As dry as a (burnt) bone : histological, physicochemical, and osteogenic reaction-related approaches to identifying the pre-burning condition of bone

Burnt human remains are recovered from a variety of contexts as a result of funerary treatment (cremation) in archaeology and fatal fires in a medicolegal context. The timing that this burning takes place compared to the time of death has received a great deal of attention in the literature, with inconclusive results. Cremation after active or passive excarnation, as a funerary rite, was mainly proposed during the prehistory of the British Isles. The postmortem (after death) transformation of the corpse starts during early diagenesis, thus taphonomic changes are thought to have the potential to determine the postmortem interval (PMI) at the time of burning. This thesis aims to provide insight into the taphonomy, early diagenesis and heat-induced changes of bone through an experimental approach with repercussions to both archaeology and forensic science. The thesis is a collection of three interdisciplinary approaches organised into three papers to identify if the corpse was burnt fresh or partially decomposed. Fleshed pig (Sus scrofa) tibiae, used as a proxy for humans, were left exposed in a field in Wytham Woods, Oxfordshire, UK, then collected at 14, 34, 91, 180, 365-day intervals prior to being burnt in an outdoor fire (≤750 °C bone temperature). Fresh (fleshed) tibiae acted as unburnt and burnt controls. In addition to this experimentation, two cremated human bone fragments from Middle/Late Neolithic (ca. 3300-2500 cal BC) Ireland were added to the physicochemical investigations. A variety of techniques were employed across academic disciplines: (1) histological, using microbial bioerosion as a proxy for decomposition by transmitted light microscopy and backscattered electron microscopy (BSEM); (2) physicochemical, investigating the structural changes through Fourier Transform Infrared Spectroscopy (FTIR) and major and trace elemental concentration changes by the wavelengthdispersive electron microprobe (EMP-WDS); and (3) the bone's osteogenic-reaction, investigating its response to sharp force trauma (SFT) and heat-fracturing compared to its collagen content. Data analyses were carried out using a variety of methods. Histotaphonomy was assessed quantitatively and systematically by a novel data labelling application developed for this study in Python and Javascript programming languages. The physicochemical results were investigated by unsupervised (PCA) and supervised (LDA) dimensionality reduction techniques and further statistical analyses (linear regression, MANOVA, multivariate comparison of means test). The latter was applied in the morphological investigations as well. All data analyses were done in Jupyter notebook in Python programming language. Results show that changes in elemental concentration of bone throughout the PMI might indicate whether the bone was burnt fresh or partially decomposed as well as if SFT was inflicted after 6 months of exposure before burning. If SFT is present in the bone, results show that the decrease of perimortem features (irregular edges, splintering, curling or uplifting) and increase of postmortem features (regular shaped, sharp edges) in cutmarks can potentially inform on the timing of SFT infliction in burnt bone. However, microbial bioerosion might not indicate early diagenesis prior to burning. Features resembling microbial bioerosion were present on the fresh burnt bones, indicating that burning is not only able to conceal features but also produce them. The concentration of elements associated with extracellular fluid (K, Na, Cl) change with the PMI and endure burning. K values under 0.07 ± 0.01 wt% in inner and midcortical zones of burnt bones suggest that bones were not burnt immediately after death. Using this criterion, results from the Neolithic bone fragments would indicate a PMI of at least weeks to months prior to cremation. Ca, P, Fe, Al, Si, and Sr are not significantly altered with burning, and Fe, Al, Si, Sr are also unaffected by the PMI. In unburnt bone increased crystallinity and carbonate loss are detectable in < 1 year, but both are obscured by burning. Structurally, the carbonate to phosphate ratios (C/P) and phosphate high temperature (PHT) are the most 5 valuable ratios for discriminating between unburnt and burnt bones. Fractures did not depend on the limited collagen loss during the one-year PMI prior to burning. All in all, the amount of potassium in burnt bone is suggested to be the most useful proxy of bone decomposition prior to burning.