8127/frame/fm

c.
Uniqueness vs. Common Source
One point about which the current coterie of practitioners in the forensic
field seem to agree is that individualization in the forensic context refers to
the assignment of items to a common source or origin rather than the
determination of the uniqueness of any one object. For this reason the process
of individualization always requires a comparison between at least two items.
According to Saferstein (1998):
A comparison analysis subjects a suspect specimen and a control specimen
to the same tests and examinations for the ultimate purpose of determining
whether or not they have a common origin.
Similarly, DeForest, Lee, and Gaensslen (1983) write:
It [individualization] may also refer to the demonstration that a questioned
piece of physical evidence and a similar known sample have a common
origin. The nature of these individual characteristics varies from one type
of evidence to another, but forensic scientists try to take advantage of them
in efforts to individualize a piece of physical evidence by some type of
comparison process.
We will continue a discussion of these concepts in depth in Section 2, The
Principles of Forensic Science.
3.
Identification
The use of the word 
identification
in the context of forensic science has been
and continues to be a major source of confusion. The first quandary results
from the common lay usage of the word identification as a descriptor of a
unique item. In particular the standard usage in the context of “human
identity” or “human identification” confers the impression of uniqueness
because we have an innate understanding that each human being is unique.
The second source of ambiguity comes from within forensic science. The
word 
identification
has often been restricted to describing the appearance or
8127/frame/ch02 Page 54 Friday, July 21, 2000 11:51 AM

---

The Evolution of Forensic Science
55
composition of an item and using those characteristics to place it in a class
or category with other items; sometimes this is a step on the way to individ-
ualization and sometimes it is an end in itself. In any case, the field has not
agreed on a consistent usage of the term 
identification
and the ambiguity is
conveyed in a passage from the first edition of Kirk’s 
Crime Investigation
:
The determination of identity is of importance to the criminalist in two
ways: (1) in establishing between two objects an identity of origin; and (2) in
determining the nature of a specimen of evidence. The first category is
ordinarily the more significant because on it rests the final determination
of the value of the evidence. It is, for example, more valuable to say that
two bullets, one of known origin, were fired from the same gun, than to
say that the bullets are the same type of ammunition. It is more valuable
to be able to say that two hairs have come from the same head, than say
that both hairs are human in origin.
Contrary to widespread conventional wisdom, individualization may not
be the necessary or even expected goal in every forensic analysis. An example
is the analysis of solid dose drugs. If controlled substances are identified, it
may well not matter from which laboratory or field they came; mere posses-
sion fulfills the element of illegality. According to Saferstein (1998):
Identification has as its purpose the determination of the physical or chem-
ical identity of a substance with as near absolute certainty as existing ana-
lytical techniques will permit.
Interestingly, Kirk also points out that when simply identifying the nature of
a substance, rather than suggesting an individualization, statistical estimates
may not be relevant because the answer is a qualitative yes or no, rather than
a quantitative estimation of the frequency of occurrence. However, he also
emphasizes, particularly with respect to presumptive chemical testing, the
necessity to state clearly other substances that might give the same result.
Probability is a factor in virtually every phase of the study of evidence. The
most important exceptions are matters of straight chemistry, physics, or
other exact science when applied directly to testing for a constituent or
property on a single choice basis.… This type of problem in which a positive
or negative but definite answer is provided by the test is not one requiring
statistical treatment.… Even in chemical testing, however, there are
instances in which a test established only a probability of identity. For
example, many of the commonly used tests for alkaloidal poisons are not
specific. That is, the same color may be given when the reagent is added to
different alkaloids.
8127/frame/ch02 Page 55 Friday, July 21, 2000 11:51 AM

---

56
Principles and Practice of Criminalistics
In the next several chapters, we will discuss a modification to the forensic
paradigm in which we suggest that identifying the nature of a piece of evi-
dence as an end in itself should be distinguished from the classification that
takes place as an intermediate to possible individualization.
One of the most cogent commentaries on individualization and statistics
was written by David Stoney (1991) in response the then new application of
DNA profiling. Stoney draws distinctions between our acceptance of der-
matoglyphic fingerprints as inherently individualizing and our seeming need
to prove uniqueness for DNA profiling.
Secondly, we must look realistically at the individualization process. Are we
really trying to prove uniqueness? I would offer to you that it is a ridiculous
notion. The contrast with fingerprint comparisons is important. We hold
fingerprint specificity and individuality up as our ideal, yet this is achieved
only through a subjective process. In fingerprint work, 
we become subjec-
tively convinced of identity; we do not prove it
[emphasis ours].
If we were to pick one concept that the criminalist must appreciate, this
would be it.