Dr. James K. Harlan might be entitled to shake a finger and say ''I told you so,'' if that were his style.
Since 1979, when he was working in the policy office of President Carter's Department of Energy, he has had doubts that a major crash program in synthetic fuels was really what the country needed.
Instead, he has favored small-scale synfuel projects whose main value would be not so much the fuel they produced as the lessons they would teach. Start off with a small project, he counsels; see what you learn from that; then apply those lessons to the next project. Only practical experience with the technologies of synfuel production - of which there is little - will show ways to cut costs.
Over the past year, as the Reagan administration has put its mark on the synfuel program, he has seen his point of view gain acceptance. Edward Noble, chairman of the United States Synthetic Fuels Corporation, has picked up the term ''learning infrastructure,'' Dr. Harlan notes. The administration generally favors a small program with heavy private-sector involvement.
The Energy Security Act of 1980 (ESA), however, which set up the synfuel program, established some ambitious goals: the production of 500,000 barrels of synfuels a day by 1987, and 2 million a day by 1992.
''The ESA deadlines were set up to get you to move quickly, but they weren't realistic,'' Dr. Harlan says. The Carter administration wanted to show it meant business on practical production of synfuels, ''rather than just financing some little R&D projects,'' he adds.
But building such production capacity all at once, in effect, would mean having to go too fast to learn - and save - along the way. Production capacity would be built at top-dollar rates.
Dr. Harlan, now with the SFC, has just set forth his analysis in a book: ''Starting With Synfuels: Benefits, Costs, and Program Design Assessments'' (Cambridge, Mass.: Ballinger Publishing Company).
In it he argues that the fundamental forces of the oil market are still at work, despite the current glut and attendant price dips. A small-scale national synfuel program is sure to be a sensible investment, he maintains. Should oil supply lines from the Persian Gulf be cut, for example, the United States could draw on its ''learning experience'' from a small-scale synfuel program and expand it rapidly to make it an alternative source of fuel.
Government support is still needed, he says, because all this ''learning experience'' doesn't leave private companies with much they can make a profit from.
Before Exxon's recent pullout from its oil shale project in Colorado, Dr. Harlan heard a man at the much smaller Union Oil project next door comment, ''I don't see what they're doing at 50,000 barrels per day - they'll just make my mistakes five times over.'' Another drawback to a crash program, Dr. Harlan adds , is that there are only about 10 architectural and engineering firms in the country with the capability for designing synfuel plants.
Whether a moderate-size program would be a good idea, however, depends on whether one sees the odds of a sudden increase in oil prices - say as a result of Soviet action in the Gulf, for example - as great enough to justify the cost-inefficiencies of a larger program.
It's like buying insurance, Dr. Harlan says. If you perceive your risk to be larger, you buy more coverage - and pay a higher premium.
''Synfuel is so costly we hope we never have to use it, but we'd better learn to do it in case we have to,'' he says.
''A credible ability to provide synfuels in case of a national emergency,'' he adds, could deter US enemies from staging an embargo and also enable the US to supply its allies with fuel in case of war.
His book, which grew out of a report he did for the Department of Energy, includes a detailed economic analysis worked out with Harvard experts. Graphs suggest what would be the most effective synfuels program under various sets of conditions.
The push for synfuel production has been weakened by current low oil prices. Reports of companies pulling out of synfuel projects, as Exxon did, are much in the news. But the synfuel program can't be allowed to flip-flop with every twitch of the marketplace, Dr. Harlan maintains. Synfuel production, unlike other energy technologies, takes long lead times: five years to build a plant, and another five or 10 for the payoff. ''We need to develop a stable consensus for two decades to get this done.''