Monday, March 26, 2018

The challenge of marrying experiential and evidentiary knowledge

Manu Joseph has this very nice article which distinguishes between the arrival of an idea and its justification, their relative merits and importance, and the way in which society perceives each,
An idea comes to some minds as an intuition, to many in the form of faith or imitation, or a convenient corroboration of a bias. An idea always arrives as a realization, spreads as a belief. The arrival of an idea is a religious moment. But its legitimacy is proved in public and private through the fabrication of rational substantiation. An argument then is reverse engineering of a religious moment. Here I am not referring to collegiate people who can debate either way, or are paid to debate for a cause in television studios. I am only referring to people who have ideas, or at least convictions. Even when they practise it, is debate as intellectually robust and pure as we are trained to assume? Isn’t it true that all debates emerge from the scripture of personal faith? Is the pre-eminence of debate then overrated? In the hierarchy of intellectual activities, why is this method of transmission of an idea more respected than the very force that creates ideas—intuition? An intuition is not a supernatural force—it emerges from dormant or subterranean knowledge. Even so, science celebrates intuition only after it has been proven to be right. Can it be that across the ages, superior thinkers have been subdued by lesser minds who were and are merely good debaters? Is the transmission of truth now entirely in the hands of the articulate, who are better at transmission than truth?...
When we debate, argue, or even write a column, we build a case, we substantiate our argument and consider opposing views. There is one thing we do not say at all—how we actually got the idea. Usually, an idea does not come to us after an argument with ourselves, or after a deep investigation into the facts. This is not how ideas usually arrive, or form. The argument does not create the idea, the idea creates the argument.... But the imitation of scientific debate in politics, economics, culture, even literature and other aspects of the subjective arts, is outrageous. In television studios and around dinner tables, people are forced to dress up their intuition or beliefs through the masquerade of logic and evidence. That is a wasteful decorum of modern intellectual life.
This has resonance with the ongoing obsession with argument in the form of evidence generation among development cosmopolitans (not practitioners) residing outside the development context. I have a plausible explanation for this in the case of international development. 

All of the doing in international development is done by the practitioners from inside, especially the government officials, for-profit social enterprises, and non-profits working in the field. A lot of the thinking about and funding of international development is done by outsiders or foreigners. There is a vast difference between how the two parties process information. 

Both sides are equipped with the analytical framework and evidence, the latter much more so than the former, though sufficient for the former to make informed enough comparative and objective judgements. Let us call this the evidentiary knowledge. Then there is experiential knowledge, which comes from a deep understanding of the context and its nuances, informed by a rich and diverse set of experiences accumulated over a long period of living and working in those contexts. By the very nature of their respective nativities and careers, the former internalise this deeply, whereas the latter are sorely deficient. Whether we like it or not, that's the way the world is and we need to accept it. 

Accordingly, the former discounts their evidentiary knowledge base with their experiential knowledge base to make judgement calls. This IT App or innovative product (in any sector) with an RCT to back its efficacy looks logically great but the practical challenges are too many that it is unlikely to make a dent on the problem which is significant enough commensurate with its opportunity cost when compared to alternative efforts (a very relevant trade-off in capacity and resource constrained environments). In contrast, the latter, with an evidentiary knowledge base which is as rich as their experiential knowledge is poor, process information overweighting evidence and logic. So the same IT App or innovative product, with its logical neatness and rigorous enough evidence, looks very appealing.

None of this is to overlook the inevitable human behavioural failings associated with both sides. In the case of the former entrenched priors socialise out evidence and manifest as prejudices. In the latter, logic crowds out reality and manifest as algorithmic reasoning.   

In conclusion, we have the classic tension between inductive and deductive inferences. The former uses inductive inference that draws on a rich set of priors to exercise judgement. In contrast, the latter uses deductive inference to rationalise from a logically drawn and rigorously proven theory of change. 

The gulf is actually much bigger. While talking of experiential knowledge, we grossly underestimate its richness and value. One only needs to keep in mind Karl Polanyi's description of tacit knowledge, "We can know more than we can tell".

1 comment:

Unknown said...

Yes I observed this in so many debate. Just like this one

Niti Aayog chief is talking about incentive based approach (naming and shaming) I think in Indian context that's better possible option. This blog repeatedly explain structural weakness and lack of political will power is main reason for administration mismanagement at ground level. while keeping in mind inter-state and intra-state capacity differences this approach may not bring major change but it will surely catalyse process and create breeding ground. Just like ambitious target kept by indian government
100GW solar power generation by 2022

All I can understood from this is India or any real time action need some 'Unconventional NUDGE'