Skip to content. | Skip to navigation

Personal tools

September 2019

Dating the spread of pottery among hunter-gatherer-fisher communities in north-eastern Europe


John Meadows, in cooperation with Carl Heron, Oliver Craig, Henny Piezonka, Olga Lozovskaya, Katia Dolbunova and many others

In the southwest Baltic, pottery first appeared in the mid-5th millennium cal BC, among Late Mesolithic hunter-gatherer-fisher communities. Although these groups had contact with pottery-making Neolithic farmers to the south, they appear to have adopted pottery production from further east. Pottery was made for thousands of years before the adoption of farming throughout the forest zone of north-eastern Europe, and in this region the term Neolithic denotes hunter-gatherer-fishers with pottery (Figure 1).


use of pottery

Figure 1: use of pottery by c.5500 cal BC, among farming societies (grey shading) and hunter-gatherer-fisher groups (coloured areas), based on legacy radiocarbon dates (H. Piezonka, 2015. Jäger, Fischer, Töpfer. Archäologie in Eurasien 30)


Although several attempts have been made to map the spread of pottery production between the Urals and the Baltic, existing radiocarbon data are deeply unsatisfactory, for various reasons:

  • A large proportion of legacy dates are old radiometric measurements with large uncertainties, on unsuitable sample types (e.g. bulk sediment, pottery total organic carbon)
  • The relationship between pottery and more reliable dating material (e.g. bone, wood, charcoal) is often unclear, and there has been no systematic attempt to integrate calibrated radiocarbon results with relative dating (stratigraphic and typological) (Figure 2)


calibrated radiocarbon dates

Figure 2: calibrated radiocarbon dates from samples all apparently associated with a short-lived hunter-gatherer-fisher site with early pottery (J. Meadows, after Karmanov et al). This is an unusual situation, which allows dates of contemporaneous materials to be compared; often legacy dates are all on one sample type, and/or come from longer-lived sites, and are thus even harder to interpret.


  • The most precise dates for early pottery are AMS radiocarbon ages of food-crust samples, which may be subject to unquantifiable freshwater reservoir effects, if fish was one of the ingredients cooked in the pot, making the dates appear spuriously old
  • There is a strong geographic pattern to the types of materials dated (Figure 3).


 Map of geographic distribution

Figure 3: Geographic distribution of >1200 legacy 14C dates apparently associated with early hunter-gatherer-fisher pottery, colour-coded by sample type. Blue: bone; black: charcoal; green: wood; brown: peat; grey: sediment; white: shell; red: pottery (J. Meadows).


In 2014, in collaboration with Olga Lozovskaya and the late Vladimir Lozovski, we began working on the Mesolithic-Neolithic Zamostje 2 site, north of Moscow (c.6500–4000 cal BC). Fishing was an essential part of the subsistence economy, as evidence by abundant fish remains, the wide range of artefacts used for fishing, and dietary stable isotopes on Mesolithic human remains (Meadows et al. 2019, Quaternary International., but herbivore bones are also common and wild plant foods were used regularly.  The Lozovskis controversially argued that pottery was not used at Zamostje 2 until the second quarter of the 6th millennium, whereas similar ceramics were dated elsewhere to 6000 cal BC or earlier.
Our current chronological model, combining existing AMS and radiometric dates on a wide range of sample types with the stratigraphic sequence and 25 new AMS food-crust dates (from both the early and middle Neolithic phases), places the adoption of pottery at Zamostje in the 57th century (soon after 5700 cal BC). We initially used EA-IRMS to distinguish food-crusts composed mainly of plant ingredients (low δ15N, high C/N) from those likely to be dominated by fish (high δ15N, low C/N), a pattern which is confirmed both by the 14C ages obtained (plant-based samples give more tightly clustered and later dates than fish-dominated samples) and by detailed biomolecular analysis of a much bigger sample set at the University of York (Figure 4). These results imply that much older dates from comparable pottery elsewhere are even more anomalous than previously thought.


bulk EA-IRMS values II

Figure 4: bulk EA-IRMS values and biomarkers in early (orange) and middle (green) Neolithic pottery food-crusts from Zamostje 2:  (A, filled diamond: amyrin present, derived from plant foods; B, = filled circle: full set of aquatic biomarkers present, derived from fish) (Bondetti et al. 2019, Quaternary International.


We are now (2016-2021) participating in the ERC-funded INDUCE project (, through which we have the resources to transform the chronology of early pottery in north-eastern Europe (600+ new AMS dates, if enough suitable samples are available). Through biomolecular, isotopic and microscopic analyses at the British Museum and the University of York, INDUCE is investigating the function of c.2000 pots from hunter-gatherer-fisher sites throughout the region from the Caspian and Urals to the south-west Baltic. The INDUCE dating programme includes

  • Dating food-crusts from typologically diagnostic sherds, usually after the full suite of analytical methods has been applied to determine the food-crust ingredients
  • Dating herbivore bones and plant fibres that are stratigraphically or functionally associated with early pottery, with species determination by ZooMS and C and N stable isotope measurement of bone collagen
  • Dating fully aquatic species across the study region, including modern fish, in order to better understand regional and local variation in freshwater reservoir effects
  • Experimental dating of different chemical fractions of well-dated Neolithic ceramics, to gain insights into the carbon sources responsible for pottery total organic carbon dates
  • From October, a specialist position for spatial-temporal modelling of the new results and legacy 14C data, based at the British Museum.

As well as providing robust chronologies for individual sites, we aim to test various scenarios for the spread of pottery in north-eastern Europe, including multiple exogenous sources and independent local development.

Document Actions
Unsere Webseite verwendet Cookies. Weitere Informationen. Einverstanden!