How do we know what we know? This is one of the transcendental questions that epistemology--the branch of philosophy concerned with the theory and origin of knowledge--seeks to answer. It’s not a question that people often think to ask, since they tend to take for granted the validity of their epistemological worldviews, but it’s critical that investigators think about its implications. When we trust a source of information we implicitly attribute authority to that source, that is, we leave ourselves open to the possibility that source may be wrong about the information communicated. This is what philosophers refer to as the Vulnerability Problem. It is a problem that, far from being merely theoretical, has very real implications for investigators and for those operating in complex markets like Latin America.
In spite of its general linguistic and cultural homogeneity (excepting Brazil), Latin America is not a monolithic region. Every country comes with its own complications. And life here is complicated, nothing is simple, easy, or straightforward, nothing is ever what it seems. A messy milieu of cultural norms/tendencies, written rules, unwritten rules, corruption, institutional weaknesses, geographic isolation materialized in poor infrastructure, lawlessness, power players, and politics permeates and affects everything. The first and most fundamental mistake outsiders make when approaching the region is to assume either that they understand how things work or that things work in the same fashion as in their home countries. In such an operating environment, we must be extra wary of the Vulnerability Problem.
The raw intelligence collected in most investigations can be divided into two categories.
Open source intelligence (“OSINT”): sourced from any openly available sources such as the Internet, media outlets, and public records.
Human source intelligence (“HUMINT”): sourced from human sources who are consulted.
Intelligence in both categories presents us with the Vulnerability Problem. For the purposes of this article, we’ll consider ways to approach this problem in the first category of information.
The Vulnerability Problem - An OSINT Example
In terms of OSINT, we principally encounter media reports, websites, blogs, databases (both official and unofficial), court filings, government files/archives, and government publications. In the first place, many people generally understand that there should be a “hierarchy” of open sources. That is, generally speaking, an official government report about X topic, for instance, should be accorded more weight than a blog or press report that discusses that government report. The government report in the example would be a primary source while the blog or press report would be a secondary source.
To take this example one step further, suppose that the government report in question was issued by the Government Auditor (“GA”), a government agency, and includes the following conclusion.
There is evidence of irregularities in the management of accounts at the Vehicular Licensing Agency (“VLA”) during the time period when Juan Perez was head of the VLA.
How would you assess the veracity of this conclusion?
What if you subsequently found a press report in The Herald, a high-circulation newspaper in the jurisdiction in question, that stated the following:
Juan Perez, former head of the VLA under the previous government of President Ayalo Azul, was arrested this morning and is being held pending trial in connection with allegations of corruption regarding contracts issued by the VLA to companies that manufacture vehicle registration decals. Current President Arveja Verde commented on the incident at a press conference today stressing her government’s commitment to root out corruption in government agencies.
Does the press report alter the way you assess the GA report?
Now think about the above example and consider just how skewed your assessment of the above situation would be if you did not know that:
Juan Perez had been appointed to the VLA by former President Ayalo Azul who is the political nemesis of current President Arveja Verde.
It is normal for incoming governments in Latin America to “clean out” agencies of prior government allies. It is also normal for incoming governments to go after the allies, appointees, etc of the previous government. It doesn’t matter if this is done under the mantle of “anti-corruption.”
It is normal for people to be jailed pending trial in much of Latin America, often, as a means of exerting pressure on those people or others.
The Herald is owned by Roberto Corneto, a businessman who is close to President Arveja Verde and supported her presidential campaign.
The current head of the GA was appointed by Arveja Verde. The GA has a history of being used as a tool to persecute enemies of the government.
The US and EU are pressuring Verde’s government and other governments in the region to “show results” in the battle against corrupt public officials.
Do the above factors change your perception of Juan Perez’s arrest? They should at least prompt some questioning of the official narrative.
And yet, it’s surprising to me how often investigators outside the region (primarily in the US and Europe) neglect to consider these sorts of factors. I once received a call from an American investigator asking for advice on investigating a subject close to the government in Panama. “I read that he’s under investigation for X,” said the American. “Where did you read that?” I asked. When he explained that he’d read it in a certain newspaper I commented that the newspaper was owned by political enemies of the current government, “Oh,” he said, “I didn’t know that.” That’s precisely the problem. In ascribing truth-values to information, one becomes vulnerable to the biases, knowledge gaps, errors, and so on, of the sources of the information, whether those sources are journalists or official government agencies. In Latin America, not all “facts” are created equal.
As such, when evaluating OSINT, at a minimum, we need to think about:
Biases and limitations: This applies in different ways to media and government/official sources. Remember that when you read a press report you’re often dealing with both the biases/mistakes/knowledge gaps of the journalist, the editor, and of the newspaper itself.
Possibility of falsification/missing information.
Apart from the above, one must also consider the limitations of OSINT. OSINT without HUMINT provides an incomplete picture. There are things one learns through HUMINT that would never be discovered through open sources. Suppose we found out, from reliable sources, that President Arveja Verde owns shares in a company that manufactures vehicle registration decals. Perhaps her "crusade against corruption" is a means of rescinding existing VLA contracts so that she can award them to her own benefit? One small bit of information can change the entire way we view a story. But, HUMINT also brings Vulnerability Problem considerations, such as source reliability, knowledge, veracity, biases, and constraints.
Another ingredient in properly mitigating the Vulnerability Problem in Latin America, which cannot be easily replicated, is the penetrating skepticism created by years of working investigations in and living in the region. All of us evaluate information through the lens of our own experiences. I would argue that without proper cultural immersion, language proficiency, and experience doing investigative field work in Latin America, outsiders will not develop the same point of view. There is, after all, a very real difference between theory and reality. In my work, I have seen cases of:
Patently false information in press reports and the purchase of falsified press reports.
Significant fraud schemes involving public officials and family members of high-ranking government officials who operate with impunity.
Cases of hit jobs ordered from prison by criminal organizations.
Persecution campaigns mounted against clients by powerful parties who can manipulate law enforcement authorities.
People jailed on false pretenses as a means of pressure either against their loved ones, their business associates, or as a means of extortion.
Cash deposits that would provoke a mild heart attack in an AML professional.
The key, I think, is that I have seen these things with an immediacy brought on by living in the region. These are issues that affect my friends, business associates, and clients; I have experienced them on a firsthand basis and they have shaped the way I view work in the region.
What is our job, as investigators, ultimately? Is it to simply gather up all these OSINT data points and present them to the client, or are we responsible for explaining to the client what those data points really mean? If it’s the latter, then we really need to start thinking about the Vulnerability Problem in our work and we need to be very careful that we understand the jurisdictions in which we are working. Otherwise, we’ll end up in the worst possible position: certain that we know something that we truly do not know.