The Methodology

The four-step deduction method

––––––

Deduction is often presented as a mysterious gift possessed by eccentric detectives, ageing professors, and occasionally by cats who appear to know things they should not know. This is unfortunate. Mystery is difficult to teach.

​

Method is easier.

​

The Deductivist teaches deduction as a repeatable process. Not intuition. Not instinct. Not "reading people". Those terms have generated enough confusion to require their own administrative department.

​

Every deduction begins with the same four steps.

​

01

Observe

Look before you interpret

​

Most errors enter the process during the first few seconds, usually disguised as efficiency.

​

A person notices a detail, immediately attaches a meaning to it, and then spends the next ten minutes defending that meaning from reality.

​

Do not do this.

​

Collect observable facts only. Clothing. Behaviour. Environment. Timing. Movement. Language. Anything directly available to the senses.

​

Nothing else.

​

A conclusion is not an observation. Neither is a feeling, a suspicion, a vibe, an impression, or whatever name intuition is currently using to avoid paperwork.

​

"It is a capital mistake to theorise before one has data."

​

02

Isolate

Separate signals from noise

​

Reality produces details at an alarming rate.

​

Most of them are innocent.

​

The objective is not to notice everything. That would be impossible and would also make grocery shopping unnecessarily dramatic.

​

The objective is to identify which observations carry information.

​

A scuffed shoe may indicate occupation. An unusual speech pattern may indicate education, region, or social environment. A blue tie generally indicates ownership of a blue tie.

​

Not all details deserve equal funding.

​

The question is simple:

​

If this detail were different, would my understanding of the person change?

​

If the answer is no, the detail may safely return to civilian life.

​

03

Hypothesise

Build the most likely explanation

​

At this stage, many people become emotionally attached to their first idea.

 

The first idea appreciates the support but remains statistically unqualified for the position.

​

Construct a specific explanation supported by the available evidence.

​

Not a story.

​

An explanation.

​

The distinction matters because stories are perfectly capable of explaining anything. This is why they are so popular and so dangerous.

A useful inference should be clear enough to be wrong.

​

"This person is likely..."

​

Good. Now continue.

​

04

Test

Attempt to prove yourself mistaken

​

Most people treat evidence as a voting system.

​

One clue suggests a conclusion. Another clue agrees. A third clue joins the coalition. Soon the conclusion is leading comfortably in the polls.

​

This is not verification.

​

Verification begins by searching for evidence that could contradict the inference. If the conclusion survives that encounter, confidence may increase.

​

If it does not survive, the conclusion is dismissed with appropriate professional courtesy and replaced.