A fine balance: Encouraging imagination, evaluating evidence
In the wake of 9/11, critics bemoaned the intelligence community’s failure of imagination. Greg Fyffe, executive director of the International Assessment Staff (IAS) with the Privy Council Office, suggests the phrase is unhelpful in understanding the role of the analyst to weigh evidence and draw useful conclusions in an unbiased, policy-neutral and professionally fearless manner. Without evidence, he suggests, imagination is of little help.
Imagination has many meanings for the intelligence analyst. It can mean “seeing significance,” a synonym for insight, the quality that makes an analytical memorandum worth reading. It can mean “understanding context,” the need to lift our eyes from the immediate subject and see how events fit into wider and wider patterns.
It can also suggest “complex deduction.” Intelligence organizations, for example, know that one consequence of global warming is rising ocean levels. This will lead to lowland flooding and population movement. Where will it happen? Where will those populations go? How will governments react as environmental crises become more threatening, and possibly act as drivers of terrorism or war? Possible answers to these questions lie in deductions based on our knowledge of actions and reactions in natural systems, and our imaginings of possible human reactions.
Imagination is also the ability to see alternative explanations for the evidence in front of us, another way of describing the basic challenge of understanding what the evidence suggests is happening. Is North Korea seeking ways to engage with other countries, and willing to put its nuclear testing program on the table? Or is it endlessly looking to squeeze out concessions without any intention of giving up anything at all? We have evidence, but we must imagine its meaning because real intentions are unknown or unformed.
Imagination may be purely speculative, linked to no evidence whatever, but leading us to ask unanswerable questions and think about distant future possibilities. What will the armed forces of the major powers look like in a world in which fossil fuels are scarce and crushingly expensive?
Imagination can mean the intellectual ability to construct alternative future scenarios, or to see how unpredictable events, such as the assassination of a leader or a catastrophic natural event, can upset all calculations. We have no evidence on such important future questions as Cuba without Castro, Chavez when the oil runs out, or Pakistan if it returns to a more robust form of power-sharing democracy.
We need to imagine diverse possibilities, see their implications, and understand the evidence that would give us a clue on the direction in which events are moving. Imagination can describe the skill of assembling widely scattered events and signals, and seeing in them a definite pattern and the intentions of an adversary.
All of these interpretations are positive – we would like our analysts to have the capacity to do all these things. But imagination can also mean invention – even fabrication. It can signal paranoia, or the undisciplined mind that multiplies possible dangers and is unable to sort out which are of the highest priority.
Analytical agencies are experimenting with different ways of weighing evidence. Analysts have always looked at alternate hypotheses, and examined signposts to test for the advent of forecasted events, but more detailed methodologies are now emerging. Analytical organizations try and ask more “what if” questions, and recognize that the first step in preparing an imaginative analysis is asking an imaginative question.
Accepting that imagination is a necessary part of intelligence analysis brings us to two central questions. First, how do we encourage analysts to be imaginative? Second, what is the relationship between imagination and evidence – what is the place of analytical intelligence products that are based on imagination rather than evidence?
Ideally, encouraging imagination takes place throughout the assessment process as a normal part of putting a paper together.
Defining the analytical questions requires us to see what is important and what will add value for readers. Question definition may come from the client, with or without input from us, or may come from the analyst or director.
Throughout the process opportunities arise to challenge the thinking of the analyst. An IAS paper usually goes in draft to missions abroad and to the Foreign Affairs regional desk. Papers are discussed at an interdepartmental meeting of experts. This often brings forward alternate understandings. Of course, there is a back and forth between the analyst and director all through the process.
The director’s role, in particular, is to challenge the evidence and to test whether it can be interpreted in different ways. I also look at how the evidence has been used, and whether the conclusions are reasonable in light of that evidence.
We explicitly test alternative hypotheses to see the degree to which different interpretations can be best supported by the available intelligence. This is valuable when there are obviously different directions that events might take.
Inviting analysts to debate very far-reaching topics during a retreat also encourages imagination. No specific analysis results, but analysts see that it is important to look at events in many contexts, follow global trends, and try to look ahead and speculate on outcomes.
Some intelligence systems with more resources are able to invest more in imagining – looking for different interpretations of events, different motivations for key actors, a better sense of what the scenario experts would call wild cards.
But what are we really getting at when we talk about the failure of imagination, and to what extent do we want imaginative assessments going to policy makers?
The phrase failure of imagination comes from the US Congressional 9/ll report, and specifically refers to the suggestion that had analysts considered the past attack on the World Trade Centre, the Bojinka plot to blow up airliners in flight, past talk about crashing airplanes into buildings, and evidence of students taking flying courses with no interest in how to land, someone should have seen the outline of the Twin Towers attack.
Many of the classic intelligence failures amount to a failure to correctly assess the intentions of an adversary. In some cases the adversary successfully camouflaged his intentions; in some cases the signs were there to see but were misinterpreted. In other cases there was insufficient evidence to make a judgment.
Sometimes the analysis has been right, but the client could not be moved from his or her own set of expectations.
· Stalin discounted all warnings of a Nazi attack on Russia despite the gathering of troops on his frontier and allied warnings;
· the German military was fooled about the location of the D-Day landings by elaborate and complex deception plans that built on the British Double-Cross system and the creation of a phantom army headed by General Patton;
· General MacArthur refused to believe China would cross the Yalu into Korea;
· Israel did not believe the Arab armies would attack in 1973, in part because Israelis could not see a sensible rationale;
· the American military was aware of the possibility of a surprise Japanese attack after negotiations broke down, but did not expect it so soon, or at Hawaii.
Each of these cases could be described as a failure of imagination. Stalin arguably had all the evidence anyone would need of an approaching German invasion. MacArthur refused to accept the significance of troop buildups across the Yalu River. Pearl Harbor and the Normandy invasions were designed to be surprises. Exceptionally good intelligence gathering as well as imaginative interpretation would have been needed to see through enemy deception.
Imagining too much is also an intelligence failure:
· the “missile gap” on which John Kennedy campaigned in 1960, reflecting fears of the military, was imaginary;
· the expectation that Cubans would rise to support the Bay of Pigs invasion was completely mistaken;
· Iraqi WMD.
Intelligence analysis will reach the wrong conclusions if we do not see all the possibilities, but it may equally fail if we imagine and act on possibilities that the evidence does not support.
The challenge of the analyst is to appreciate the different possibilities and search for further evidence to confirm a hypothesis, or to alert the decision-maker of alternate possibilities that must be prepared for. It is important to be rigorous in evaluating evidence, and not rush to confirm a favourite hypothesis.
Often when we talk about analytical failure, we could as easily talk about collection failure, or perhaps system failure – the right intelligence was not collected, or if collected was not disseminated, or if disseminated was lost in the noise. If the right material has been collected and made available to analysts, they still might get it wrong, but we need more precise descriptions of why this may have occurred than “failure of imagination.”
A distinction also has to be made between imagination that alerts us to possibilities, and imagination as warning. Imagination as warning is not useful unless it is backed by evidence.
I can speculate about where terrorists might strike in Canada, and events could prove me correct. But why would anyone shut down the Toronto subway system, or fence off the CN Tower, or close train stations, unless I had some solid reason based on evidence?
The value of imaginative or speculative analysis, where there is no evidence, is to point to possible events, and to suggest the signals or signposts that could suggest that a speculative scenario might develop.
Understanding intent, as opposed to capability, is one of the most difficult challenges in intelligence analysis. Collecting security intelligence on possible terrorist attacks is even more challenging, because a small number of people are likely involved, and there may be no pervious signal that these individuals need to be watched.
Imagination drives the overall consciousness of security officials on what steps need to be taken to protect the public and public places – what are the possible targets, who needs to be watched, what defensive measures have to be taken. Again, this needs to be complemented by intelligence and analysis, or the possibilities are legion and the best hope rests on being lucky.
The phrase “a failure of imagination” is unhelpful in understanding the challenge that analysts face. It’s vague enough to hang a saint and no use in saving a sinner. The heart of good intelligence analysis is the professional weighing of evidence, and the use of careful reasoning to estimate what that evidence means. If the evidence is missing, imagination won’t save the day.
Greg Fyffe is executive director of the International Assessment Staff with the Privy Council Office. This article is adapted from a presentation to the Canadian Association for Security and Intelligence Studies conference in September.