Rants into empty space.

A footnote on why “user created” taxonomies doesn’t sound at all boring for suttacentral

Bhante Sujato at https://discourse.suttacentral.net/t/a-suttacentral-creativity-multiplier/33579

I got the idea indirectly from a talk at the Digital Buddhism conference by Christian Wittern. Just an amazing man! He’s built this system that allows user input for a taxonomy for texts. That’s great, but (a) it’s hard and (b) its, well, a taxonomy :yawning_face:

But the thing that he said that I cottoned on to was the idea that it was not about creating a systematic and perfect taxonomical system. It was about giving people a chance to work on something they are interested in.

So that really made me see it in a different light.



I really like this line of thought – the idea of some user-created taxonomies maybe sounds boring on the surface – but there is understanding and meaning in how others structure things – if we are ok with incompleteness that incompleteness and inability to “ef” the “ineffable” (to borrow a phrase from douglas adams) is part of what maybe creates meaning and understanding – distinguishes us from the sort of incompleteness of an “AI hallucination” and I don’t see it as something to iron out “These ambiguities, redundancies and deficiencies” when they occur as a result of human creative classification

The concrete example I think of in how I would use such a user-created taxonomy tool is “a taxonomy of ‘delight’ in the suttas’.” I thought it would be cool one day to find all the contexts “delight” or nandi or other words translated as “delight” is mentioned in the suttas, and categorize those contexts into some taxonomy so later someone else could see connections and meaning – maybe if we can embrace the idea of incompleteness of a taxonomy it’s meaningful – we don’t impose the demand on the author for its completeness – maybe it takes a kind of intellectual maturity to see the benefit of intentionally creating incomplete classification scheme – just like John Louis Borges essay above – probably sounds like madness to one who is still wedded to some idea all classification schemes must be complete and any that isn’t is “irrational”. (another quote from the essay)

“He knows that there are in the soul tints more bewildering, more numberless, and more nameless than the colours of an autumn forest… Yet he seriously believes that these things can every one of them, in all their tones and semitones, in all their blends and unions, be accurately represented by an arbitrary system of grunts and squeals. He believes that an ordinary civilized stockbroker can really produce out of his own inside noises that denote all the mysteries of memory and all the agonies of desire (G.F. Watts, p. 88, 1904).”

The idea of dropping the pretence a taxonomy needs to be complete and instead embraces incompleteness may have deeper significance than just giving people a chance to work on something they are interested in.

does that make sense?

One response to “A footnote on why “user created” taxonomies doesn’t sound at all boring for suttacentral”

  1. jones Avatar
    jones

    Also in terms of it being “hard” well – I don’t know anything about how suttacentral is structured IT wise – but having said that – just imagining all the suttas as text files in folders or in one massive json file or other datastructure – is the hardness in how conceiving it? What is a taxonomy anyway? a taxonomy is a structure of how we categorize “things” – in terms of an information system – we could imagine this as a list of things and a list of those lists of things – basically a tree structure. Those “things” could be symbols – atomic. something that allows for lists and lists of lists and the “processing” of those lists of symbols, – a list processing language if you will – “LISP” – is a language made just for this sort of open ended exploratory type programming. And you need not define symbols to define the structure – so lisp has always alowed this sort of exploration – whereas other languages require you to do the components and low level and have that well defined and build up, lisp allows you to create a program “top down” in a sort of “wishful thinking” (the analogy is casting a spell in one popular text book SICP) where you sort of design programs by imagining how they would run – you in other words have symbol stand in for something and not worry too much yet about what it is or refers to – in other languages this would result in dumping core or not running or something – in lisp its totally normal to have some symbol not yet defined and be manipulating and doing operations on that. So it leaves itself open to defining “things” – of course in the context of suttas this might be a “word” a “phrase” a “sutta” a “parallel” a “reference” – we may even come up with new things “a rhyming couplet” – we can still have a taxonomy and it may not yet be specific – ie. on all rhymes – that is where the “helper functions” may come in – someone writes a function that looks at the ending of all lines and searches for simularity – maybe checking levenstein distance or the like.

    I imagine some system that enabled user created taxonomies would just be writing some functions to operate on suttacentrals existing datastructures to have sort of some primitive atoms or things and allow users to make their own “things” by writing helper functions on these datastructures, then grouping these functions in lists and lists of lists and so on is the taxonomy.

    in terms of a user interface the “lists of lists” readily represents as a tree with some nodes and lines between them – a bit of .js that takes s-expressions in to represent them as a tree in the browser for instance

    Reply

Leave a comment