MBTI Cognitive Functions Explained: The Layer Under Your Type

Here, “extraverted” and “introverted” describe the orientation assigned to a mental process in the model. They do not simply mean talkative and quiet. Sensing and Intuition concern taking in information; Thinking and Feeling concern evaluating it. Feeling is not a measure of emotional intensity, and Thinking is not an intelligence score.

The definitions below are an editorial paraphrase of the Foundation's Processes page. The questions are our own unscored prompts. Anyone can use any prompt.

These are perspectives, not separate talents. Brainstorming does not prove Ne, remembering an event does not prove Si, and caring for someone does not establish Fe rather than Fi. Training, role requirements and the situation can all affect what you do.

The following describes the Foundation's type-dynamics framework. It does not establish how much time a person spends using a process or how accurately a stack predicts a particular action.

The Foundation also discusses development and stress within its theory. We do not use an age schedule or a stress reaction to assign a function. A source explaining its own model is different from an independent study showing that its predictions work.

J/P indicates the kind of preferred process presented outwardly in this framework: J points to Thinking or Feeling; P points to Sensing or Intuition. I/E determines whether the dominant or auxiliary is the outwardly oriented preferred process.

For INTJ, J points to outward Thinking (Te). Because the code begins with I, that outward process is auxiliary; the model assigns introverted Intuition (Ni) as dominant. For INTP, P points to outward Intuition (Ne); with I, it is auxiliary, while introverted Thinking (Ti) is dominant.

This explains how the reference table is constructed. It does not mean changing a questionnaire answer switches all of your mental processes, or that people with different patterns cannot use the same skills.

Source: the Foundation's interactive Processes table, checked for all 16 codes. The tertiary column retains the alternatives it shows rather than presenting one popular fixed-stack convention as the only official version. The table is a model reference, not observed population data or a function-test result.

A four-letter code is the input to this model mapping. The resulting row does not independently validate the code. If you are uncertain about your questionnaire result, reading a row is not a second measurement.

These contrasts explain vocabulary; the examples are editorial, not observations of real typed people.

Ne/Ni is also not a contest between creativity and insight. You can generate alternatives and then develop one interpretation in the same task.

Choose a real, ordinary decision. Write what happened before selecting a function label. Repeat in a different context and include an alternative explanation. This record has no score and cannot verify a dominant function.

Fictional example: Alex is choosing a venue for a team event. Alex remembers that last year's room was too noisy, compares two alternatives, asks about accessibility needs, and uses a cost limit to choose. The record includes remembered experience, alternatives, people and external criteria. It does not show that Alex “is Si,” “is Fe” or has a particular stack. Trying a second venue decision with different constraints may reveal what was situational.

Use the record to notice an overlooked perspective: gather a direct observation, consider an alternative, clarify criteria, or ask an affected person. You can do this without knowing a type code or taking a test.

The Myers & Briggs Foundation is a primary source for the framework it describes. Its definitions and table explain what the labels mean within that framework. They are not an independent validation study of eight measured functions, a fixed stack's predictive accuracy, or this site's questionnaire.

This site's preview asks about four preference axes. It is not an eight-function measurement instrument, and a four-letter result does not demonstrate a brain mechanism. For measurement terminology, use the existing test-accuracy guide; reliability of preference scales and agreement on four-letter codes are different from validating a function hierarchy.

The useful output here is clearer vocabulary and a concrete record of your decisions. Avoid inferring intelligence, career ability, morality or another person's intentions from a function label.