Identification Systems in Developed Countries: Taking Germany as an Example

Eyewitnesses hold equally important positions in criminal justice procedures across the globe. Therefore, ensuring the accuracy of eyewitness identifications of defendants has become a critical issue. In Germany, through academic theoretical discussions and recognition by legal practitioners, several basic principles of identification have been summarized (refer to Schweling, MDR 1969, 177ff.):

Basic Principles for Conducting "Identifications" in Criminal Investigations

Identifications should be conducted as "choice-based" identifications rather than single-suspect identifications (i.e., "yes-or-no" identifications).

Before identification, the identifier must first state the characteristics of the suspect.

The individuals presented for selection in the identification must not have major differences in appearance.

No potentially suggestive arrangements shall be present prior to the identification.

The identifier must be told prior to the identification: "The actual criminal suspect is not necessarily among the persons being identified."

Some of these principles have also been adopted into Section 18 of the German Guidelines for Criminal and Fine Proceedings (Richtlinien für das Straf- und Bußgeldverfahren). In practice, the Criminal Police Office in Nuremberg, for example, typically arranges six to eight individuals for choice-based identifications during each identification procedure.

Criteria for Courts to Determine the "Admissibility of Evidence for Identifications"

Whether the identification during the investigation process complied with the aforementioned principles will become the basis for courts to determine whether it can be admitted as evidence. Therefore, several criteria for determining the "admissibility of evidence for identifications" have been summarized in litigation procedures:

Data concerning the identification process (such as photographs of all identified persons, audio and video recordings of the identification process, etc.) must be thoroughly attached to the case file; otherwise, the identification lacks admissibility as evidence.

The judge must investigate the identification process and explain the reasons for admitting it as evidence (especially for identifications that do not fully comply with the aforementioned principles); otherwise, the judgment shall be deemed illegal.

Repeated identifications (subsequent identifications by the same identifier) lack admissibility as evidence under the law of evidence.

The court can only ever review the initial identification; therefore, if the initial identification has flaws, it cannot be remedied.

Courts have also affirmed these determination criteria in judgments. In a 1961 precedent (BGH 16, 204 (1961)), during the investigation of a case, two children aged eleven and thirteen respectively identified a suspect from among fifty photographs, but the suspect refused to submit to an in-person identification. Subsequently, on a chance occasion, the suspect encountered these two children and was pointed out. The court considered this a "repeated identification," denied its admissibility as evidence, and emphasized in the judgment that "it is exceptionally difficult, particularly for children, to resist suggestion."

Research reports published by two scholars, Brian L. Cutler and Steven D. Penrod (published in Identifizierung von Tatverdächtigen durch Augenzeugen, 31f., Günter Köhnken and Siegfried L. Sporer (eds.)), point out that the reliability of identifications by young children increases with age, but for those aged 9 to 11, it is still only 60% to 70%, and this is statistical data obtained under conditions without suggestive prompting.

In addition, some research reports have also indicated that identifications involving siblings or relatives are more likely to result in errors, and the identifier's emotions as well as the context during the identification will also affect the accuracy of the identification.

The Urgency of Reforming Our Country's "Identification" System

Factors that German theory and judicial practice deliberately strive to avoid because they affect the accuracy of identification—such as "single" (yes-or-no) identifications, "repeated" identifications, suggestive prompting, and the complete absence of records of the identification process—are all commonly occurring in the identification processes of criminal cases in our country. Consequently, the possibilities of suggestion or error arise extremely easily. Unfortunately, our country's criminal investigation practice has historically failed to establish any requirements or regulations regarding how "identifications" are conducted, and front-line police and investigative personnel conduct identifications carelessly and imprudently. In particular, eyewitness identifications will undoubtedly be judged by judges as evidence with immense probative force; hence, if an erroneous identification occurs, it will lead to severe consequences. Yet human memory is so easily susceptible to interference, making identifications extremely prone to suggestion and error. Therefore, from how an "identification" is conducted during the investigation process to how an "identification" is scrutinized during the trial process, strict layers of principles are required to act as safeguards; otherwise, "identification," the type of evidence most easily judged as having immense probative force, will often simultaneously and most easily become the basis for miscarriages of justice and the source of wrongful imprisonments.

Similar incidents, we believe, continue to occur within the criminal investigation and trial processes of our country. Therefore, referencing the practices of developed countries to establish strict identification procedures and scrutiny principles is a matter of utmost urgency.