It would be great if it were possible to have a core vocabulary of, say, 50 words. Realistically it might have to be 100, but that's getting more complex than a team, or a communicative practice, might want to be handling.
There is a need for a working procedure to decide - to evolve, to tune - which words are in, and which are out, as regards being allocated core roles in the work of the vocabulary.
> What's on this page needs modiifying, in the light of a shift in methodology. The keywords in this wiki are now derived from the 'backstories' of the patterns in the foprop pattern language. Revision of this page, to be done xxx
Some clue to how to achieve this selective focusing is here: >Address words as if they were persons, in living relationships with other words - family, employers, enemies, workmates, the neighbours, 'the authorities', friendship networks, etc. And, in these relationships, doing particular kinds of work, in particular segments of society.
So in some way, the 50 core words need to be the ones with the strongest, richest afiliations? The words that others 'look up to' or defer to, or are tacitly ruled by.
Aiming closer to 50 than 100 is an important discipline, because it will focus attention on the words that can do the most work, or the most beneficial work. But arriving at those 50 might involve considering 500, getting to know how they work, identifying the networks they travel on, and triaging them to see if they rank higher or lower in the influence hierarchy.
How long this takes, and how complex this is, are practical matters to be addressed early.
Looking at how Williams did it might help - but he'd been on the case for 30 years. Also, he was dealing with words that are much-used.
In contrast, many of our key words are little used? While some others are over-used, with many overlays of meaning. Money, for example!
>This is one reason why focusing on key forms of practice (which is what pattern language does) rather than pieces of language, can be an effective strategy.
The two should be run alongside one another?
Looking at how the Groupworks Collective developed their pattern language of group facilitation certainly will help. Although they were at work on pattern language rather than keywords, the process is basically similar I think.
>"Getting it down from \[115] to the "final" 91 patterns was brutal — relationships were broken in this endeavour"
They arrived at 91 patterns in eight families, after considering 700, and taking four years, collaborating across the breadth of the USA. See Patternwork in the Groupworks language
> More figuring needed. More to be added xxx