8127/frame/fm

Bertillon fingerprint.
The photograph Bertillon produced to illustrate his “points.”
(Courtesy of Professor Pierre Margot, Institut de Police Scientifique et de Criminologie
de l’Université de Lausanne, Switzerland.)
He also admits that they would likely not stand up to a meticulous comparison.
It is also necessary to add that these patterns, which appear so similar after a preliminary
observation would not stand up to a meticulous comparison.
We are left to ponder his fascinating conclusion that echoes the lament of many a defendant
whose DNA profile is indistinguishable from the evidence.
The two examples of deceiving similarities that have just been presented show that the assertion
of identity relies less on the number of common particularities than on the undoubted absence
of dissimilarities. It is therefore, in the end, an 
induction based on a negative result
[emphasis
added].
In the end, he rescues fingerprinting from the oblivion where he apparently intended to send it
by giving the nod to personal expertise.
However, from a point of view purely philosophical, such conclusions are generally unconvincing.
Their value is uniquely derived from the accepted and already proven competence of the expert.
It is totally personal, whereas a non-identify may be shown without discussion by the finding of
obvious dissimilarities that may be verified by anyone.
We refer the interested reader to the full article by Champod, Lennard, and Margot listed in the
references below.
References
Champod, C., Lennar, C., and Margot, P., Alphonse Bertillon and Dactyloscopy, 
J. Forensic Ident.
,
43(6), 604–617, 1993.
Rhodes, H. T. F., 
Alphonse Bertillon, Father of Scientific Detection
, George G. Harrap, London, 1956.
8127/frame/ch02 Page 49 Friday, July 21, 2000 11:51 AM

---

50
Principles and Practice of Criminalistics
this concern. As early as 1879, Dr. Rudolph Virchow, a German pathologist,
realized the limitations of hair evidence. In comparing hairs, he accounted
for the reality that visible traits vary between different hairs of an individual,
even along the same hair, and that hairs from different people can look
similar. His understanding is evident in a portion of his testimony, “The
appearance of the hair is such that it is not inconsistent with having come
from the accused” (Thorwald, 1966). Unfortunately, not all hair analysts since
that time have followed his example. In the now infamous Canadian case of
Guy Paul Morin (
Commission on Proceedings Involving Guy Paul Morin
, 1998)
Morin was arrested almost solely on the basis of a single hair of dubious
quality found adhering to the necklace of the murder victim. This case is a
poster child for the necessity of the criminalist conveying, and the investiga-
tors and attorneys understanding, the limitations of the evidence.
In the last paragraph of the chapter on hair in the first edition of 
Crime
Investigation
(1953), Kirk writes of his hopes for the individualizing potential
of hair, in the process employing the inevitable fingerprint comparison:
That hair is actually characteristic of the individual is very probable since,
if it were not, it would be the exception to the general rule of biological
individuality. It therefore merits the most careful and extensive investigation
aimed at fulfilling the tremendous possibilities that it presents in this direc-
tion.… Simple and rapid methods capable of producing decisive results in
this field may be possible, but not without much more extensive and thor-
ough research investigation.… It seems safe to predict that such efforts
might well be rewarded ultimately with one of the most valuable sets of
techniques for establishing personal identification, using evidence that can-
not readily be kept from the possession of law enforcement agents, as
fingerprints are at present.
Earlier in the chapter Kirk intimates that the individualizing potential of hair
might lie in its transient chemical makeup rather than its inherent biological
properties.
The 
minor constituents
of hair have received only limited attention and it is
possible that these may lead to considerable advance from the standpoint
of criminal investigation.… The intake of arsenic, lead, silica, and other
minor constituents which are regularly detected in hair is certainly variable
and to some extent a function of the occupation, diet, and medication of
the individual. Thus these constituents might be expected to vary accord-
ingly and to be useful in determining the source of a hair, provided that
sensitive enough methods for the analysis can be developed and applied
without destruction of a prohibitively large amount of available evidence.
8127/frame/ch02 Page 50 Friday, July 21, 2000 11:51 AM

---

The Evolution of Forensic Science
51
In the second edition of 
Crime Investigation
(1974), buried in a section about
using neutron activation analysis to detect trace elements in hair, is the
acknowledgment that:
It may be mentioned, however, that the analysis of a single hair has not
been found adequate for personal identification to date, both because of
the difficulty of making a complete analysis of a single hair, and because of
the fact that variations may be considerable from hair to hair of the same
person.
Interestingly, his final paragraph in the 1974 edition, which rests on assump-
tions that hair evidence will necessarily exhibit “biological individuality,”
remains virtually the same.
In 1953, the same year Watson and Crick elucidated the structure of
DNA, Kirk predicted, “The criminalist of the future may well be able to
identify him directly through the hair he dropped, the blood he shed, or the
semen he deposited.” About 30 years later, Alec Jeffreys of the United King-
dom in fact presented the first “DNA fingerprint.” Jeffreys makes no apology
for his very deliberate descriptor of this technique. To set the record straight,
it is worth pointing out that Jeffreys’ first efforts, using a technique known
as multilocus RFLP, may well have been individualizing. In this technique, a
single DNA probe is used to investigate numerous locations (
loci
) in the
genome simultaneously. Complex patterns, containing substantial amounts
of information, are produced and initial calculations (Jeffreys et al., 1985)
suggested that the pattern produced by a single probe might well be unique
to an individual.
However, technical considerations and interpretation concerns led to the
eventual adoption of a single-locus system, in which only one location at a
time is analyzed. With each locus-specific result, a segment of the population
is eliminated but, in and of itself, the result is not individualizing. This was
even more true of the loci initially used for PCR-based techniques, because
each locus generally exhibited less variability. For both kinds of systems,
much of the power depends on the same concept that Bertillon employed,
multiplication of individual trait frequencies to reduce sequentially the por-
tion of the population exhibiting the combined traits. The difference is that
population data for genetic traits have been amassed in great amounts allow-
ing a truly quantitative assessment of the rarity of a composite profile. In
fact, genetic-based analyses (serology and DNA) are the only forensic tests
for which hard data exist from which to estimate the frequency in the pop-
ulation of an evidence sample exhibiting a particular. This same concept
could certainly be employed in dermatoglyphic fingerprint analysis, but the
8127/frame/ch02 Page 51 Friday, July 21, 2000 11:51 AM

---

52
Principles and Practice of Criminalistics
underlying population data (available in a raw form in the national finger-
print database, AFIS) has not been systematically organized and analyzed.
No doubt exists that DNA analysis provides the potential to individual-
ize.* This was not true when the first few single-locus RFLP probes came
online, however this conclusion can no longer be escaped. When 9 RFLP
probes or 13 PCR-amplified STR loci produce population frequencies as low
as 1 in quadrillions (that’s 15 zeros), the odds of the evidence coming from
another person are so small that a reasonable person is convinced of indi-
viduality. Interestingly, many in the forensic community continue to hold up
fingerprints and toolmarks, including bullet comparison, as the standard for
determining individuality. But without the hard frequency data now begin-
ning to be expected because of the work performed for DNA, these claims
are starting to be reexamined.