accused, the interests of victims of crime or their families in apprehending and convicting perpetrators, and the interests of society. Whether the interests of society in seeing that justice is done should count as much as the interests of the accused or the victim is open to question. (Here there is an obvious overlap with an ethical analysis from the perspective of rights, and assessment of the consequences of instituting a new practice should include the effects of the new practice on the rights of the people involved.)
Especially when a practice is new and information on projected consequences is scanty, there are problems with relying on balancing the good and bad consequences as a mode of ethical analysis.9 People who favor one policy or practice predict a balance of good consequences over bad ones, and detractors do the opposite.
One important factor contributing to uncertainty about the use of DNA typing technology is the existence of disagreement among scientific experts.10 When experts disagree about the use of techniques or statistical methods (such as extrapolations based on population genetics) or about the interpretation of data, the uncertainty is of a different sort from uncertainty that stems simply from scanty evidence drawn from actual consequences. The latter uncertainty can be remedied by gathering more data before a technology is introduced as an accepted standard. If controversy among experts persists, disagreements can erupt whenever empirical evidence is analyzed and specific conclusions are to be drawn.
An overall judgment that DNA technology in forensic science is superior to existing forensic methods requires comparing intersubjectively verified scientific evidence on the reliability and validity of the new method with evidence on the other methods. Certainly, as a personal identification method, fingerprinting is the definitive forensic technique and has many advantages. It has almost 100 years of development, which has established empirically that a person has unique fingerprints; fingerprints can even distinguish between identical twins. Fingerprints are easily detected and developed, and large electronic fingerprint databanks exist all over the world. A fingerprint is a directly observable impression that does not generally involve extensive chemical or biochemical manipulation. Rarely do fingerprint experts differ in conclusions reached after examination of fingerprint evidence. The limitations of the technique are derived from the fact that usable fingerprint evidence is left at crime scenes relatively rarely and can indicate only the presence of a person at a scene.
Another method of identification commonly used in forensic laboratories is forensic serology, i.e., analysis of physiological fluids for genetic markers, such as ABO antigens, enzymes, and serum proteins. The major drawback of these analyses is the degree of specificity provided. The usual battery of serological tests might still allow characterization of a person only as a member of a larger population sharing the same markers. Depending on the panel of tests, the likelihood that a randomly selected person