehistoric period by Selirand (1982a, 18). This may be the result of
the hypothesis by Alttoa about the possible location of the early
settlement site, but the finds themselves do not allow the present
authors to agree with the suggestion.
The first archaeological excavations in today's sense of the
word took place in 1989, as a water pipeline was established at Munga
Street (Fig. 2: 1). These investigations, as well as the excavations in
the territory of the churchyard of St. John's (Fig. 2: 2, 3;
1990-1991), a small area in the block north of Market Square (Fig. 2: 4;
1991) and at the Tartu Gate of the town (Fig. 2: 5, 1992) were directed
by archaeologist Heiki Valk. On the basis of the results of his
excavations, Valk published the first article concerning the formation
of Viljandi on the basis of archaeological studies (Valk 1993). He
supplemented it with data collected in the meantime by 2001 (Valk 2005).
According to Valk, the most intensive layer connected with the
early town development was around St. John's, as well as in the
area to the north. As during the excavations near the choir of St.
John's, no remains of buildings could be traced, Valk proposed that
the area might have been used as the first market place of the town
(Valk 1995). Thus he stated that habitation near St. John's had
probably started together with the founding of the potter's
workshop, and considerably earlier than in other parts of the town (Valk
2005, 103 ff.). It seems that his dating was mostly based on the
locally-made pottery, as he stated that the pottery decorated with
straight and wavy lines should belong to the 13th century (Valk 1993,
222). The habitation traces north of the market-place, and next to Tartu
Gate he dated to the second half of the 13th century, and those at Munga
Street to the beginning of the 14th century. These dates were based on
radiocarbon dating (Valk 2005, 103 f.), which, however, had not been
duly calibrated. Valk also stressed that excavations near the Tartu Gate
and those in the vicinity of the market square had produced a
considerably larger share of imported ceramics in comparison to local
products than those in the vicinity of St. John's. He interpreted
the large share of local pottery, found in the area, as an indication of
population of ethnic Estonian origin (Valk 2005, 104).
[FIGURE 2 OMITTED]
Large-scale archaeological excavations took place in the town
centre, north of the market square in 1993-1996 under the guidance of
Aare Kodar (Fig. 2: 6). As currently only the finds are available for
detailed study, the results of these investigations can be used for this
cause only partially. Still, the great amount of material dating from
the last third of the 13th century should be pointed out.
On the basis of the results of his investigations at the site 4
Pikk Street (Fig. 2: 8), in the territory of a medieval potter's
workshop, Andres Tvauri has concluded that the production of pottery in
that workshop took place between 1224 and the end of the 13th century
(Tvauri 2000a, 21). The first of these dates has obviously been taken
from written sources, the second one is probably a generalization of a
radiocarbon dating, collected from below the town wall in the same area,
where the layer containing production remains of the workshop has been
piled up against the town wall (Tvauri 2001a, 106 f.). Still, it seems
that the excavations did not give a definite answer to the question
whether pottery production had ended in the workshop by that date, let
alone the question of how long similar pottery was in use in Viljandi.
During the last decade, most of the archaeological investigations
in the medieval town area have been small-scale, usually connected to
the repair of pipelines. Still, remarkable information has been obtained
during these works, most notably from the northern part of the town, the
area between Vaike-Turu Street and the town wall (Fig. 2: 13), next to
the former cinema "Rubiin" (Kriiska et al. 2007). As such
monitoring works have been numerous, but the early finds collected are
relatively few, and have so far not led to alterations in the general
scheme of town development, we will not introduce them here, but
information available to date will be included in the find analysis, and
discussion about the development of different parts of the medieval
town.
The latest large-scale investigations took place east of the
mentioned former cinema and just south of the northern part of the town
wall (Fig. 2: 16; Tvauri 2010). In the excavation results, Tvauri has
pointed out strata of "13th-16th century", and a drainage
system just above the virgin soil, but no clear construction remains
(Tvauri 2010, 158 ff.).
In addition to topics mentioned above, special articles have been
dedicated to such topics as the construction and appearance of the town
wall of Viljandi (Tvauri 2001b), the archaeological ascertaining and
dating of medieval streets (Haak 2003) and the formation and destruction
of the suburb in front of Tartu Gate (Haak 2006). Although all of these
are based on the dating of finds collected during archaeological
investigations, the sources have not been discussed in depth in these
writings.
As a conclusion about the existing studies it may be said that the
question of the formation and development of medieval Viljandi has been
quite topical. Still, the usage of archaeological finds for that cause
has been limited, and some of the collected material (imported pottery
in the first place) has only been marginally used. There lies the reason
for the current attempt to include the dating of archaeological finds
from the earliest contexts to the sources available for discussing the
genesis of Viljandi.
13th-century finds from Estonian towns and castles
Strange as it may seem, most of the conclusions regarding the 13th
century in Estonian towns have either been made on the basis of pottery
finds, or the location of the stratum considered above virgin soil or
strata clearly of prehistoric origin. At the same time, articles
considering the "earliest" pottery can be counted on the
fingers of one hand. Thus, a short introduction to what kind of finds
can be considered as evidence of 13th-century habitation is necessary
(see also Mall & Russow 2000, 120 ff. for that cause).
The shift from "Final Iron Age" to "medieval
period" in Estonia is by no means obvious in the material culture
(see also Russow et al. 2006, 159). Although "new" item types
appear, which have not been used in pre-conquest Estonia (e.g.
proto-stoneware, highly decorated redwares, and greywares from Germany
and neighbouring territories, window glass, some decorations, etc.),
there are many everyday items which follow earlier traditions. On the
other hand, the find material of the period has not been a subject of
thorough studies either. So, usually the dating of the deposits in
question is based on coins, the existence of imported wares, or the type
of local pottery, similar to that found in hill forts, which have been
in use during the 13th century conquest, or the existence of brick
fragments in the strata. (5)
There are other item types which have undergone clear changes
during the 13th century (e.g. brooches, crossbow bolts, etc.), but such
finds are relatively scarce in urban contexts, and examples have been
collected either from cemeteries, hoards, or castles (weaponry). The
question concerning the end of 13th century on the basis of
archaeological material is even more complex, as the only indisputable
change recognized so far is the appearance of Siegburg stoneware and the
absence of certain earlier pottery groups (e.g. Paffrath-type globular
pots and possibly also highly decorated redwares).
So, the current study starts within the same framework. In order to
establish contexts that can be dated to the period of town formation, we
have started from their position in stratification, as well as the items
collected thereof. The results of our analysis were summarized in Table
1 (see next section). Criteria for dating the pottery finds are
discussed in the following chapter.
The finds from Viljandi
In the framework of the current study, the finds from the earliest
complexes, collected from 17 excavation plots (see Table 1) were
revisited. For that cause, all the archaeological investigations in the
territory surrounded with the medieval town wall, where finds from
strata deposited just above the virgin soil had been unearthed, were
taken into account. In order to get some kind of an overview from the
intensity of 13th-century habitation, all the stratigraphically later
contexts which included a reasonable amount of 13th-century finds were
included as well (i.e. contexts where one or two fragments, which can be
dated to the 13th century, appear side by side with overwhelmingly
16th-century material, were omitted).
Local pottery
The amount of potsherds from 13th-century contexts is enormous, and
for a better understanding of the situation, a large share of finds,
which have been mixed up with later strata should be added. It should
still be noted, on the one hand, that at least at the current state of
research, only rim sherds, or those which give an overview of the
ornament, can be used for dating purposes. The situation with the
chronology of medieval local pottery is, however, unsatisfactory. As for
the available data, the complex of the potter's oven (Fig. 3) can
be used as an "early" example, and the complexes at the Castle
of the Teutonic Order, as indicators of the last quarter of the 13th or
the beginning of the 14th century. As no later ovens have been found so
far, the origin of the later pottery remains uncertain.
[FIGURE 3 OMITTED]
It has long been stated that pottery with wavy and line ornament is
probably the oldest (Valk 1993, 222), as it resembles the products of
the Final Iron Age. To sum up, current investigations have not yet
produced a reliable chronology, while the complexes of early 14th
century also contain numerous fragments of such wares, so the existence
of the ornamentation alone cannot be used as a clear indicator of
13th-century habitation. The main focus on dating should thus be on
vessel shapes, especially rim shapes, with ornamentation as a supportive
argument.
The rim profiles of vessels of local clay, used in south and east
Estonia during the 13th century, have been classified according to Pskov
ware types, namely their local variants 3: 2 and 3: 3 have been
identified (Tvauri 2000b, 100 ff.). While the former form was in use
already since the 12th century, the latter was introduced during the
second half of the 13th century (ibidem). As the vessels are rather
similar in broader terms, but have some local differences, an attempt to
elaborate the classification might produce a more precise typology,
especially if more closely-dated complexes could be included.
As marked by Tvauri (2005, 50), vessels with a curved rim and
without a neck should be the oldest. These have been found in abundance
from the territory of the potter's workshop (Fig. 4). However, it
should be noted that from the analysed contexts, there exist none with a
greater number of rim sherds, which have not included a few sherds of
the 3: 3 type. Also the early contexts from the Castle of the Teutonic
Order, most likely originating from the 1260s-ca. 1300 (Haak &
Rannamae in print), included both variants. It can also be stressed that
those rim sherds of the 3: 2 type, where the connection to the vessel
body could be established, had a neck.
The lowest strata at the yard of the Museum of Viljandi, which seem
to be the oldest context on the basis of the imported wares, included
both variants 3: 2 and 3: 3, although the former dominated in the strata
just above the virgin soil. In the burning chamber of the potter's
oven, only pots belonging to subtype 3: 2 were found. So it seems that
the proportion of the share of these ware types might give additional
arguments for dating, but the current situation of research does not
allow including a more specific dating to the shift.
[FIGURE 4 OMITTED]
Imported wares
While concentrating next on the earliest fragments of imported
wares, found from the town area, it should be noted first that the use
of the finds connected with detailed discussions about the town genesis
in the 13th century is not without problems. Several reasons for that
can be pointed out. Firstly, fragments which could be dated to a narrow
period on the basis of morphological or typological criteria are lacking
among the finds of Viljandi. Although archaeological investigations in
west Estonian towns have unearthed rim fragments of proto-stoneware
jugs, or wall fragments with rolled stamp ornament, typical to the first
half of the 13th century (Russow 2006, 144), not even one such fragment
has been found from Viljandi. The wall and bottom fragments without
specific characteristics allow in some cases to suggest that the vessel
might have been produced even before 1250, but no solid proof can be
found to such a statement. Paffrath-style globular pots, connected
directly to the early phase of colonization, have been used until the
last decades of the 13th century, or even later, as shown by the results
of archaeological investigations in the Viljandi castle during recent
years (Haak 2004, 115). Thus, the use of the existence of fragments of
such vessels in the find material as an indicator of settlement
distribution is rather doubtful.
A second problem which cannot be ignored is the physical appearance
of the sherds. Usually the sherds are extremely small, and their breaks
show strong traces of wear. In addition to that, the find complexes
include very few conjoinable fragments. Both of these characteristics
refer to the fact that the deposition process of the broken vessels has
included many steps. An explanation may be the fact that a share of the
finds has been collected from excavations in the street areas of
Viljandi and thus the findspots might not indicate the original place of
usage of the vessels. The finds from the excavations in the yard of the
Museum of Viljandi (VM 10326; 10942), and near St. John's Church
(VM 10258) can be pointed out as exceptions. In spite of the
reservations listed above, the finds of sherds of imported wares from
the early stages of Viljandi still remain an important component in
assessing the topographical situation in the town.
In order to determine the dating of the earliest find complexes of
Viljandi, it is not sufficient to take into consideration only the finds
from that particular town. For a more reliable evaluation, it is
unavoidable to compare the finds from Viljandi to material related to
the founding period of other urban centres. From Estonian towns, we can
currently rely on analyses of material from coastal towns of north and
west Estonia, i.e. Tallinn, Haapsalu, Lihula and New Parnu (Mall &
Russow 2000; 2004; Russow 2006). Of the urban centres of south Estonia,
the authors are familiar with pottery finds from Tartu. Material
unquestionably dating from the first half of the 13th century has been
found from three centres: Lihula, Tallinn and Tartu. Comparison of the
earliest complexes shows that a common trait of all these settlements is
the existence of glazed redwares from southern Scandinavia and the Low
Countries, in addition to Paffrath-style globular pots and
proto-stoneware. In Lihula and Tallinn the redwares are more numerous,
while in the inland town of Tartu the number of such finds is somewhat
smaller. (6) As a hypothesis, it can be stated that a large amount of
fragments of glazed redwares in the early contexts of Estonian towns
might indicate a greater age of the settlement. As an example, among the
finds from excavations in Haapsalu and New Parnu, both of which emerged
during the second half of the 13th century, the share of glazed redwares
is almost non-existent. At the same time, both proto- and near-stoneware
can be identified among the finds, which allows to date the formation of
urban settlement in these centres to the last quarter of the 13th
century and the turn of the 13th and 14th centuries respectively, on the
basis of the different frequency of such finds (see more at Russow 2006,
especially Appendix 3). (7) Still, it should be stressed that without a
doubt, coastal towns had better opportunities for acquiring imported
ceramics than inland towns, so that the more fragile glazed redwares
could not be distributed in so large areas as the more easily
transportable early forms of stoneware.
The analysis of the pottery from Viljandi leads to parallels with
material from the west Estonian coastal towns, with certain
reservations. On the basis of archaeologically collected finds, glazed
redware jugs have not featured widely here either. Only two fragments
have been recognized so far among finds from the town area (8), so the
earliest "horizon of usage" of imported pottery from western
Europe consists of proto-stoneware which can be dated to the middle/or
third quarter/ of the 13th century (Fig. 5). These complexes were
unearthed in archaeological excavations in the surrounding of cinema
"Rubiin" (VM 11087) and in the yard of the Museum of Viljandi
(VM 10326; 10942). To be added are two fragments of proto-stoneware from
Munga Street, so it can be stated that all the oldest fragments of
imported wares originate from the northern part of the town, the only
exception being one sherd from St. John's churchyard. Thus, on the
basis of the imported wares, a more widespread habitation of the town
area took place at the end of the 13th century, as the distribution of
pottery from southern Lower Saxony, found from almost all archaeological
investigations in the area, seems to indicate (Fig. 6). As a preliminary
hypothesis, it can be stated that the diversity of imported vessels, or
more specifically, the number of ware types simultaneously in use in
Viljandi, grew remarkably by the end of the 13th century.
[FIGURE 5 OMITTED]
[FIGURE 6 OMITTED]
Habitation in the town area
Phase 1. The period before ca. 1250
Unlike the medieval castle of Viljandi, which was built on top of
the prehistoric hill fort, there exist no traces of a prehistoric
settlement site in the territory of the medieval town. The find of
several ard marks in the area around St. John's (Tvauri 2001b; Valk
2005, 99) have led to the suggestion that the territory of the medieval
town was used as a field during the prehistoric period (Valk 2005, 99).
As written sources state that the construction of the castle started in
1224 (Wartberge 1863, 4), this has usually been seen as a starting point
for activities in the territory of the medieval town (Alttoa 1978; Valk
2005, 102 f.; see also Johansen 1955, 119 ff. for a different view). The
existence of roads in the would-be town territory is fairly likely, but
as no archaeological traces have been found, we will not stress it
again. However, the existence of roads connecting the castle area to the
town territory would become a necessity once habitation had started in
the area.
Of the settlement traces investigated so far, only the earliest
layers in the area of the potter's workshop might be connected to
that stage. However, the fact that coins which could be dated to the
period before the 1260s, imported pottery more characteristic of the
early stages, and even the products of the potter's workshop which
were found in the burning chamber were all missing from the town
territory should lead to the conclusion that although habitation in the
town territory had started during the period in consideration, it has
not been intense, as finds from that period are not comparable to those
from other Estonian urban centres, as stated above.
The question whether the remains of daub from the area in front of
the castle (VM 10873: 3; Valk 2002) should indicate a house, which
originates from the same period, is disputable. On the one hand, the
earliest layers did not contain imported pottery, and the sherds of
local ceramics can be connected to the production of the potter's
workshop. On the other hand, the existence of such a house during the
third quarter of the 13th century is not excluded either.
Phase 2. Ca. 1250-1300
Whatever the situation was before ca. 1250, the town formation has
been rather intense since then. Pottery production in the potter's
workshop continued, as the layers that have deposited onto virgin soil
just next to it contain also fragments of imported pottery, which can be
dated to that period. If habitation had started in the area of the
market square even before 1250, it clearly saw an increase during the
period in question. This can be seen in the large share and variety of
imported wares (highly fired products of Siegburg and from Cologne
foothills, Paffrath-type wares, south Lower Saxony, etc.). It extended
toward Tartu Gate, but probably also the territory east of the market
square, as the excavations next to the town wall also yielded finds from
the same period (e.g. pottery decorated with wavy lines). The earliest
finds from the excavations at the north-west corner of the medieval town
(Fig. 2: 20) also included similar wares (Tvauri 1997, 83). In the block
of houses west of Lossi Street, and north of Vaike-Turu Street there
were areas that yielded similar finds, but the existence of drainage
ditches might show that this area was not suitable for early
habitation--it became possible once drainage was erected. It seems
likely that at first place, the inhabited areas were rather small and
these were few and far between (Fig. 7, dark shaded areas). The
difference between two investigated areas in the latter quarter: one
with early finds (Fig. 2: 14) and the other (Fig. 2: 17) without any
early finds despite larger investigated area may hint that not all plots
may have been inhabited simultaneously.
[FIGURE 7 OMITTED]
However, it still seems that the whole area surrounded with the
town wall had not been inhabited even by 1300. Investigations at the
town wall in the northeastern corner of the medieval town unearthed a
brown layer of soil, interpreted as the original soil of the area, which
did not contain any finds, but was heaped up against the town wall
(Tvauri 1999; 2001b, 100; 2010, 158 f.). The layers above it contained
finds from the 14th century (see Table 1). In addition to the border
areas, there exist no solid data of habitation in the southern part of
the town, except the territory of the potter's workshop, and the
house (?) remains mentioned in the last section. As seen in Fig. 7
(lightly shaded area), the area habited is oblong in east-west
direction, and should have considered the territory of the northern moat
as its natural border. It seems that the natural hollow in this area has
been a clear border to human activities at that time, as the
investigations north of the moat, in the territory of the medieval
suburb, have not produced contemporary finds. It seems that human
activities in the territory of the suburb started during the first half
of the 14th century (see also Haak 2006, 71).
According to the existing radiocarbon dating, collected from
charcoal below the town wall, the wall was built by the end of the 13th
century or shortly thereafter. (9) If the town wall was built in one
stage, as has been stated (Tvauri 2001b, 100, 105 f.), the area chosen
for the medieval town had been finally approved by that time. The
question whether the town wall was actually constructed all at once
remains, however, uncertain, as it has been located only on a few spots,
and charcoal could be collected only from under the western part of the
wall. However, stratification as recorded by Tvauri supports the view
that by the end of the 13th or early 14th century Viljandi was
surrounded with a town wall.
Phase 3. Ca. 1300-1350
It seems that a lack of space in the town area appeared only at the
beginning of the 14th century, now surrounded with the wall, so that
areas with less favourable conditions were adapted for living. This goes
for the area of Munga Street, as radiocarbon dating from a woven branch
fence of a drainage ditch (10) indicates. Approximately the same dating
can be given to the earliest find complex from the area (see Table 1).
The eastern part of Pikk Street may have been inhabited even
slightly later, as the finds from there include Siegburg stoneware. The
same goes for the most intensive layer around St. John's. However,
as the complex also includes finds of clearly 13th-century origin, it
cannot be excluded that some kind of a habitation had taken place in the
area before the intense layer, connected to the marketplace, had
deposited, and the finds were mixed up during the later stage.
The finds from the area north of the market square, next to the
Tartu Gate and at Munga Street also show signs of permanent habitation
during the period, but it seems that the activities of the potter's
workshop have ceased, and the territory has not experienced any imminent
restructuring. The finds from the layers just above the remains of the
stoves contain a few fragments of pottery from that period, but these
can be connected neither to any structures nor to layers which have
primarily deposited in that area.
The beginning of habitation in the territory of the northern suburb
should also be mentioned here, although the reasons for that are
uncertain. It seems likely that the suburban settlers were active in
trades considered dangerous or unacceptable within the town walls (e.g.
smithery, as the finds include remarkable amount of iron slag). Fire
hazard might also have been the reason for shutting down the
potter's workshop.
By mid-14th century, Viljandi had become a Livonian small town,
clearly distinguishable from nearby villages also on the basis of
material culture. It should be noted, however, that the inhabitants
still had land within the walls to use for gardening even at the
beginning of the 16th century (Freymann 1918. No. 3). In addition to
that, the townspeople owned arable land next to the town. So,
connections between townspeople and agriculture were very active in
small towns during the whole Medieval Period.
Discussion
As the present analysis has shown, the "early" material
of Viljandi seems to be pretty homogeneous. The stratigraphically
earliest contexts all include finds which arrived in Estonia only around
ca. 1250, and many of these also contain finds from the 14th century.
While comparing the two areas regarded as possibly the oldest: the
territory near the market square and that surrounding St. John's,
we should conclude that the contexts which have yielded 13th-century
finds, are much more intense near the market square. Thus, we cannot
support the idea of earliest settlement "core" just in front
of the castle. Moreover, finds with a roughly similar dating to those
from the market square have even been collected in the excavations just
next to the Tartu Gate. The finds from Munga Street are of slightly
later origin, as it seems likely that the complex formed around the year
1300.
The complex collected from the churchyard of St. John's is of
still later origin. On the basis of Siegburg stoneware, it cannot have
deposited earlier than during the second quarter of the 14th century.
Still, the interpretation of this complex is hindered by the presence of
a few sherds, dating from the 13th century. These, however, happen to be
the only traces of permanent activities during the 13th century.
Thus, on the basis of datings of imported pottery, we cannot
support the view that habitation in the territory of the medieval town
was first centred around the supposed market-place at the site of later
St. John's. On the basis of finds of imported ware, the situation
seems to have been quite the opposite: the largest share of
"early" finds (Paffrath-type ware, Siegburg proto-stoneware
and Cologne foothills, etc.) has been deposited in the blocks north of
the medieval market square. These originate from early contexts, while
many strata with 13th century finds could be distinguished in
stratigraphical sequence in the area.
The question of the dating of the potter's workshop is not
easily solved with the help of the few sherds of collected imported
wares. Still, it should be noted that the layer just above the burning
chamber of one of the ovens also contained fragments of imported wares,
which can be dated to the mid-14th century at the earliest. In that
sense, we should also consider the possibility that the potter's
workshop (though perhaps not that particular stove) also functioned at
the beginning of the 14th century. At the very least, wares very similar
to those collected from the waste of the workshop are present in the
deposits of early 14th century.
From outside the medieval town, no sherds of imported wares earlier
than 14th-century origin have come to light. Although finds of pottery
with wavy and straight lines has been collected from these areas, it
should still show that the suburbs in the true sense of the word started
to emerge no sooner than during the 14th century.
If we look at the scarcity of settlement remains that could be
dated before 1250, naturally the question arises whether a settlement in
the town area existed at all, which could have been granted town
privileges. Arguments supporting this idea have usually been indirect
ones. (11) However, as historian Tiina Kala has recently questioned the
early dating of Tallinn as a town (Kala 2002, 401 ff.), we can note that
there is no "must" in Viljandi being granted the privileges
before 1283: all earlier dates are of hypothetical nature.
Interpretation of the archaeological data regarding Viljandi in the
first half of the 13th century, would not favour a rapidly developing
urban centre.
A small comparison to other medieval centres of Estonia would
reveal a strikingly similar situation. There are a few written notices,
which have been used to show a quick formation of medieval centres,
which cannot be actually seen in the collected archaeological material.
It seems that such a possibility has so far been a priori excluded:
either by dating some (in our opinion) not so precisely datable finds or
strata to the "intermediate" period (e.g. Maesalu & Vissak
2002, 154 f.), or assigning the earliest settlement core to areas not
investigated archaeologically (cf. Russow 2006, 164). Although neither
of these possibilities can be denied without a good argument12, the
possibility that the formation of centres of medieval type took somewhat
longer also needs consideration. So far, finds of "medieval"
type which can definitely be dated to the first half of the 13th century
have been obtained only from Toompea in Tallinn, Lihula and Tartu. If we
include castles, we should add Otepaa and probably also others, but the
fact remains that it took some time for the newcomers as well to form
the "medieval" structures we mostly know of later sources.
The hypothesis of the distribution of imported against local
pottery as an ethnic indicator is intriguing, but should, in our
opinion, by no means be overestimated. The neighbourhood of a
potter's workshop is a much more likely reason for the abundance of
local pottery, including wasters. This is not meant to diminish the role
of local population in the town formation process: there are several
indicators for that (e.g. the abandoning of settlement sites in the
close vicinity of the town, the presence of pottery with decoration
similar to the Final Iron Age, as well as ornaments, etc.; remains of
clay daub in the medieval town area). The question of ethnicity in a
Livonian town of the medieval period is a most interesting one, but
should, in our opinion, be examined in a more complex manner, which
would exceed the limits of the current article.
Finally, to return to the question whether Viljandi was a typical
"founded" town, we should return to the question at what time
we can see evidence for the founding of the town. If we see the
foundation in connection with the castle, it should be remembered that
the convent house was constructed most likely at the beginning of the
14th century (Alttoa 2003, 104). The existence of large-scale
fortifications in the outer baileys during the 13th century can also be
doubtful. As stated above, the dating of the town wall is not much
earlier. However, the fact that there has been only one spot near the
medieval town, where 13th-century settlement traces have been found,
leads to the conclusion that the territory where settlers could start
their activities was strictly defined. Whether it was natural causes,
land ownership, or any other reason, we cannot determine, but all the
existing signs lead to that conclusion. However, the planning of a town
and castle complex in the middle of the 13th century, just after the
conquest, would seem a bit too optimistic. However, a monumental project
by the Livonian Order, which was never realized (Valk 2005, 106) is a
reasonable hypothesis. We would still question the dating of this
project. Almost unanimously, the "plan for Viljandi" has been
dated to the period just after the German conquest. Although the
possibility that the plan was still up to date half a century later
cannot be excluded, it is just as possible that it was only drawn
considering the existing situation at the end of the 13th century, when
Viljandi had already become a place of importance. The fact that many
remaining features of Viljandi (Convent House, town wall, permanent
settlement of several areas, several street pavements) can be dated to
that period, should indicate that the development of that period
followed an existing plan. The question remains whether this was the
first and original plan, or whether it had been redrawn.
Conclusions
In the period of increasing understanding of the interpretive
nature of the whole archaeological process, the aim of the current
article is by no means to state that the question of the town
development process of Viljandi has finally been solved. Still, we
believe that the arguments presented so far are convincing enough to
show that the development of the medieval town of Viljandi cannot be
seen simply as an expansion from the castle northwards. Although the
links between the castle and the town are clearly visible, it seems
likely that the cradle of the medieval town was situated either at the
marketplace or at the potter's workshop, in other words, it needs
trade and handicraft to "make" a medieval town. In addition,
we cannot outline a clearly bordered "cradle area", rather
there were several "seeds" that were later shaped to the
bordered pattern that we know from late 16th-century documents and the
17th-century town plan.
The relations between seemingly totally irrelevant events and
circumstances make the investigative process quite exciting and allow
new conclusions to be made. Fortunately, there has been a lot of
creative thinking in earlier works on the town development in Viljandi.
The current article cannot manage all that, but hopefully it has added
the dating of imported ceramics, and the reconsideration of the
"contemporary trends" of town formation in the 13th century to
the pack of variables which have to be taken into account while
discussing urbanism of the 13th century in Livonia. What is more, it is
not only new archaeological investigations which change the existing
concepts. Although it is often new excavations that offer valuable
insight into the archaeological sources investigated now and then, there
are only too many archaeological investigations in medieval Livonia,
which have stopped together with the end of fieldwork. The more
comparative material we have, the better is our understanding of the
trends and peculiarities, and finally, the more we find problems still
without a reasonable solution.
There was not much chance to demonstrate the importance of complex
usage of all available sources, but hopefully it was clear enough that
the possibilities of a reliable interpretation of archaeological finds
were greatly diminished by the lack of contemporary written data. An
analysis of the situation in either Tallinn or Riga, with an attempt to
use all kinds of available sources, would thus be most interesting.
doi: 10.3176/arch.2013.1.04
Acknowledgements
We would like to thank archaeologist Eero Heinloo (AEG NPO) for
showing us the finds from his excavations at New Parnu, Herki Helves
(Museum of Viljandi) for his help in photography of the finds, and
Kristiina Zadin (Tartu City Museum) for her help with figure 7. Special
thanks to the reviewers for their useful comments. The research was
supported by the Estonian Ministry of Education and Research target
financed research project SF0130054s12, and ESF grant ID 134.
References
Alttoa, K. 1977. Viljandi vanalinna kaitse- ja ehitustegevuse
reguleerimistsoonide detailplaneerimine. Uurimistood. Koide I.
Ajalooline oiend. Tallinn. Manuscript in the Archives of the National
Board of Antiquities.
Alttoa, K 1978. Viljandi linna kujunemisest.--Ehitus ja
Arhitektuur, 2, 48-54.
Alttoa, K 1995. Louna-Eesti arhitektuur 15. saj.--16. saj. I
poolel. Unpublished master's thesis. Manuscript in the Library of
the University of Tartu.
Alttoa, K 1999. Viljandi linn 1599. aastal. Kommentaarid.--Viljandi
Muuseumi aastaraamat, 1998. Viljandi, 154-162.
Alttoa, K. 2003. Viljandi ordulinnus. Uurimisseis ja
probleemid.--Viljandi Muuseumi aastaraamat, 2002. Viljandi, 92-108.
Amelung, Fr. 1898. Geschichte der Stadt und Landschaft Fellin von
1210 bis 1625.--Jahresbericht der Felliner litterarischen Gesellschaft
fur die Jahre 1890 bis 1895. Fellin, Beilage, 1-232.
Angermann, N. 2001. Das Hamburgische Recht in Nordosteuropa.--Die
Stadt im europaischen Nordosten. Hrsg. R. Schweitzer, W. Bastman-Buhner
& J. Hackmann. Aue-Stiftung, Helsinki, Lubeck, 65-75.
Arman, H. 1965. Linnaehitus.--Eesti arhitektuuri ajalugu. Ed. H.
Arman. Eesti Raamat, Tallinn, 31-51.
Bergengrun, A. 1897. Beitrag zur Geschichte Fellins a. d. J.
1592.--Sitzungsberichte der Gesellschaft fur Geschichte und
Alterthumskunde der Ostseeprovinzen Russlands aus dem Jahr 1896. Riga,
37-40.
Bronk Ramsey, C. 2009. Bayesian analysis of radiocarbon
dates.--Radiocarbon, 51: 1, 337-360.
Freymann, G. v. 1918. Urkunden und Aktenstucke aus dem Fellinschen
Stadtarchiv 1481-1783.--Jahresbericht der Felliner litterarischen
Gesellschaft fur die Jahre 1912 bis 1917. Fellin, 1-400.
Haak, A. 2003. Keskaegse Viljandi tanavate vork.--Viljandi Muuseumi
aastaraamat, 2002. Viljandi, 71-91.
Haak, A. 2004. Archaeological investigations at Viljandi Castle of
the Teutonic Order and in Medieval Viljandi.--AVE, 2003, 107-121.
Haak, A. 2005. Viljandi linna kujunemisest peamiselt
arheoloogiliste allikate pohjal.--Konverentsikogumik "Linna
asutamine, esmamainimine, inimtegevuse jaljed". Narva linnuses, 7.
septembril 2005. (Narva Muuseumi toimetised, 5.) Narva, 17-28.
Haak, A. 2006. Tartu varava eeslinna tekkest, havingust ning
taaskujunemisest. Uusi andmeid arheoloogilistelt kaevamistelt
1996-2005.--Viljandi Muuseumi aastaraamat, 2005. Viljandi, 68-87.
Haak, A. & Rannamae, E. in print. Tracing the castle crew.
(Zoo)archaeological search for the inhabitants of Viljandi Castle (south
Estonia) in the late 13th century.--The Castle as Social Space. Ed. K.
Predovnik. (Castrum Bene, 12.) Ljubljana.
Heinloo, E. & Vissak, R. 2012. Investigations on the territory
of the medieval New-Parnu.--AVE, 2011, 151-166.
IK 1991 = Finds from the investigations at Pikk Street (1991).
Stored in the collections of Kabinet of Archaeology, University of
Tartu.
Jakubowski, J. & Kordzikowski, J. 1915. Inflanty I. Polska XVI
Wieku, XIII.--Zrodla Dziejowe. Tom XXIV, Cze^c I-sza. Warszawa.
Johansen, P. 1938. Paide linna asutamisest.--Opetatud Eesti Seltsi
Toimetused, XXX. Liber Saecularis. Tartu, 201-208.
Johansen, P. 1955. Lippstadt, Freckenhorst und Fellin in
Livland.--Westfalen, Hanse, Ostseeraum. Veroffentlichung des
Provinzialinstituts fur westfalische Landes- und Volkskunde. Reihe I.
Heft 7. Aschendorffsche Verlagsbuch-handlung, Munster, 95-160.
Johansen, P. 2005 (1941). Hansa tahendus Liivimaale.--Paul
Johansen. Kaugete aegade sara. Comp. J. Kivimae. Ilmamaa, Tartu,
355-390.
Kala, T. 2002. Tallinna tekkeloo peegeldumine kirjalikes
allikates.--Keskus--tagamaa--aareala. Uurimusi asustushierarhia ja
voimukeskuste kujunemisest Eestis. Ed. V. Lang. (MT, 11.) Tallinn,
Tartu, 391-408.
Kelch, C. 2004. Liivimaa ajalugu. Eesti Ajalooarhiiv, Tartu.
Kenkmaa, R. 1937. Eesti keskaegsete linnade tekkimine ja arenemine
XIII ja XIV sajandil.--Eesti ajalugu, II. Ed. H. Kruus. Eesti Kirjanduse
Selts, Tartu, 83-119.
Kildyushevskii, V. I. 2006. Pskov pottery in the 12th to 16th
centuries.--The Pottery from Medieval Novgorod and its Region. The
Archaeology of Medieval Novgorod. Ed. C. Orton. University College
London, London, New York, 79-115.
Kreem, J. 2002. The Town and its Lord. Reval and the Teutonic Order
(in the Fifteenth Century). (Tallinna Linnaarhiivi toimetised, 6.) Ilo,
Tallinn.
Kriiska, A., Haak, A. & Lohmus, M. 2007. Arheoloogilised
valitood Viljandi linnas Tallinna ja Oru tanava vahele rajatud
kaugkuttetorustiku kraavi alal.--Viljandi Muuseumi aastaraamat, 2006.
Viljandi, 101-129.
Lang, V. 2004. Varalinnalised keskused (aolinnad) hilismuinasaegses
Eestis.--Kui vana on Tallinn? Ed. T. Kala. (Tallinna Linnaarhiivi
toimetised, 8.) Ilo, Tallinn, 7-27.
Leighly, J. 1939. The towns of Medieval Livonia.--University of
California Publications in Geography, 6: 7, 235-314.
LGU = Livlandische Guterurkunden. Bd. I (1207-1500). Hrsg. von H.
v. Bruiningk und N. Busch. Jonck & Poliewsky, Riga, 1908.
Lowis of Menar, K v. 1922. Burgenlexicon fur Alt-Livland. Riga.
Maesalu, A. & Vissak, R. 2002. On the older topography of
Tartu.--The Medieval Town in the Baltic: Hanseatic History and
Archaeology, II. Eds R. Vissak & A. Maesalu. Tartu Linnamuuseum,
Tartu, 145-163.
Mall, J. 2004. Arheoloogilise kultuurkihi spetsiifikast Tallinna
vanalinna territooriumil. Linnusest ja linnast. Uurimusi Vilma Trummali
auks. Eds A. Haak, E. Russow & A. Tvauri. (MT, 14.) Tallinn, Tartu,
249-268.
Mall, J. & Russow, E. 2000. Kohalik ja importkeraamika
Tallinnas aastail 1200-1550.--EAA, 4: 2, 120-128.
Mall, J. & Russow, E. 2004. Archaologie und Stalinismus--die
Anfange der Stadtarchaologie in Estland und die Ausgrabungen auf dem
Tallinner Rathausplatz 1953.--Zeitschrift fur Archaologie des
Mittelalters, 2003, 31, 145-158.
Neumann, W. 1911. Der Stadtplan als geschichtliche
Urkunde.--Mitteilungen aus der Livlandischen Geschichte, XXI. Riga,
84-99.
Poltsam-Jurjo, I. 2009. Liivimaa vaikelinn Uus-Parnu 16. sajandi
esimesel poolel. Argo, Tallinn. Pullat, R., Hein, A., Leimus, I. &
Viires, A. (comp.) 2006. Johann Christoph Brotze. Estonica, Tallinn.
Raadik, E. 1960. Viljandi arhitektuuri ajalugu feodalismi perioodil
(XIII-XIX sajandi keskpaigani). Unpublished diploma thesis. Tartu.
Manuscript in the Department of History, University of Tartu.
Reimer, P. J., Baillie, M. G. L., Bard, E., Bayliss, A., Beck, J.
W., Blackwell, P. G., Bronk Ramsey, C., Buck, C. E., Burr, G. S.,
Edwards, R. L., Friedrich, M., Grootes, P. M., Guilderson, T. P.,
Hajdas, I., Heaton, T. J., Hogg, A. G., Hughen, K. A., Kaiser, K. F.,
Kromer, B., McCormac, F. G., Manning, S. W., Reimer, R. W., Richards, D.
A., Southon, J. R., Talamo, S., Turney, C. S. M., van der Plicht, J.
& Weyhenmeyer, C. E. 2009. IntCal09 and Marine09 radiocarbon age
calibration curves, 0-50,000 years cal BP.--Radiocarbon, 51: 4,
1111-1150.
Renner, J. 1995. Liivimaa ajalugu 1556-1561. Olion, Tallinn.
Russow, B. 1993. Liivimaa kroonika. Tallinn.
Russow, E. 2006. Importkeraamika Laane-Eesti linnades 13.-17.
sajandil. Tallinna Ulikooli Ajaloo Instituut, Tallinn.
Russow, E., Valk, H., Haak, A., Parn, A. & Maesalu, A. 2006.
Medieval archaeology of "European" context: towns, churches,
monasteries and castles.--Archaeological Research in Estonia 1865-2005.
Eds V. Lang & M. Laneman. (Estonian Archaeology, 1.) Tartu, 159-192.
Selirand, U. 1981. Ausgrabungen auf dem Territorium des
Franziskanerklosters in Viljandi.--TATU, 30: 4, 423-428.
Selirand, U. 1982a. Viljandi, Kauba tn. 12 Riia varava 1981. a.
valiuurimiste aruanne. 1. koide. Leiuainese analuus. Tallinn. Manuscript
in the Archives of the National Board of Antiquities.
Selirand, U. 1982b. Uber die Untersuchungen des
Franziskanerklosters in Viljandi.--TATU, 31: 4, 398-401.
Tuulse, A. 1942. Die Burgen in Estland und Lettland. (Opetatud
Eesti Seltsi Toimetused, XXXIII.) Tartu.
Tvauri, A. 1997. Archaeological excavations in the old part of
Viljandi.--AVE, 1997, 81-86.
Tvauri, A. 1999. Aruanne Viljandi linnamuuri arheoloogilistest
uuringutest kvartali nr 162 alal 1999. aastal. Tartu. Manuscript in the
Archives of the National Board of Antiquities.
Tvauri, A. 2000a. Pihkva pottsepad Viljandis ja Tartus.--EAA, 4: 1,
21-30.
Tvauri, A. 2000b. Loode-Vene paritolu slaavi keraamika Eestis
11.-16. sajandil.--EAA, 4: 2, 91-119.
Tvauri, A. 2001a. Archaeological investigations in the old town of
Viljandi, at 4 Pikk Street.--AVE, 2000, 103-107.
Tvauri, A. 2001b. Viljandi linnamuuri arheoloogilised uuringud
aastatel 1997-1999.--Viljandi Muuseumi aastaraamat, 2000. Viljandi,
92-110.
Tvauri, A. 2004. Liivi soja aegne Vene keraamika Eesti linnustes ja
linnades.--Linnusest ja linnast. Uurimusi Vilma Trummali auks. Eds A.
Haak, E. Russow & A. Tvauri. (MT, 14.) Tallinn, Tartu, 395-409.
Tvauri, A 2005. Eesti hilisrauaaja savinoud (11. sajandist 13.
sajandi keskpaigani). (MT, 16.) Tartu, Tallinn.
Tvauri, A. 2010. Archaeological excavations at Lossi 21,
Viljandi.--AVE, 2009, 157-163.
Valk, H. 1993. About the role of the German castle at the
town-genesis process in Estonia: the example of Viljandi.--Castella
Maris Baltici, 1. Ed. K. Drake. Almqvist & Wiksell International,
Stockholm, 219-223.
Valk, H. 1995. Keskaegse Viljandi sunni- ja arenguloost.--Opetatud
Eesti Seltsi aastaraamat, 1988-1993. Tartu, 173-174.
Valk, H. 2002. Aruanne arheoloogilistest fikseerimistoodest
Viljandis Pikk tn 15 ja 17 veetrassil 05. juunil 2002. a. Manuscript in
the Archives of the National Board of Antiquities.
Valk, H. 2005. The genesis of Viljandi (Fellin): archaeological
data.--Riga und der Ostseeraum.
Von der Grundung 1201 bis in die Fruhe Neuzeit. Eds I. Misans &
H. Wernicke. (Tagungen zur Ostmitteleuropa-forschung, 22.) Marburg,
95-107.
Viljandi linn 1599. aastal. Translation from Old Polish by K.
Vabamae.--Viljandi Muuseumi aastaraamat, 1998. Viljandi, 114-153.
Wartberge, H. 1863. Hermanni de Wartberge Chronicon Livoniae.
Scriptores rerum Prussicarum, Bd. 2. Leipzig.
Westren-Doll, A. 1929. Burg und Stadt Fellin zu polnischer
Zeit.--Sitzungsberichte der Gelehrten Estnischen Gesellschaft, 1928.
Dorpat, 67-81.
(1) Some of the ideas presented here were first introduced in a
presentation given by Arvi Haak at a seminar on town formation in Narva
(7 September 2005) and published as Haak 2005. The current paper is a
significantly extended version on that issue.
(2) The first to point out this difference of numbers of surviving
houses was art historian Kaur Alttoa (1977, 46).
(3) The medieval town and parish church of Viljandi, situated next
to the market place, was dedicated to St. John and St. Clara (Viljandi
linn 1998, 114).
(4) Earlier authors have proposed 1680s (Alttoa 1999, 156) or even
the year 1688 (Westren-Doll 1929, 74) as the year when the map was
compiled. As the map itself has not been dated, the current dating is
based on the inventory of the Swedish War Archives (Krigsarkivet;
http://www.ra.se/KRA/0406h.html#Fellin). Of course, this dating is
arbitrary as well.
(5) Although suggestions have been made of brick manufacture in
Estonia before the 13th-century conquest, these are not based on solid
data and can be dropped. However, although a few earlier examples exist,
the appearance of iron nails in greater quantities seems to be a
phenomenon that starts in Estonia with the Middle Ages.
(6) This is currently only a subjective assessment based on finds
from the intramural area and suburbs of Tartu. It cannot be excluded
that a statistical analysis could show that a similar proportion of
glazed redwares was used in Tartu as in the coastal towns.
(7) This view might change soon as the finds from recent
excavations (2011) in New Parnu include a higher proportion of
proto-stonewares as the earlier investigations, which tended to
concentrate on the southern part of the town, i.e. not on the initial
core area of the urban settlement. See also Heinloo & Vissak 2012,
especially fig. 8.
(8) These finds originate from mixed contexts and therefore are not
shown in Table 1. In addition to that, archaeological excavations in the
Castle of the Livonian Order have unearthed ca. ten fragments of glazed
redwares.
(9) TA-2615, 740 [+ or -] 30 BP, calibrated with program OxCal
v4.1.7 (Bronk Ramsey 2009); 1257-1286 cal. A.D. with 68.2% confidence
(atmospheric data from Reimer et al. 2009). See also Tvauri 2001b for
results of archaeological investigations of the town wall of Viljandi.
(10) TA-2216, 620 [+ or -] 40 BP, calibrated with program OxCal
v4.1.7 (Bronk Ramsey 2009); 1297-1325; 1344-1394 cal. A.D. with 68.2%
confidence (atmospheric data from Reimer et al. 2009).
(11) R. Kenkmaa (1937, 89 ff.) took it as a prerogative that the
inhabitants of the castle needed various services, thus there was a need
for several activities typical of a medieval town. Thus, the town should
have emerged rather quickly. K. Alttoa (1978, 53 f.) compared Viljandi
to Paide, where the temporal distance between the founding of the
castle, and the acquiring of the town privileges was ca. 25 years. On
the special situation of Paide, see also Johansen 1938.
(12) See Mall 2004 for a hypothesis for the missing of finds from
the first half of the 13th century from most contexts in Tallinn lower
town.
Arvi Haak, Institute of History, Tallinn University, 6 Ruutli St.,
10130 Tallinn/Tartu City Museum, 23 Narva Road, 51009 Tartu, Estonia;
arvi.haak@katarina.ee
Erki Russow, Institute of History, Tallinn University, 6 Ruutli
St., 10130 Tallinn, Estonia; erki.russow@tlu.ee
Table 1. Finds from earliest archaeological contexts of Viljandi
No. Excavation Collection Investigator Total number
number of potsherds
from 13th-
early 14th-
century
contexts
1. Munga Street (1989) VM 10235 Valk 69
2. St. John's VM 10258 Valk 846
churchyard (1990)
3. Pikk Street 4 and 6 IK 1991 Valk 3054
(Inkeri Keskus;
1991)
4. Laidoneri Square 10 VM 10326 Valk 66
(1991)
5. Tartu gate (1992) VM 10322, Valk 21
10323
6. Laidoneri Square 10/ VM 10536 Kodar 24
Lossi Street 14
(1993)
7. Laidoneri Square 10 VM 10942 Kodar 151
(1994-1995)
8. Vaike-Turu Street 6 VM 10533 Valk 61
(1996)
9. Pikk Street 4 (1998) VM 10620 Tvauri Ca. 5000
10. Town Wall (1999) VM 10830 Tvauri 353
11. Pikk Street (2000) VM 10810 Haak 44
12. Pikk Street 13 VM 10873 Valk 22
(2002)
13. Eastern part of Pikk VM 10932 Haak& 17
Street (2002) Tvauri
14. Vaike-Turu and Oru VM 11087 Kriiska 32
Streets (2004) etal.
15. Eastern part of Pikk VM 11001 Veidi 381
Street (2004)
16. Hariduse Street 3A VM 11120 Haak 10
(2006)
17. Lossi Street 21 VM 11272 Tvauri 14
(2009)
No. Excavation Total % of imported Siegburg
number of wares from proto-
imported the pottery stoneware
wares assemblage
1. Munga Street (1989) 25 36.2 2
2. St. John's 39 4.6 1
churchyard (1990)
3. Pikk Street 4 and 6 4 0.1
(Inkeri Keskus;
1991)
4. Laidoneri Square 10 20 30.3 1
(1991)
5. Tartu gate (1992) 4 19.0
6. Laidoneri Square 10/ 21 87.5 2
Lossi Street 14
(1993)
7. Laidoneri Square 10 25 16.6 7
(1994-1995)
8. Vaike-Turu Street 6 11 18.0
(1996)
9. Pikk Street 4 (1998) 0 0.0
10. Town Wall (1999) 5 1.4
11. Pikk Street (2000) 5 11.4
12. Pikk Street 13 0 0.0
(2002)
13. Eastern part of Pikk 1 5.9
Street (2002)
14. Vaike-Turu and Oru 15 46.9 8
Streets (2004)
15. Eastern part of Pikk 28 7.3
Street (2004)
16. Hariduse Street 3A 0 0.0
(2006)
17. Lossi Street 21 8 57.1 0
(2009)
No. Excavation Siegburg Mid-rhenish Paffrath
near near ware
stoneware stoneware
1. Munga Street (1989) 1 1
2. St. John's
churchyard (1990)
3. Pikk Street 4 and 6
(Inkeri Keskus;
1991)
4. Laidoneri Square 10 3 4
(1991)
5. Tartu gate (1992) 1 1
6. Laidoneri Square 10/ 5 1
Lossi Street 14
(1993)
7. Laidoneri Square 10 10 2
(1994-1995)
8. Vaike-Turu Street 6
(1996)
9. Pikk Street 4 (1998)
10. Town Wall (1999)
11. Pikk Street (2000)
12. Pikk Street 13
(2002)
13. Eastern part of Pikk
Street (2002)
14. Vaike-Turu and Oru 1 1
Streets (2004)
15. Eastern part of Pikk
Street (2004)
16. Hariduse Street 3A
(2006)
17. Lossi Street 21 0 0 0
(2009)
No. Excavation Southern Greywares Siegburg
Lower stoneware,
Saxony 1st half
of 14th c.
1. Munga Street (1989) 11 6
2. St. John's 6 2
churchyard (1990)
3. Pikk Street 4 and 6 2
(Inkeri Keskus;
1991)
4. Laidoneri Square 10 6 3
(1991)
5. Tartu gate (1992) 2
6. Laidoneri Square 10/ 3
Lossi Street 14
(1993)
7. Laidoneri Square 10 1 3 1
(1994-1995)
8. Vaike-Turu Street 6 5 2 1
(1996)
9. Pikk Street 4 (1998) Ca. 5000
10. Town Wall (1999)
11. Pikk Street (2000) 2
12. Pikk Street 13
(2002)
13. Eastern part of Pikk
Street (2002)
14. Vaike-Turu and Oru 1 2
Streets (2004)
15. Eastern part of Pikk 5
Street (2004)
16. Hariduse Street 3A
(2006)
17. Lossi Street 21 5 0 1
(2009)
No. Excavation Local Incl. % of Wavy
wares rim rim orn.
sherds sherds
1. Munga Street (1989) 44 5 11.4 2
2. St. John's 807 153 19.0 9
churchyard (1990)
3. Pikk Street 4 and 6 3050 384 12.6 223
(Inkeri Keskus;
1991)
4. Laidoneri Square 10 46 5 10.9 3
(1991)
5. Tartu gate (1992) 16 4 25.0 2
6. Laidoneri Square 10/ 3 2 66.7 1
Lossi Street 14
(1993)
7. Laidoneri Square 10 126 21 16.7 11
(1994-1995)
8. Vaike-Turu Street 6 50 5 10.0 1
(1996)
9. Pikk Street 4 (1998) 326 6.5 125
10. Town Wall (1999) 348 40 11.5
11. Pikk Street (2000) 39 1 2.6
12. Pikk Street 13 22 4 18.2 2
(2002)
13. Eastern part of Pikk 16 1 6.3
Street (2002)
14. Vaike-Turu and Oru 17 0.0 3
Streets (2004)
15. Eastern part of Pikk 353 51 14.4
Street (2004)
16. Hariduse Street 3A 10 1 10.0
(2006)
17. Lossi Street 21 6 2 33.3
(2009)
No. Excavation Line Wavy and % ornamented
orn. line orn. from all
sherds
1. Munga Street (1989) 2 3 15.9
2. St. John's 42 4 6.8
churchyard (1990)
3. Pikk Street 4 and 6 268 163 21.4
(Inkeri Keskus;
1991)
4. Laidoneri Square 10 7 3 28.3
(1991)
5. Tartu gate (1992) 1 2 31.3
6. Laidoneri Square 10/ 1 66.7
Lossi Street 14
(1993)
7. Laidoneri Square 10 11 3 19.8
(1994-1995)
8. Vaike-Turu Street 6 3 1 10.0
(1996)
9. Pikk Street 4 (1998) 489 35 13.0
10. Town Wall (1999) 0.0
11. Pikk Street (2000) 8 20.5
12. Pikk Street 13 2 3 31.8
(2002)
13. Eastern part of Pikk 1 6.3
Street (2002)
14. Vaike-Turu and Oru 2 29.4
Streets (2004)
15. Eastern part of Pikk 0.0
Street (2004)
16. Hariduse Street 3A 0.0
(2006)
17. Lossi Street 21 0.0
(2009)
