Strong Inference 1964 - John R Platt

Systematicity

"These rapidly moving fields are fields where a particular method of doing scientific research is systematically used and taught, an accumulative method of inductive inference that is so effective that I think it should be given the name of ‘strong inference.’"

Steps to strong inference

  1. Devising alternative hypotheses
  2. Devising a crucial experiment (or several)
    • with alternative possible outcomes
    • each will exclude one/several of the hypotheses
  3. Carrying out the experiment
  4. Recycling the procedure

“We may write our scientific papers so that it looks as if we had steps 1, 2, and 3 in mind all along. But in between, we do busywork”

Inference through exclusion

“Any conclusion that is not an exclusion is insecure and must be rechecked”

“science advances only by disproofs

“a theory is not a theory unless it can be disproved

“we need to try to formulate multiple alternative hypotheses sharp enough to be capable of disproof.”

4 tips by Platt

“be explicit and formal and regular about it [the inductive method]”

Be problem-oriented

vs. method-oriented

be systematic

“devote a half hour or an hour to analytical thinking every day”

be explicit

“write out the logical tree and the alternatives and crucial experiments in a permanent notebook”

be excluding

  • ask “what experiment could disprove your hypothesis?” or “what hypothesis does your experiment disprove ?”
  • ask yourself : Conflicts are best to be had between ideas rather than (wo)men

New Project

sentence-level combinatorics

  • which elements are integrated with each other?

  • in what order?

  • what are they integrated into? -> RQ: are there intermediate abstract representations?

passive/active sentences

Vorige avond heeft de kook de kat getrapt

Last night the chef kicked the cat

Vorige avond werd de kat door de kook getrapt

Last night the cat was kicked by the chef

premise

We compute intermediate abstract representations (i.e. ‘agent’ and ’patient) as we integrate sentence elements.

Hypothesis

Word order and verb passive markers will be integrated into an “agent” concept independently of word semantics.

-> If true, neural pattern when reading first and second noun will differ early on

-> neural patterns will differ whether related to semantic content or “agent” concept