"You can't be a student and plagiarize. Are weto say that it's O.K. for a professor?" Tostesontold The Times from Denmark.

Tosteson did not respond to several telephonecalls from The Crimson yesterday after his returnfrom Boston.

Professor Richard Marius, director of theExpository Writing Program, said that "If Frazierhad been a freshman, he would have been [asked toleave] for a year. It is a very serious offense,not to make some kind of acknowledgement."

"Academic institutions live on mutual trust andcertain known criteria of integrity. The wholeinstitution falls when someone does somethingwhich betokens a lack of integrity," said Marius.

But colleagues charged that Frazier wasunfairly being used as an example. "The wholeaction was out of proportion," said Dr. JohnRomano, a professor of psychiatry at theUniversity of Rochester. "It is a reflection ofHarvard's anxiety over its past record." Romano,who said he has known Frazier for many yearscalled Frazier "a man of integrity."

Tseng Case

Romano and others suggested that Harvard'sactions were in response to criticism last monththat the Medical School was too slow to act inanother issue of scientific ethics.

The Boston Globe reported in October thatformer ophthamology fellow Dr. Scheffer C.G. Tsengmisrepresented results of his research on an eyemedication and profited from the sale of stocks inthe company that manufactures the drug.

Tosteson knew about the misrepresentation atleast four months before The Globe reported thecase, but did not act on the matter, according toa report by Harvard Watch, a student groupaffiliated with consumer advocate Ralph Nader.

`Sloppy'

Frazier referred to papers that do not exist inone article, and cited sources correctly in somepapers but later attributed them incorrectly, saidScatena.

"He really made sloppy mistakes," said Scatena."The more I studied, the less I understood him."

"He may have been set in the mode of `I gottapublish.' It is easy to get caught up in it," saidScatena.

Statistical Errors

There were also statistical errors in Frazier'sarticles, Scatena said. Frazier derivedpercentages from other researchers' dataincorrectly by using too large of a test group, hesaid.