Search

Die vorliegende Dissertation beschäftigt sich mit der Fragestellung, ob das Drahtziehverfahren bereits im Altertum zur Anwendung kam und nicht, wie in der bisherigen Forschung angenommen, erst im Frühmittelalter einsetzte. Um diese Fragestellung zu klären, wurden Drähte unterschiedlicher Produktgattungen untersucht und experimentalarchäologisch mit der Expertise des Goldschmieds und Restaurators Frank Willer nachgeahmt. Während einige Drähte aus einem Blech gezogen sind, konnte für den Draht des Kettenpanzers aus der Tiefenau bei Bern anhand einer metallographischen Analyse das Ziehen aus einem massiven Rohling nachgewiesen werden. Auch die Untersuchung der für das Drahtziehen benötigten Werkzeuge unterstützt die Annahme, dass das Drahtziehverfahren früher als bisher vermutet bekannt gewesen war. Durch eine materialwissenschaftliche Analyse der Zieheisen und -bronzen konnte nachgewiesen werden, dass diese Werkzeuge den Anforderungen für das Ziehen von Drähten entsprechen. Ein weiteres Argument für das Drahtziehen im Altertum besteht aus der chronologischen und geographischen Korrelation zwischen dem Einsetzen von Kettenpanzern und dem Aufkommen von Zieheisen innerhalb des Latènekreises. Für die Produktion von Kettenpanzern wurde Draht in großen Mengen benötigt, weswegen die Nutzung des im Vergleich ökonomischeren Drahtziehverfahrens naheliegend ist. Somit lässt sich das Ziehen von Drähten zwar schon vereinzelt für die Bronzezeit nachweisen, doch eine weite Verbreitung dieser Innovation geschieht erst ab der Früh- oder Mittellatènezeit mit der Nutzung von Kettenpanzern. Solch eine verzögerte Diffusion einer Innovation stellt keine Ausnahme innerhalb der Technikgeschichte dar und ist dadurch zu erklären, dass eine Innovation, die nicht fortgeführt wurde, an einem geographisch und zeitlich unterschiedlichem Punkt der Geschichte erneut auftreten kann, da es sich um eine konvergente Entwicklung handelt, bei der dieselben Rahmenbedingungen und Anforderungen vorliegen., The doctoral thesis examines the question of whether the wire drawing process was already used in antiquity and not, as assumed in previous research, first began in the early Middle Ages. In order to elucidate this question, the wires of different products were examined and imitated experimentally with the expertise of the goldsmith and conservator Frank Willer. While some wires were drawn from a sheet metal, by metallographic analysis the wire of the mail armour from Tiefenau near Bern was drawn from a solid blank. The investigation of the tools required for wire drawing also supports the assumption that the wire drawing process was known earlier than previously assumed. A materials science analysis of the drawplates showed that these these Bronze and Iron Age tools met the requirements for wire drawing. Another argument for wire drawing in antiquity is the chronological and geographical correlation between the introduction of mail armour and the frequent use of drawplates within the Latène circle. The production of mail armour required considerable quantities of wire, so the use of the more economical wire drawing process compared to hammering or torsion techniques is obvious. Although the drawing of wire can be proven for the Bronze Age in certain cases, this innovation was not widely spread and established until the Early or Middle Latène Period onwards with the use of mail armour and the accompanying higher demand for wires. Such a delayed diffusion of an innovation is not an exception within the history of technology and can be explained by the fact that an innovation that was not continued can reappear at a geographically and temporally different point in history, since it is a convergent development in which the same underlying conditions and requirements are present.

Historical materials are abundant. Yet, piecing together how human knowledge has evolved and spread both diachronically and synchronically remains a challenge that can so far only be very selectively addressed. The vast volume of materials precludes comprehensive studies, given the restricted number of human specialists. However, as large amounts of historical materials are now available in digital form there is a promising opportunity for AI-assisted historical analysis. In this work, we take a pivotal step towards analyzing vast historical corpora by employing innovative machine learning (ML) techniques, enabling in-depth historical insights on a grand scale. Our study centers on the evolution of knowledge within the `Sacrobosco Collection' -- a digitized collection of 359 early modern printed editions of textbooks on astronomy used at European universities between 1472 and 1650 -- roughly 76,000 pages, many of which contain astronomic, computational tables. An ML based analysis of these tables helps to unveil important facets of the spatio-temporal evolution of knowledge and innovation in the field of mathematical astronomy in the period, as taught at European universities.

Many learning algorithms such as kernel machines, nearest neighbors, clustering, or anomaly detection, are based on the concept of 'distance' or 'similarity'. Before similarities are used for training an actual machine learning model, we would like to verify that they are bound to meaningful patterns in the data. In this paper, we propose to make similarities interpretable by augmenting them with an explanation in terms of input features. We develop BiLRP, a scalable and theoretically founded method to systematically decompose similarity scores on pairs of input features. Our method can be expressed as a composition of LRP explanations, which were shown in previous works to scale to highly nonlinear functions. Through an extensive set of experiments, we demonstrate that BiLRP robustly explains complex similarity models, e.g. built on VGG-16 deep neural network features. Additionally, we apply our method to an open problem in digital humanities: detailed assessment of similarity between historical documents such as astronomical tables. Here again, BiLRP provides insight and brings verifiability into a highly engineered and problem-specific similarity model., Comment: 12 pages, 10 figures