The model rests on an idea called the lexical hypothesis, which we touched on at the start. Put plainly: the personality characteristics that matter to people in everyday life tend, over generations, to become single words in their language. If a trait is socially important enough that we keep needing to point it out in others, a word for it tends to stick. Collect those trait words, study how they group together, and the major dimensions of personality should reveal themselves in the structure.
Through the 1990s and 2000s, researchers ran exactly this kind of study in language after language. They gathered the trait adjectives of Korean, French, German, Italian, Dutch, and others, and analysed how the terms clustered. The striking result was that these analyses repeatedly recovered six factors rather than the five that earlier English-language work had emphasised. The sixth kept turning out to be the Honesty-Humility material.
That convergence is what prompted Kibeom Lee and Michael Ashton to formalise the model and to build an inventory for measuring it: the HEXACO Personality Inventory-Revised, or HEXACO-PI-R, which is the questionnaire Hexis is based on. The official inventory, along with its scoring keys and supporting materials, is made freely available by its authors at hexaco.org for anyone who wants to read further.
We have described this history in plain terms on purpose, and have avoided dressing it up with invented citations or statistics. The short version is that the six dimensions were not designed at a desk and imposed on the data; they emerged from the data, in multiple languages, and the model was written afterwards to account for what kept appearing.