Remarks on the age of the bronze Venus on display at
the Pitti Palace in Florence
A privatly owned bronze Venus is currently on display at the Pitti Palace in Florence. The
sculpture was discovered some 30 years ago at a Paris scrap merchant and has raised
debate ever since concerning its date of fabrication and artists attribution. According to a
brochure written by one of its current owners (1) thermoluminescence (TL) data of the
casting core left inside the bronze indicate a 16th century fabrication date, while excluding
a 17th century date, thus supporting an attribution to Giambologna (1529-1608).
The aim of the following remarks is to point out that a reliable fabrication date of the
bronze cannot be derived from the TL-data presented, and that the attribution of the
bronze to Giambologna based on these data is highly speculative. For a person skilled in
the art, the TL-data have two major shortcomings:
1- Errors are underestimated: according to the scientific report (2, thereafter "report")
attached to the owners brochure, two TL-measurements have been performed, one in
1996 by the Rathgen Forschungslabor Berlin, and another in 2008 by Oxford Authentication
Ltd. The 1996 measurement suggested a fabrication year around 1582 with an error of plus
or minus 22 years (67% probability), while the 2008 measurement suggested a fabrication
date anywhere between the years 1308-1608 (without error and probability indications).
For a person skilled in the art, these dates are unreliable for the following reasons:
The error of ± 22 years attributed to the 1582 date of the first TL-measurement is greatly
underestimated. According to the report (2) it was calculated under the assumption that the
observed spread of measurements was due to random errors only and followed a normal
distribution. This assumption is not justified. TL-age measurements almost never follow a
normal distribution, while random errors constitute only a fraction of their total error. TL-
measurements are usually dominated by systematic errors intrinsic to the TL-dating technique
itself, such as those originating from the unknown history of radioactive force acting on the
artifact, a possibly incomplete heating of the object during fabrication, an accidental exposure
to X-ray radiation altering the stored dose, an unsuitable sampling location for the TL-
measurements, and other factors summarized by a person which was affiliated with the
commissioner of the first TL-measurement of the bronze Venus (3). While these errors can
rarely be quantified in a straight forward manner, they are generally considered to amount
typically to ± 20% of the TL-age (3). By comparison, the ± 5% error claimed for the TL-age (414
± 19 years) of the present bronze is smaller by a factor of four. The question concerning this
unusually high accuracy has not been adequatly adressed in the report.
As to the second TL-measurement, the report contains no indication as to how its errors
were estimated. Thus, a person skilled in the art may choose between two alternatives.
Either the range of production dates given (1308-1608) refers to the limits of a 67% (one-
sigma) probability of a normal distribution as for the first TL-age measurement, or it
represents the maximum possible errors as estimated by the authors. Neither one of these
alternatives has been discussed in the report. Thus, from a scientific point of view, the
range of production dates claimed is of very limited value.
Considering these inconsistencies, and given the lack of information on how the error of the
second TL-measurement has been estimated, a person skilled in the art cannot exclude a
production date of the bronze in the late 17th century.
2- Lack of data: Appendix IV of the owners brochure lacks crucial details. Apart from the error
estimates for the second TL-measurement, it contains neither TL-curves, nor a signed
document of Oxford Authentication Ltd. allowing their results to be checked independently.
Without such a document, their TL-age estimation must be believed on a bona-fide basis. The
report also remains silent about the exact location from where the sample of the second TL-
measurement has been taken. Normally, this location is indicated in a drawing and included -
together with other details and the name of the experimentalist - in the reports. Finally, the
time at which the bronze underwent artificial irradiation (X-ray radiography at the
Rijksmuseum Amsterdam, X-ray CT at the Frauenhofer Institut in Fürth) is not indicated in the
report (see Table 2). This omission may have serious consequences, because if such
irradiations had been performed before the TL-measurements, they would tend to artificially
increase the TL-age, i.e. simulate an older fabrication year of the bronze. An indication that
such an irradiation might have taken place before the second TL-measurement comes from
the shift of possible production dates towards older TL-dates as compared to those of the first
TL-measurement, including years as early as 1308. In this regard the absence of TL-curves in
the report must be deplored because their behavior allows a person skilled in the art to
determine if such an irradiation had indeed occured.
Altogether, the lack of data in the report does not allow a person skilled in the art to dismiss
the possiblity of an artificial irradiation before the date of the second TL- measurement (2008),
and thus to exclude a false TL-age.
In conclusion, given the absence of correct error estimates for the TL-age, and
considering a general lack of data, the claimed 16th century fabrication date of the
bronze Venus is highly uncertain, and thus its attribution to Giambologna speculative.
Literature:
1- AlexanderRudigierandBlanca Truyols: "GIAMBOLOGNA Court Sculptor to Ferdinando I,
his art, his style and the medici gifts to henri iv", 376 pp. 600 col. ills. Revised Edition (AD
ILISSVM, London, 2019)
2-Ernst Pernicka: Analysis Report, Appendix IV in ref. (1)
3- Jane Basset: Thermoluminescence dating for European sculpture: A consumer’s guide;
American Institute for Conservation of Historic & Artistic Works (AIC). AIC Objects Specialty
Group Postprints, Volume 14, 2007 Pages: 32-46
Brussels, 31 Dec. 2019 BANS/KY-Authentication