Vergelijkend natuurwetenschappelijk dateringsonderzoek
For the first time in the Netherlands a start was made — besides dendrochronology (d) and in comparison with it - with thermoluminescence (TL) datings of brick and 14C datings (14C) of wood and organic material contained in mortars. In this summary the methods as well as the most accurate results will be assessed.
The oldest stages of the churches at Garnwerd (1229 ± 3 d), Oosterwijtwerd (1237 ± 5 d) and Marsum (1306 ± 6 d) could be ascertained by means of dendrochronology. The roof of the church at Britsum proved not to be older than 1464 ± 5 years, while the brick of the western block (1244 ± 75 TL) and the nave/choir (1324 ± 70) have only been roughly dated by means of thermoluminescence.
The townhall of Haarlem dates from 1369 (d), which is roughly confirmed by the thermoluminescence dating of the wallwork (1379 ± 65 TL). It is obvious that the dendrochronological dating method is still the most reliable and accurate. Nevertheless, the wood samples from the cloister shed of Ter Doest (near Bruges, 1272-1294 14C) and those of the church at Oostum could not be dated in this way because of too wide annual rings.
In combination with the dendrochronological datings of the Lairesse wing (1328-'29 d, Binnenhof. The Hague) and Oudegracht 219, Utrecht (1307-'08 d) the 14C dating of the outer sapwood proves to be 24 years early on average. With respect to a 14C dating of organic material contained in mortars two of the nine tests are possibly correct: the northern front of Ter Doest (1293-1421 14C) and the rear side of Kamperstraat l l Zwolle (595-1431 14C).
It concerns unburnt fibres, twigs, leaves and the like, which ended up polluting the undried mortar. However, a non-corresponding 14C score of 78% raises questions. Five times it concerns a material identified as charcoal. It is likely that these are the remnants of the fuel that was added to the lime when it was burnt. Dependent on its nature, this fuel could have originated/died off at various moments. For instance, this appears to be the case with respect to the mortar gathered from the western block of the church at Britsum (approximately 4450 BC 14C). Besides, it is possible to mistake dark shells for charcoal. The samples drilled from the cellars of the Binnenhof complex (with water and diamond drills), did not produce any practicable 14C or TL outcome.
From the twelve calculations made by R. Kotalla five (=42%) do not match an assumed or otherwise obtained dating at all (Rolzaal, transverse cellar and Lairessezaal, all at the Binnenhof in The Hague, western and northern fronts Ter Doest). In the case of the Binnenhof the permanently moist situation in the cellars may have played a part, particularly in comparison with the piece of brick for which it did work, from the drier interior of the western front of the Ridderzaal (Binnenhof The Hague, 1288 ± 6 d, 1352 ± 65 TL, 1259-1232 14C). The failure of the two TL datings of Ter Doest cannot be accounted for, unless there was a fire there.
If verifiable, the datings according to the TL method average out at 13 years earlier than the datings by way of dendrochronology. The margin of ± 65-75 adopted for the 13th and 14th centuries could be reduced to ± 50 years on the strength of the (4) verifiable cases. The margins of the TL are comparable to those of the 14C research, although those of the 14C are considerably smaller with samples of a mass of over l milligram.In the absence of old wood a TL dating of brick seems to be more reliable so far than the 14C dating of organic material contained in mortar. For TL research it is best to take samples from dry, higher-situated wall interiors, for instance at points where there are unfinished breaks. The 14C dating of charcoal in mortar is unreliable, even impracticable in this research, which may have to do with inexperience in taking samples by means of pulverization. Van der Borg mentions that in general samples smaller than 0.4 mg charcoal are problematic because of limited accuracy as a result of necessary corrections. Moreover, the odds are that the sample is not representative for the material to be dated. Through unburnt fibres, through distinctly deviating materials, a dating of mortar may indeed be achieved, as Van Strydonck demonstrated earlier on the basis of candle-grease, horse hair and also charcoal. On the other hand, wood cut from the outermost annual rings of old beams produces results that are to be compared to those of dendrochronology, although the margins do not correspond completely.