Politicians, academicians and lawyers who met here for a weekend conference on political parties divided sharply along partisan lines on whether courts should restrict gerrymandering, but agreed that the creation of a southern regional primary is a bad idea.

The two issues -- unrelated, except that both are timely -- dominated the agenda of the quadrennial conference sponsored by the Institute of Politics at Harvard's John F. Kennedy School of Government and the Republican and Democratic national committees.

RNC Chairman Frank Fahrenkopf opened the gerrymander debate with a plea that as a matter of democratic principle and political comity, gerrymandering should "go the way of the spittoon."

He called for courts to employ a variety of "objective" standards -- such as the compactness of a district, the "community of interest" of its residents, the respect that the lines show for existing political or geographic subdivisions -- to determine whether legislatures had drawn lines to maximize the number of seats the party in control of the legislature would win.

But he said that if the Supreme Court does not venture into the "thicket" of gerrymandering, Republicans will try to stick it to Democrats after the 1990 reapportionment, as they say Democrats did to them after 1980.

"There will be a temptation to follow the old axiom that he who has the gold makes the rules," said Fahrenkopf, who predicted that his party will control a majority of state legislative chambers by 1991. Democrats hold a 2-to-1 edge.

Democrats here responded that they would prefer to take their chances with legislatures. Bruce Cain, a law professor at the California Institute of Technology, said that any "objective" standards a court might employ are biased toward Republicans.

He noted that Democrats tend to live bunched tightly in inner cities and that a compactness standard would place them in geographically small, heavily Democratic districts -- thereby consigning the party to the "wasted votes" that come with districts that are 70 percent or 80 percent Democratic.

Moreover, he argued that a standard that measures the fairness of a redistricting plan by asking whether the votes a party receives are out of sync with the seats it wins is unfair because Democrats represent more "socially disadvantaged" constituents who do not vote.

On another subject, the movement under way among southern legislatures to bunch their primaries and caucuses together in the second week of March 1988 drew negative reviews on the grounds that if the aim is to help nominate a centrist, southerners are seizing on a solution that could backfire.

Donald Fowler of South Carolina, chairman of the Democratic Party's "Fairness Commission," said the South is no longer demographically much different from the rest of the country. And DNC member Mark Siegel noted that in 1984, when the Rev. Jesse L. Jackson was a candidate for the Democratic presidential nomination, black turnout in many southern primaries was at record-high levels -- 41 percent in Alabama and 32 percent in Georgia, according to television network exit polls.

He also argued that, rather than decreasing the impact of the Iowa caucuses and the New Hampshire primary, a southern regional primary would increase their importance. The candidate who surpasses expectations in those two states gets a slingshot of favorable news media attention that, according to poll findings cited by Siegel, tends to peak within a week to 10 days -- precisely when the southern primaries would be held.

Other conferees said a southern regional primary would set off an avalanche of early primaries and caucuses. "We may be heading unintentionally towards a national primary . . . and that's a bad thing" both for the selection process and for the parties, said institute director Jonathan Moore.

Richard Lodge, Democratic state chairman of Tennessee and a proponent of the regional primary, said the change "couldn't be any worse than the current system."