From Brazil to Iran to South Korea, the one phrase that crops up in every conversation is 'transfer of technology."

The idea of picking the brains of the industrial countries, and using their know-how as a base on which to build for the future, has become something of an obsession with a number of developing countries.

The Brazilian government agreed in 1975 to buy eight atomic power stations from West Germany for an estimated $10 billion-even though it knew it undoubtedly could purchase them cheaper by putting them out for bids individually-in return for German agreement to supply Brazil with nuclear technology.

"The leadership of Brazil said, 'What we are doing is starting from the shoulders of giants,'" recalls Jose Goldemberg, chairman of the Brazilian Society of Physicists. "The leaders said: 'The Germans are transfering all the technology, so we'll start from that point and we'll become very smart, and we'll become independent between 1990 and 1995.'"

The Brazilian motivation in seeking this "technology transfer" makes a lot of sense.

Minister of Mines and Energy Shigeaki Ueki notes that for more than 30 years. Brazil has been buying generators and turbines for its hydroelectric plants from France, West Germany, the Soviet Union, Japan, Sweden and the United States.

"After 30 to 40 years, we will still don't have our technology and we still depend on the importation of certain parts," says Ueki. "If we had in the past a policy of technological transfer, maybe today Brazil would be one of the more important countries exporting hydroelectric turbines and generators around the world."

But a number of scientists like Goldemberg-all ardent supporters of the concept of transfer of technology, think the Brazilian government's approach is not the right way.

"What kind of training are we having?" asked Goldemberg. "The government says, 'Well; we sent some engineers to Siemens.' And what are they doing? They are learning how to copy blueprints.

"That's not transfer of technology," argues Goldemberg. "These engineers are not learning why the size of the machinery is the way it is. That's the real know-how."

Rogerio Cerqueira Leite, former head of the physics department at the University of Campinas, agrees.

"Some of the Brazilian engineers are learning how to operate the units in Germany," Leite says, "but that has nothing to do with absorbing the technolgoy."

Many scientists and engineers think the only way a developing country can achieve a true transfer of technology is to participate right from the outset in designing the nuclear power station it is buying.

"But you can't do that in buying large nuclear power plants from companies like Westinghouse and Germany's Kratwerk Union," says Reza Khazaneh, president of the Isfahan Nuclear Technology Center. "The commercial secrets are of such value that understandably, they are not willing to give you their know-how." stSo Iran - which like Brazil is determined to achieve this transfer of technology - is considering the intriguing notion of building a small nuclear power plant as part of the Isafahan center.

"We went to Belgium, which is operating a small nuclear plant in the nuclear center of Mol near Brussels, and together we have studied the feasibility of constructing a small nuclear plant here," Khazaneh says.

Khazaneh says that for a plant this size - which would cost about $200 million and have less than 3 percent of the electrical output of the big atomic power stations under construction - Iranians could be involved in everything from the design up through construction.

"We can take the responsibility for such a small facility," he says. "If we want to get a spectrum of this technology, we can't do it through purchase of large nuclear power plants. It's absolutely impossible."

Khazaneh also notes that a small nuclear power plant of this kind would be useful in training the people who will operate Iran's large atomic power station.

"The alternative is a simulator, but a simulator is a toy," he says. "There are functions in nuclear power plants that you can never learn with a simulator. If you want to face hazards of all kinds, you have to have a real power plant."

Whether this scheme, in light of Iran's current difficulties, will ever be approved is at this point an open question.

But Khazaneh, who thinks the project has a 50-50 chance, remains hopeful.

"It means getting a deep know-how into the technology," he says, "and not remaining a buyer forever from advanced countries." CAPTION: Picture, South Africa's Safari research reactor: 'Getting a deep know-how.' South African Atomic Energy Board