Forensics is the science of individualization. This quote from Paul Kirk shows how important the identification process is to forensics. The definition of identification in forensic science can be very different from that accepted in science, where the term « identification » is simply used to describe the assignment of an object to a particular class. In forensic science, the identification process is ultimately aimed at individualization. For forensic pathologists, the identification of an object means that it is possible to distinguish that object from all objects viewed. In the forensic literature, the problem of identity of origin is often addressed with reference to « class » and « individual » characteristics (Table 1). Comparisons that only lead to agreement on class characteristics (without significant differences) will lead to conclusions about « group identification ». Only if individual characteristics are present in relation to class characteristics can positive conclusions of identification or individualization be drawn. The definitions of « class » and « individual » characteristics are only conventional methods for describing selectivity. We will see that the problem of the inference of origin identity and its relevance to forensics is more complex than a simple dichotomy between class and individual characteristics. As Table 2 shows, we will distinguish between forensic fields that often lead to individualization and those that rarely lead to individualization (according to the current state of the art), but more often to corroborate evidence of varying strength. Philosophically, the identity of the source cannot be recognized with certainty and must therefore be derived. As Kwan has shown, the hypothetico-deductive method (supported by statistical inference methods) provides a reasonable explanation of how criminal lawyers infer the identity of the source.
The identification process can be thought of as a process of reduction, from an initial population to a restricted class or finally to unity. The initial population represents objects or witnesses, depending on the type of evidence. We have the combination of two factors: • A relevant population of controls or objects defined by their size (and/or other peculiarities). In other words, any member of this source population can be considered a possible source. • A reduction factor resulting from the combination of matching characteristics of a certain selectivity. In fact, the reduction is proportional to the rarity or random occurrence of these traits observed in this population. As Kwan suggests, « It is the sheer rarity of a trait that is important as the rarity of that trait in terms of the amount of suspect sources considered. It is important to emphasize that scarcity is relative to the situation.
What is the statistical significance of conclusions such as: « It is possible (or likely or very likely) that this pressure/mark was made by this particular element »? Such verbal scales can be translated into numbers expressing Pr(IDlE, N). It has been shown that these terms are understood fairly uniformly (in a digital conversion) by experts or lawyers. The digital conversion is shown in Table 4 (columns l and 2). From these probabilities, their corresponding frequencies (f) can be inferred using equation 2 taking into account an open quantity framework (N = 5 billion), as shown in Table 4 (column 3). The analysis of Table 4 leads to a paradox. In the open setting (the one that is supposed to be adopted in key forensic areas such as shoe marks, firearms or tool marks), the rarity of common characteristics corresponding to the most negative verbal statement is quite low (2.0 x 10-7).