OTTAWA — If Louis Riel had been hanged in 1885 because of an informer among his Metis rebels, Canada's spy agency might still be blocking release of that history-changing revelation 125 years later.
That hypothetical scenario was put to the Canadian Security Intelligence Service last week as part of a legal battle over the government's refusal to fully disclose decades-old intelligence gathered on socialist icon Tommy Douglas.
The agency couldn't say for certain whether it would release the identity of Riel's hypothetical betrayer or withhold the information on the basis of national security.
The equivocal response offers a glimpse into the lengths CSIS will go to protect the secrets of the spy trade, no matter how old or historically significant.
Nicole Jalbert, Access to Information co-ordinator for CSIS, testified she routinely nixes disclosure of intelligence files that would reveal confidential sources or intelligence agents, even if they're long dead and the information gathered decades ago.
She said that rule applies to "more contemporary times" - including the files on Douglas, gathered as far back as the 1930s.
Such files must remain secret "perhaps not forever" but "maybe longer" than 100 years.
Paul Champ, lawyer for The Canadian Press reporter who initiated the legal battle over the Douglas dossier, pressed Jalbert to specify how much time must elapse before CSIS considers it safe to release such information. He wondered for instance what she'd do with a file that suggested Riel had been betrayed to the Northwest Mounted Police by fellow Metis leader Gabriel Dumont.
Government lawyer Gregory Tzemenakis called the question "hypothetical" and "unfair." Nevertheless, Jalbert said in a "very, very old" case like that she'd seek the advice of people at CSIS who deal with "human sources."
"I would leave it to their discretion to make the decision," she said during examination on her affidavit filed with the court in January.
"It's not mine to make when we're dealing with human sources."
In a case that is "very, very old like that, yes, it's not clear in my mind that I have to protect it so I will, I will take an extra step and see if we can release it," she added.
Jalbert distinguished between a "very, very old" case like the hypothetical Riel scenario and those like the Douglas case, which are "just old." In the latter instance, she said she routinely refuses to release the names of informants without checking with anyone else at CSIS.
Some of the intelligence was gathered before the Second World War, but she insisted the Douglas case is "still quite contemporary."
"These people still have families and reputations. How are you going to recruit somebody if you can't guarantee their anonymity?"
The battle over Douglas's intelligence file began in November 2005, when reporter Jim Bronskill made an Access to Information request for the RCMP dossier on the fabled prairie preacher who served as premier of Saskatchewan and federal NDP leader, and who is widely hailed as the father of medicare.
The information was gathered by the now-defunct RCMP security service, transferred to CSIS and eventually to Library and Archives Canada.
Some material was released to Bronskill, showing the RCMP had shadowed Douglas for more than three decades, attending his speeches and even eavesdropping on conversations. His links to the peace movement and members of the Communist Party were of particular interest.
But on the recommendation of CSIS, the archives refused to release large parts of the Douglas dossier, citing national security. The exemptions were upheld by the information commissioner of Canada so Bronskill took the minister of Canadian Heritage, who oversees the archives, to court to force disclosure.
Jalbert's affidavit said disclosure of the dossier could endanger informants and jeopardize the spy agency's ability to conduct secret surveillance. She argued that "the origin of information, its extent and the methods by which it was obtained must remain a secret."
During last week's examination, Champ questioned the qualifications of Jalbert, who started with the federal public service as a secretary in 1977, to determine whether the Douglas dossier should be fully disclosed.
Jalbert, who has worked as an Access to Information officer since 1984 with both Foreign Affairs and CSIS, acknowledged she has no expertise in the spy agency's operations or political history. But she was trained by veteran intelligence officers and, in any event, she said determining what shouldn't be released is "not rocket science."
"You don't have to be an intelligence officer working in that milieu to know what is sensitive and what isn't. ... Good judgment will take an individual very, very far," she said.
Moreover, Jalbert said, CSIS has developed guidelines when advising Library and Archives Canada whether a document should be released.
The guidelines specify the name of any RCMP officer linked to "a covert op" must be withheld. Apart from that, the guidelines say all names and ranks prior to 1945 can be released and outline a sliding scale of higher ranked officers who can be identified in later years.