Association and Reconstruction — Inference of Contact
183
are drawn based on what the
examiner
thinks is
most likely
, as opposed to
expositing the alternative possibilities and evaluating the evidence in light of
them.
One example of such an overinterpretation occurred in a case involving two
suspects accused of torturing and shooting the clerk at a
Stop ’N Rob
con-
venience store. A detective from another jurisdiction, who
had attended
several bloodstain pattern interpretation workshops, was retained to prepare
a reconstruction based on the blood patterns at the scene. He wrote an
11-page report, including every movement of the victim and suspects over
a 30- to 45-minute time period. He included in his reconstruction a
5-minute time period where all activity ceased for the three individuals,
basing this on a photograph of what he interpreted as a blood clot. He
concluded that the victim was shot, bled for a period of 5 minutes on the
floor, and then was beaten by the assailants. How did he arrive at his detailed
reconstruction? He was given one suspect’s
confession, who indicated that
they had shot the victim, sat around eating sandwiches and mocking him
for 5 minutes, and then beat him for good measure. The detective used this
as a starting point for his reconstruction, and picked out stains from the
crime scene photographs that appeared to him to be clotted blood. As a
nonscientist, he couldn’t quite comprehend the criticism that it was not
possible to distinguish between a blood clot and a bloodstain based on a
photograph. Happily, he did not include a reconstruction of the suspect’s
meal, although perhaps only because no blood was involved in it.
The numerous reconstructions of the Marilyn
Sheppard murder provide
a classic example of how completely antithetical conclusions can apparently
be reached from the same evidence. We contrast two examples, one from
Paul Kirk (1955) a criminalist, the other from Gregg McCrary (1999) a
behavioral scientist (profiler). Although these individuals come from com-
pletely different backgrounds, resulting in varying perspectives, the pream-
bles to each of their conclusions sound eerily similar:
Kirk — “… the facts are clear, and the conclusions inescapable.”
McCrary — “The … evidence allows for
only one logical conclusion, …”
Both men worked from essentially the same physical evidence results, except
that McCrary had the benefit of DNA evidence that only became available
several decades after Kirk worked on the case. Kirk’s inescapable conclusion
was that Sam Sheppard
did not
murder Marilyn Sheppard; McCrary’s one
logical conclusion was that Sam Sheppard
did
murder Marilyn Sheppard.
That both would reach mutually exclusive conclusions from the same evi-
dence accents the dual dangers of ignoring inconvenient evidence and
emphasizing only evidence that tends to corrobarate one’s hypothesis. Both
8127/frame/ch07 Page 183 Friday, July 21, 2000 11:46 AM
184
Principles and
Practice of Criminalistics
reconstructions are highly subjective and speculative. We highlight sections
of each to illustrate the seduction of overinterpretation and the dangers of
ignoring evidence that does not fit your theory.
We start with McCrary on the issue of a sexually motivated attack*:
The body of Marilyn Sheppard was found on her bed in the early morning
hours of July 4, 1954. Subsequent investigation revealed that her pajama
top had been unbuttoned and pushed up exposing her breasts and that one
pant leg of her pajama bottoms had been pulled off of her leg. The victim’s
knees were at the foot of the bed with her lower legs hanging vertically from
the end of the bed. Her legs were slightly spread. Blood smears on her ankles
along with the blood pattern higher on the bed sheet suggest that the killer
pulled her down toward the foot of the bed into that position…
With
her breasts exposed, one pant leg removed, her legs slightly spread
and hanging off the foot of the bed, one would have to consider the potential
that this was a sexually motivated crime. However, the evidence does not
support this hypothesis. Having manipulated the victim as described above,
the offender positioned her so that a large horizontal wooden bedpost is
over her pelvic and vaginal area thus presenting a physical obstruction which
would prevent anyone from sexually assaulting the victim in that position.
In fact the evidence shows that the victim was not sexually assaulted…
An examination of the vaginal smears
taken from the victim during
the autopsy found no spermatozoa present.
It is odd that McCrary assumes that if Marilyn Sheppard was sexually
assaulted, it would only have been in the final position in which she was
found. What eliminates the very reasonable possibility that she was raped in
another position on the bed and then killed? This is but one of many examples
of failing to entertain an alternate hypothesis. Further, McCrary misrepre-
sents available facts when he states that “the evidence shows that the victim
was not sexually assaulted.” In examining vaginal swab slides taken at autopsy
in preparation for DNA analysis in 1997, Mohammed Tahir states explicitly
that he observed sperm. Because that work was
performed for the plaintiff
in the civil case, it is inconceivable that McCary did not have access to it.
One might legitimately ask if this is simply an example of ignoring evidence
that is inconvenient to the proffered theory of the incident.
Later in the report, McCary states that:
There is also inconsistent behavioral evidence in that if this was a sexually
violent offender who engaged in overkill to subdue his victim it is most
likely that the sexually violent behavior would have continued unabated
and he would have ripped the victim’s clothing off, and brutally raped her
leaving clear evidence of vaginal and anal trauma.
* Excerpts reproduced here and in Sidebar 5 with all typographical
and grammatical errors
intact.
8127/frame/ch07 Page 184 Friday, July 21, 2000 11:46 AM
