User:Taxelson/Late Roman gap

From Cybis Wiki
Jump to: navigation, search

Late Roman gap. There are a lot of western European oak chronologies which goes back to about the "migration period" or "later" Roman time. There are also a great number of Roman time chronologies from Germany, Netherlands, France and England which forms a consistent complex. But the problem is that very few links - if any - exists between those two complexes! This means that the whole Roman time complex may possibly not have been correct anchored - i.e still floating!

The accepted anchoring seems to be based upon the work of Ernst Hollstein and Berndt Becker only. The Hollstein link, however, is proved to be very weak - and his material do instead imply a very different position for the roman complex on the timescale.[1] and according to the Becker chronology (Southern Germany) there are no information available about the quality of the critical Late Roman/Migration period linkage.[2] It is maybe not sure that the Becker chronology did not use the Hollstein chronology as a template. We do not know.

Other links, that supports the current anchoring of the Roman complex, are actually not strong enough to prove the case.


List of available master chronologies and collections

Anchored in present time

Anchored in Roman time

  • HollsteinBestRoman 340 BC - AD 336. Trier area, Germany. Data retrieved from diagrams, analyzed and made available by Lars-Åke and Petra Larsson[6] The young end may possibly have severe problems, probably due to contamination with younger material in the mean chronologies, see The Hollstein pseudobridge below.
  • Becker: South Germany curve 370 BC - AD 1950[4]
  • Becker: Eight curve diagrams for Roman time samples are also retrieved and avaliable.[7] 24 BC - AD 181
  • Netherlands 154 BC - AD 205: Collections ITRDB: neth005.rwl info, ITRDB: neth009.rwl info, ITRDB: neth010.rwl info, ITRDB: neth011.rwl info,ITRDB: neth012.rwl info, ITRDB: neth013.rwl info, ITRDB: neth019.rwl info, Ester Jansma et al., 154 BC - AD 205
  • England: A collection of chronologies from England compiled of not yet published measurement data from Sheffield University and English Heritage, compiled by Petra Ossowski Larsson.[8]
    • Southern England: 13 chronologies, 256 BC - AD 207
    • Northern England: 5 chronologies (including one from QUB - Carlisle), 372 BC - AD 103
  • Ireland (mainly): "LateBC" QUB material compiled by Petra Ossowski Larsson. 1155 BC - 69 BC[9]

Known possible bridges supporting the current hypothesis


No information about the quality of this link is available. No measurement data available. Nothing said about the number of samples and sites involved, nothing about the length of overlap or t-values found. It is therefore unwise to rely on this link if not certified by independent material.

Denmark - Netherlands - Hollstein best Romans

The chronology which can be extracted from ITRDB: neth002.rwl info and ITRDB: neth006.rwl info (Ester Jansma et al.) covers AD 248-395 (as dated) dates towards WestDK AD 200 - 1986, although with not too impressing values: corr=0.31, overlap=147, T=4.0. I.e values not strong enough to certify a dating, although the block AD 280-362, gives corr=0.56, overlap=82, T=6.0 - and it is not uncommon that distant chronologies as those sometimes are varying according to corr values from period to period.

The relation between the Netherlands AD 248-395 and Hollstein best Romans gives: overlap=59, corr=0.40, T=4.1. That is far from an indisputable match... Towards South Germany the overlap is the all available 147 years, but only with corr=0.34, which gives T=4.4. This means that this bridge is too weak to rely on.

Ireland - English Roman

The old end of the Belfast Chronology, consisting of five local chronologies from Northern Ireland, should overlap the English Roman complex with more than 200 years 13 BC to AD 207. Most of those chronologies, however, contains very weak old end tails with rings produced by young trees or plants, which use to give untypical growth pattern. Therefore it can be argued that it should be a good idea to truncate them. If that is done we become an overlap of 172 years and corr=0.35, T=4.8.[10] So there is a bridge, but too weak to be satisfying. And if not truncated chronologies are used, we will see much lower values.

Alternative bridgings

The Hollstein bridge (207 years shortcut!)

When the HollsteinBestRoman is tested towards HollsteinBestLateToYear401, a match appears, implying a date for the HollsteinBestRoman 207 years later than expected, i.e AD 543 instead of AD 336! The match is rather impressing, indeed: overlap: 142, corr(P2Y)=0.53, T=7.4; corr(Ba/Pi)=0.46, T(Ba/Pi)=6.1. This is normally considered as a safe level for crossdating.[11] The reason for this confusing match is not clear, although a lot of work has been laid to understand it.[12] Contamination with erroneously dated samples of the mean curves plotted by Hollstein may be a reason, but no evidences for that is yet found. And there are no spectacular correlation between neither of the unexpected matching Hollstein ends towards the South German/A100 chronology at their +/-207 years positions, but very good at their expected positions. Could this be the true date for the roman complex, although most historians would have great problems to accept that view! No other obvious bridges to the Roman complex, corresponding to the -207 years shortage, have yet been found to confirm this revolutionizing re-dating.


  1. Lars-Åke Larsson & Petra Ossowski Larsson: An introduction to "The validity of the European chronology".
  2. Larsson: 'Retrieving some Becker data. The "Merovingian gap" is discussed by Becker but the "Migration period" is totally left in silence.
  3. Lars-Åke Larsson & Petra Ossowski Larsson: Merging Hollstein curves - an interpretation of the data Download: data set WebCite-archive
  4. 4.0 4.1 L-Å Larsson: Retrieving some Becker data. Download: Becker Chronology file WebCite-archive
  5. (WebCite-archive)
  6. Lars-Åke Larsson & Petra Ossowski Larsson: Merging Hollstein curves - an interpretation of the data Download: data set WebCite-archive
  8. For details see: An Irish tree ring chronology. Downloadable file: (WebCite-archive)
  9. Consists of mean chronologies for AnnaholtyBog, Ballymacombs4, Corlea, DerraghanMore, Dorsey, GarryBog4, Navan, SwanCarr. Available in (WebCite-archive)
  10. The collection of truncated chronologies are available in the zip-file available from QUB - Northern Ireland first century AD
  11. Torbjörn Axelson and Lars-Åke Larsson: What is a good TTest value
Personal tools