These sound like the words that made your mamma wash out your mouth with soap:

"Bivariate ecological regression analysis."

"Multivariate analysis."

"Homogeneous precinct analysis."

"Ecological inference."

The phrases refer to statistical techniques used to analyze voting behavior.

While tedious if not torturous, they go to the heart of the federal voting rights lawsuit filed by five American Indians from the Wind River Reservation against Fremont County, which is on trial before U.S. District Court Judge Alan Johnson in Casper.

Fremont County has maintained the dispute centers not on race, as required by the Voting Rights Act, but on myriad issues related to the history and political situation of American Indians, said Scott Detamore, lead attorney for the county, with the Denver-based Mountain States Legal Foundation.

Some mind-bending statistical heavy lifting was necessary during testimony Monday and again Wednesday to determine why candidates preferred by Indians usually do not win elections.

A key federal voting rights case discussed this week, "Thornburg v. Gingles," states those who claim an elections system violates Section 2 must show three things:

- The minority group is big enough and geographically compact enough to make a majority in one or more single-member districts.

- The minority - in this case the Indians on the Wind River Reservation - is politically cohesive, or votes as a bloc.

- The majority - in this case non-Indians in Fremont County - votes as a bloc usually to defeat candidates preferred by the minority.

The Gingles case also requires a look at the total circumstances in voting, such as a history of discrimination.

Detamore and Laughlin McDonald, the Indians' lead attorney, hired statistical experts in voting practices - Steven Cole for the Indians and Ronald Weber for the county - to testify at the trial this week.

By studying the 1996, 2000 and 2006 general elections, both Cole and Weber found Indian cohesion, but Weber found less cohesion than Cole.

On Wednesday, Weber said he agreed with Cole generally but disagreed with Cole's analysis of racial polarization - that Indians voted cohesively one way and non-Indians voted cohesively another way.

Cole, who testified Monday, cited one contest in Fremont County where three Indian-preferred candidates were on the ballot, but only one won.

Voters in Fremont County would have shown their voting patterns were not racially polarized if all three of those candidates won, Cole said.

Weber responded Cole set too high a standard, and the analysis should be done differently.

"I think it should be candidate-by-candidate," Weber said.

The trial arose because Patricia Bergie, Pete Calhoun, Gary Collins, James E. Large and Emma Lucille McAdams - members of either the Eastern Shoshone or Northern Arapaho tribes - claimed the at-large, or countywide, method of electing county commissioners dilutes Indians' voting strength in violation of Section 2 of the Voting Rights Act and the 14th and 15th Amendments.

The want an election for a redistricting plan for single-member districts, because the size and population of the reservation is such it would constitute a single-member district, according to their federal court complaint filed in October 2005.

If that doesn't work, they want the court to implement a redistricting plan, according to the complaint.

The tribal members only have to prove the at-large system dilutes their political power according to Section 2 of the Voting Rights Act, and not whether it was caused by racism, said McDonald, who is lead attorney for the Indians and from the Atlanta office of the American Civil Liberties Union Foundation.

Reporter Tom Morton can be reached at (307) 266-0592, or Tom.Morton@casperstartribune.net.