Last week, Washington officials trekked to the Desmond Americana Hotel in Albany to hear witnesses give their versions of how and why a bridge carrying the New York State Thruway over Schoharie Creek collapsed on April 5, killing 10 persons.

With that kind of death toll, the National Transportation Safety Board hearing attracted a lot more attention than many of the judicial and quasijudicial inquiries into why a building project has failed or run into trouble.

But it wasn't all that different. In a litigious society, such attempts to pin down the blame in construction accidents are increasingly common -- and are spawning an entirely new branch of technology. Business executives sued (or worried about being sued) in the wake of such an accident now turn to a forensic engineer. There's a lot of controversy over whether the development of a specialty of applying engineering techniques to the particularized needs of litigation is a good idea.

"Turning testifying into a business has real potential for subverting the system," warns Arthur D. Kornblut, head of the Construction Law Forum of the American Bar Association. there's no doubt that it has happened.

The National Academy of Forensic Engineers, formed just four years ago, already has 250 members. The Association of Soil and Foundation Engineers is almost through drafting ethical recommendations to cover members' testimony in adversarial proceedings. The American Society of Civil Engineers has embarked on a similar project. And a construction news weekly published by McGraw-Hill last month featured the new specialty in a cover story titled "Experts for Hire."

Essentially, forensic engineers serve two functions in a liability dispute. Lawyers count on them to sift through all the evidence and give an honest reading of just how much the client is responsible for what happened.

"If the view is totally adverse, the information is still helpful," insists New York construction lawyer Robert A. Rubin. "We may want to settle out."

But when culpability is not so clear-cut, it is the job of the forensic engineer to appear in a hearing or court as an expert witness and make a convincing case for why the client should not be held liable. Just being ready to make such a presentation may be enough to persuade the other side to accept liability and not risk letting the money damages be set by a jury.

It is estimated that no more than 5 percent of the controversies in which forensic engineers are employed actually go to trial. When they do, though, most often the fact-finders have to choose between conflicting interpretations from opposing teams of equally impressively credentialed witnesses.

The idea of engineers taking part in a trial is not new. But the recent booming market for forensic science has moved the ethical considerations to the forefront.

Engineers like to believe that they approach a problem objectively and find the "true" solution. That ethos fits uneasily into the adversarial process, where each side is expected to make the strongest possible argument and it is up to the umpires to choose between those presentations.

The National Academy, for instance, began by adopting the existing ethical code of the National Society of Professional Engineers. It demands that engineer be "objective and truthful" in all reports and testimony.

But in the commentary on that provision, the new group notes: "When retained by one party to litigation, the forensic engineers have a duty to that client." In practice, says a draft report from the academy's ethics committee, that means giving the facts but not volunteering any opinions that are adverse to the person who is paying the bill.

Another ethical key is to do real research, especially historical research. Some professional groups are worried about expert witnesses testifying about the most up-to-date knowledge and techniques in an accident involving an old structure. You must be careful to talk about what accepted practice was at the time the project was designed, the soil engineers group tells its members.

Oddly enough, the more an engineering firm specializes in helping out in lawsuits, the less attractive it is to the lawyers. An engineer who is handling construction projects for clients most of the time may bring a broader understanding to investigating the causes of an accident. And a jury may rely more on such testimony than on that of someone who earns his keep by appealing to lawyers.

The jury is perceived as putting the highest trust in forensic engineers who are not hired by either side -- in other words, in government investigators.

The National Bureau of Standards has a cadre of building-technology experts who look into the reasons for accidents. They are, for instance, investigating the collapse in April of an apartment complex going up in Bridgeport, Conn.

The bureau personnel will not assign blame, but they will testify in a trial on their findings. That kind of competition tends to worry some of the new for-profit experts. They say that while their conclusions may be just as sound as those from Washington, the laymen who make up juries are likely to put more weight on the views of the bureaucrats