Archaeological Questions
Status of Absolute Chronology for Austria’s Pre- and Protohistory
Since the rise of the 14C method, the absolute chronology of European Prehistory changed dramatically. Prior to 14C dating, the beginning of the Neolithics in Central-Europe was dated at about 4000 BC. Afterwards it changed to 4800. Later, by using the tree-ring calibration, the beginning moved back to 5600 BC.
The same occurred with other periods, for example the beginning of the Bronze Age. Originally it was believed to date at about 1800 BC, later at about 2000 BC, but now at about 2200 BC.
The Linear Ceramics is an early culture which appears at the beginning of the Neolithics in larger parts of Europe. New 14C dates for this culture in Austria, from the sites of Brunn/Wolfholz, Rosenburg and Asparn/Schletz, inspired us to re-evaluate the absolute chronology of Linear Ceramics (Lenneis et al., 1996). In this paper the temporal distribution of the new dates was discussed, however, without the aid of group calibration. In the paper "Contribution to Absolute Chronology of the Neolithics and Bronze Age in Eastern Austria" (Stadler, 1995), the duration of Linear Ceramics in Austria was compared with that of the settlement of Brunn/Wolfholz and the Linear Ceramics in all of Europe. In another paper (Stadler and Lenneis, 1996), the possibility of separating the chronological phases of Linear Ceramics - also with 14C dates - was investigated, with partial success (mainly for the new Austrian dates). This had been tried before by Breunig (Breunig, 1985), however, without success. In addition, different factors influencing 14C dates were examined (e. g. old wood problem vs. short-lived material).
For other prehistoric cultures and groups an absolute chronological framework is being developed. The starting point is a database for 14C dates of Pre- and Protohistory in Europe, created by Stadler (Stadler, 1995). Data were taken from the big monographs containing large data sets for the last 20 years (Pape, 1979; Breunig, 1987; Ehrich, 1992; etc.). Data for the last few years, which were not yet available for the big monographs, were extracted from more recent papers. The database is not yet complete - it probably will never be - but it comprises at this time more than 13000 dates. This is considerably more than the 2000 dates in the book of Breunig (Breunig, 1987). In our project the database will be updated continuously employing part-time students of prehistory.
In near future the database will be made accessible through Internet. Users will also get information about results of data analysis, such as group calibrations of the many different culture groups (> 400) for the whole prehistoric period in Europe. Scientists who enter additional data not yet contained in the database for a particular group will get access to the whole data set belonging to this group.
Dendrochronology is an alternative method for establishing more exact absolute chronological frameworks. However, due to the lack of continuous tree-ring series for the different regions in Austria, it is currently less useful than 14C dating. The number of datable objects for dendrochronology is very limited due to the bad preservation of wooden objects in usual archaeological contexts. Exceptions are for example the lakeside settlements of the Mondsee Culture and the excavations in the salt mine of Hallstatt. The latter findings are particularly important since 14C dating for the Hallstatt Culture is virtually impossible due to an increase of atmospheric 14C going back in time during this period, which tends to cancel the loss due to radioactive decay. As a result, the relevant section of the 14C calibration curve runs almost horizontal in Fig. 6.
The nearly horizontal trend of the curve between 780 BC and 410 BC makes it virtually impossible to assign a specific date from radiocarbon measurements.
Among a variety of selection criteria for samples, two are particularly important: First, the samples should have a good archaeological context, which means that their position in the context should be known and the stratigraphy should be clear. Second, these strata or contexts should also contain well-defined and archaeologically datable remains. Only in special cases contexts without a good archaeological datation should be dated with 14C (e. g. graves without grave-goods)
Because of the shape of the calibration curve (see Figs. 1, 2, 6) not all segments are equally well suited for 14C dating. Nevertheless, in a first explorative phase of this project, a large increase of the 14C data base from Palaeolithics until the Middle Ages is expected. At present, the number of 14C dates from Austria amounts to 447 (Stadler, 1995). With this project the number should increase substantially.
After this exploration, samples will be taken from archaeological contexts belonging to "secure" segments of the calibration curve, where it runs reasonably steep and unambiguous. At that time, the new VERA AMS system may be ready for high precision (<0.5%) measurements. Multiple measurements of samples from "secure" sectors of the calibration curve should be carried out, thus reducing the uncertainty of the dates to a minimum.
The following time segments are at our disposal: For the Mesolithics some parts are available which can be well dated with 14C. Also the beginning of Neolithics including the beginning of Linear Ceramics, the Painted Pottery Culture - with some gaps. The Late Neolithics from 2700 BC to the Early Bronze Age and till the Late Bronze Age with some gaps may also be a good period for 14C dating. As discussed above, 14C dating for the Hallstatt Culture from about 780 till 410 BC is not useful (Fig. 6). Dates from early La Tène may be well suited, and also from Late La Tène till the middle of the 3rd century AD - with gaps. Late Antiquity samples from about 350 till 440 AD can be dated exactly. The last phase of Longobard settlement in Pannonia till their withdrawal to Italy in 568 and the beginning Avar settlement in the Carpathian basin from 568 till 700 - the end of Middle Avar time - may be dated exactly with 14C (see Fig. 2). From the late Avar time maybe samples from about 770 to 830(?) can deliver certain results. From the end of the Early Middle Ages the period from 970 to 1040 should be dated exactly. Thus archaeological finds from the "foundation" of Austria (996) - in this year - 1996 - Austria celebrates its millennium - could be dated exactly. In the Middle Ages and later segments from 1160 to 1290, from 1400 to 1510 and from 1640 to 1660 could all yield good dates.
Table 1 presents the periods, where 14C dating is useful.
Some projects related to 14C dating in the last decade concentrated on the dating of short-lived samples, such as burnt seeds or similar material. In this project we want to analyse also these samples, but in archaeological reality these are rare. Much more frequent are samples of charcoal or human and animal bones.
With charcoal there is the problem, because it may originate from wood which was already old when it was burnt, and thus deliver results which are too old. But with the determination of the wood species and with large series it should be possible to eliminate these erroneous dates.
The dating of bones has other difficulties. With the analysis of (purified) collagen of the bone one should be able to date the time period when the individual lived. Dating of bone collagen from long-lived species does not give the exact date of the death of the individual because of the relatively long turnover time of bone collagen (i.e. up to several tens of years for humans), which varies with the biological age of the individual.
The Museum of Natural History in Vienna owns in its Anthropological Department a large collection of human bones - about 35000 individuals - from the entire Pre- and Protohistory. Many of them come from regular excavations and have often good archaeological contexts and documentation. For the sample procurement it is necessary to have a good database of the human bones, i.e. samples from well documented and archaeologically well dated bones. Since this database does not exist yet, we plan to co-operate with Maria Teschler from the Anthropological Department to create with the help of her students such a database (part-time student employment).
The relatively young department of Archaeozoology in the Museum of Natural History in Vienna owns the largest collection of animal bones from the Pre- and Protohistory in Austria. So it would be possible to obtain a large number of well documented animal bone finds. But here we have the same problem as mentioned above. So a database for the animal bones must be created as well, together with Erich Pucher from the Archaeozoological Department and part-time employed students.
Samples will be collected in Austria mainly in the big institutions related to excavations: The Bundesdenkmalamt, the Departments of Prehistory, Anthropology and Archaeozoology of the Museum of Natural History, and the Institute of Prehistory of the University of Vienna.
Since we have good connections to scientists from the neighbouring countries, Czech Republic, Slovakia and Hungary, they will help us to collect a reasonable amount of samples from these areas. Table 2 presents a list of scientists, who want to contribute samples. Travels to or from these countries will be necessary and some work must be invested in the sample procurement and input into a database.
An exiting challenge for absolute dating would be the verification of relative chronologies. We propose to investigate at least two of them, both obtained with archaeological statistical methods.
The Early Bronze Age Cemetery from Franzhausen I, Lower Austria (2200-1500 BC)
So far, for the evaluation of the cemetery of "Franzhausen I" (Neugebauer, 1995) different mathematical methods were used. Besides seriation (in the form of reciprocal averaging) a newly developed method called "Analysis of the Next Neighbours" was applied (Stadler, 1996a; 1996b; 1996c).
This new method, - some kind of spatial analysis - was applied to the data from Franzhausen I. These data comprise archaeological finds, and situation of finds as well as anthropological data and various other data. The methodology starts with distribution maps of different types or features. Two varieties of this method were developed and are available in our program package SERION for the PC (DOS-Version).
Analysis of the Next Neighbours concerning one type: Distribution maps of one type are evaluated with statistical tests to find out whether the distributions show significant deviations from randomness. If the distributions are non-random, they can be used to form a matrix for relations between finds.
Analysis of the Next Neighbours concerning two types: Here the distributions of two types are compared with each other. Distributions showing significant similarity are used to form a matrix for relations between finds. With this step it is also possible to find graves where men and women belong to each other, even though they usually have few objects in common.
In both cases seriation is made with these matrices. Afterwards the eigenvectors of these are clustered in groups, which are then plotted on the map of the cemetery. These groups may be interpreted as a chronological sequence, even though some of them may also be contemporaneous.
Fig. 7 presents the result of the Analysis of the Next Neighbours concerning one type.
Both procedures, seriation and Analysis of the next Neighbours, result in relative chronologies. We propose to compare these chronologies with 14C dates from the same cemetery. In case of agreement, the relative chronologies could be calibrated to absolute ones with the 14C dates. So, with a small number of 14C-dated samples (50 to 100) a network of absolutely dated graves could be obtained.
Absolute Calibration of the Avar Age relative Chronology (568-ca.830 AD)
The seriation of more than 30000 Avar graves from the Carpathian basin yielded a useful relative chronological sequence. Fig. 8 displays the result of seriation (reciprocal averaging). The correctness of the sequence was confirmed with (only) 30 gold coins contained in these graves, but only for the Early Avar period, because later no gold coins were deposited in the graves (see Fig. 9) . More than that the sequence could also be calibrated with the help of these coins. 14C dating would increase the number of absolute dates substantially. Thus, the former calibration for the Early Avar period would be confirmed, completed and hopefully refined. The Middle Avar period - and maybe also some parts of the Late Avar period - would be calibrated for the first time. The last overview of these results was presented by Stadler (Stadler, 1996a).