"Data from proficiency tests and other examinations suggest that forensic errors are not minor imperfections . . . Handwriting error rates average around 40% and sometimes approach 100%."
Source:
Michael J. Saks and Jonathan J. Koehler. "The Coming Paradigm Shift in Forensic Identification Science."
Science, New Series, Vol. 309, No. 5736 (Aug. 5, 2005), pp. 892-895
"The writer of a forged signature can be identified from examination of his known handwriting but only under rather special circumstances.
The qualifying portions of this statement may surprise many readers.
There is a rather widespread belief that document examiners can identify
the writer of any forgery from specimens of his handwriting. Unfortunately,
in a great many instances this is far from true. It seems appropriate,
therefore, to consider this problem, paying particular attention
to those cases in which the forger can be identified, but also looking
into those more common instances when he cannot."
"There are two requirements which must be satisfied before a positive
identification can be made. First of all, the forged signature must have
been written in the natural handwriting of the forger. Signatures for
the most part are short, and even a moderate degree of disguise may
prevent accurate identification of a single specimen. The second condition
is the need for a very large quantity of the forger's known writing. One
or two specimens of his signature alone are virtually useless for solving
the problem. If a large quantity of day to day general writing is not
available, then naturally written request specimens of the forged name
must be obtained. Only under these conditions is identification of the
forger possible."
"In considering this subject it is only proper to look carefully at the
various pitfalls which confront the examiner when he is called upon to
identify the forger from his handwriting. It has already been stated
that this identification problem is among the most difficult and is a more
frequent source of error than almost any other handwriting problem.
Disregard of the dangers that confront the examiner can easily cause
him to be mistaken. These include:
1. Inadequate standards.
2. A tendency to discount differences as either being representative of disguise or chance variation.
3. An identification based only on general handwriting habits and not upon
individual ones.
4. The influence of the surrounding circumstances, that is, factors which
have no part in the handwriting identification."
"Let us consider these points briefly.
"Inadequate Standards. With only a very limited amount of known
handwriting the standards may not be truly representative of the suspect's writing habits. This would be especially true if they were written
under unusual writing conditions. Chance variations in these standards
may assume undue importance, causing error in the identification;
while if a larger quantity of known handwriting was available, it would
appear at once that the suspect could not have written the forged signature. More often, though, when the standards are limited the error is
one of not identifying the guilty person. Identification of the forger
from his handwriting requires far greater quantities of handwriting than
are necessary to show that the signature is not genuine,1 and it is always
risky to base an opinion on a rather limited amount of handwriting.
Unless the similarity between the forged signature and the forger's
writing is so great that no significant differences exist, an identification
must be cautiously approached.
Differences Are Disguise. In any fraudulent signature in which no
attempt is made to imitate another's writing, there is always a likelihood
that the writer may attempt to change his own handwriting. In
a signature made up in a large measure of capital letters he can be
quite successful. If the fraudulent signature contains disguise the
chance of his identification is greatly lessened. On the other hand, differences which might be attributed to disguise because of their poor
quality of execution may actually be the natural writing of a writer of
less skill than the present suspect. If there are indications that the
forged signature contains elements of disguise, there is less likelihood
that the forger can be identified, and the conservative and scientific
document examiner may find it necessary to qualify his identification or
refrain from any at all.
General Writing Habits. By far the most common errors in the
identification of the forger arise out of overemphasis of general writing
characteristics which are common to both the suspect's writing and the
forged signature. By general characteristics are meant those writing
habits which are part of a basic writing system or which are modifications
of the system of writing found among so large a group of writers
that they have only slight identification value. These might include an
open top "o" and "a" or a looped "t" form, which occur in many rapid,
careless handwritings. When an unexperienced or unscientific examiner
discounts the differences between the known and forged signature as
disguise and then makes his identification on the basis of three or four
general similarities, error is almost certain. General similarities can
certainly form a part of the basic identification, but there must be a
very unique combination of them and of the individual or personal
writing habits, with no fundamental differences, in order for the identification to be accurate.
Surrounding Circumstances. It is not uncommon for an aggressive
investigator to submit a signature forgery case to the examiner pointing
out evidence other than the handwriting which seems to clearly establish
the suspect's guilt. The investigator will want to detail all of these
extraneous circumstances to show how clear and convincing is this proof.
More than one case is on record in which a person was accused by eyewitnesses of writing a forged signature when it could be shown clearly
from an examination of his handwriting that he did not write it. Furthermore, these findings were subsequently substantiated by the confession of the guilty person.
Extraneous factors which incriminate the suspect have no place in
the document examiner's consideration. He must base his identification
entirely upon the handwriting and those conditions under which it was
written. If he cannot find sufficient similarity between the suspect's
writing and the forged signature for him to make an identification without
the assistance of other evidence, then he has no place as a witness
against the suspect."
Source:
Ordway Hilton. "Can the Forger Be Identified from His Handwriting?"
The Journal of Criminal Law, Criminology, and Police Science, Vol. 43, No. 4 (Nov. -Dec., 1952), pp. 547-555