By analyzing minute emotional reactions in real-time during normal conversation, FIML practice disrupts the consolidation, or more often the reconsolidation, of “neurotic” responses.
In FIML, a neurotic response is defined as “an emotional response based on a misinterpretation.” The misinterpretation in question can be incipient (just starting) to long-seated (been a habit for years).
The response is disrupted by FIML practice and, thus, tends not to consolidate or reconsolidate, especially after several instances of learning that it is not valid.
A neurotic response is a response based on memory. The following study on fear memories supports the above explanation of FIML practice.
Memories become labile when recalled. In humans and rodents alike, reactivated fear memories can be attenuated by disrupting reconsolidation with extinction training. Using functional brain imaging, we found that, after a conditioned fear memory was formed, reactivation and reconsolidation left a memory trace in the basolateral amygdala that predicted subsequent fear expression and was tightly coupled to activity in the fear circuit of the brain. In contrast, reactivation followed by disrupted reconsolidation suppressed fear, abolished the memory trace, and attenuated fear-circuit connectivity. Thus, as previously demonstrated in rodents, fear memory suppression resulting from behavioral disruption of reconsolidation is amygdala-dependent also in humans, which supports an evolutionarily conserved memory-update mechanism. (Source: Disruption of Reconsolidation Erases a Fear Memory Trace in the Human Amygdala)
FIML practice works by partners consciously and cooperatively disrupting reconsolidation (and initial consolidation) of neurotic memory (and associated behaviors). FIML both extirpates habitual neurotic responses and also prevents the formation of new neurotic responses through conscious disruption of memory consolidation.
FIML probably works as well as it does because humans have “an evolutionarily conserved memory-update mechanism” that favors more truth. Obvious examples of this update mechanism can be seen in many simple mistakes. For instance, if you think the capital of New York State is New York City and someone shows that it is Albany, you will likely correct your mistake immediately with little or no fuss.
Since FIML focuses on small mistakes made between partners, corrections are rarely more difficult than the above example though they may be accompanied by a greater sense of relief. For example, if you thought that maybe your partner was mad at you but then find (through a FIML query) that they are not, your sense of relief may be considerable.