8127/frame/fm

 
c.
Inference and Interpretation
While those espousing interpretation as the primary goal of a forensic exam-
ination use inference as the basis for communicating that interpretation, they
also frequently fail to distinguish clearly between source determination and
an inference of contact. Sometimes it is difficult even to determine which of
these ideas a writer has in mind. Early in their book, Robertson and Vignaux
(1995, p. 34), articulate a global “prosecution hypothesis” as “The accused
was present at the scene.” This clearly encompasses an inference of contact.
Later they describe the alternative (defense) hypothesis as “obtaining a match
by chance,” (p. 58) a clear reference to source determination. They then make
the point that “the match could not have occurred unless the prosecution
hypothesis was true.” Their prosecution hypothesis is predicated on contact
and transfer (the accused was present at the scene), while their defense or
alternate hypothesis relies only on source determination (obtaining a match
by chance).
Oddly, the distinction between source determination and inference of
contact becomes increasingly clear with more ambiguous evidence. In such
instances, the leap from source determination to an inference of contact
requires more assumptions, hence is more evident. Rarely, for example,
would an analyst argue that a physicochemical correspondence between a
brown cotton evidence fiber and a brown pullover sweatshirt is strong evi-
dence supporting an inference of contact.
 
d.
The Source of the Confusion
We have outlined some of the intellectual confusion that we perceive exists
in criminalistics today. Perhaps the first step toward clarification is to look
backward in an attempt to discern the origin of the confusion. Both anthro-
pometry and fingerprints were popularized as direct personal identification
evidence. Notwithstanding attempts by Galton, and later by others, to quan-
tify the likelihood of a coincidental match, the inference of source was gen-
erally accepted without much question. And because few or no inferences
need be made to connect a person with this type of evidence, the subtle
distinction between personal identification and inference of contact went
essentially unrecognized. Even Locard’s work with trace evidence focused on
the 
 
identification
of the dusts that he was convinced could provide informa-
tion about a person’s occupation and whereabouts. Although he was the first
to articulate the concept of transfer, if he appreciated the assumptions needed
to strongly infer the source of a relatively common fiber, or the further
inferences needed to conclude, from the discovery of a fiber at the scene, that
a person committed a crime, it is not prominent in his writings.
Relatively little documented work exists that explores a formal scientific
framework for extending a source determination to an inference of contact.
8127/frame/ch07 Page 168 Friday, July 21, 2000 11:46 AM

---

Association and Reconstruction — Inference of Contact
169
Taroni (1998) summarizes some of the early efforts. Perhaps, in part, because
of our 20th century love affair with technology, the focus in criminalistics
has been to develop more efficient, more sensitive, and more highly discrim-
inating tests, all directed toward increasing a random match probability in
support of an opinion of individualization. This has cast source determina-
tion as the central, or even sole, player in the use of physical evidence to help
solve crimes. Ironically, Kirk must accept much of the responsibility for
fomenting the very thing that he decried. For in declaring 
 
individualization
as the primary function of criminalistics, he placed the focus squarely on the
physicochemical analysis of objects, and shifted emphasis away from the
interpretation of data in the context of the crime. By focusing on individu-
alization, the field was doomed to progress that Kirk describes as “technical
rather than fundamental, practical rather than theoretical, transient rather
than permanent.”
Source determination depends on technology and our knowledge of the
nature of the evidence; scientific logic furnishes the means to summarize the
possibilities that could have produced the physical evidence. As we have seen,
the application of science consists of forming inductive inferences from the
analytical results, and making a statement about the likelihood of encoun-
tering the evidence under competing hypotheses. The truly scientific aspect
of our work lies not in the technical minutiae of the analysis itself, but rather
in the way that the analysis is constructed and interpreted (Evett, 1996). We
spend the rest of this chapter exploring some of these relatively undeveloped
ideas about inference and association.