Cleric wrote: ↑Thu Jan 09, 2025 4:24 pm
So to enter in greater concreteness, can you identify something in the present structure of language that prevents
you from approaching the deeper musical reality of the Logos?
I think I already expressed that quite clearly, for example in what I proposed to call LaaS, Language as a Service,
here, which of course hints to the Ahrimanic degeneration of contemportary linguistic use. I will also try to write something new, however, the follwing that I wrote a month ago is still a good way to word my experience.
***
To continue the reflection on LLMs and the nature of language, I believe there are at least two ways in which nowadays the nature and potential of language can be missed, and essentially are missed. The question is not straightforward because of how language intersects our spectrum of activity
in between thinking and feeling.
One way to misuse language is what Cleric referred to as “verbal tokenization”, when we are “too lazy to think in pictures”. I believe this dependence on tokens extends beyond episodic laziness. It's also a generalized phenomenon. It consists of the concrete difficulty, without proper spiritual development, to come to the
reality of a collective concept (such as for instance the concept of triangle) without pictorial thinking and eventually imaginative approach. Without that, the word “triangle” remains a mere word, as Steiner says. As a consequence, the common denominator among all triangles is experienced as only a verbal symbol.
The sense is that the only collective element is the mere token. This is one major problem of our language habits. And, even when pictorial thinking and imagination are worked with, this imprisonment in the words-in-themselves tends to persist. It’s a pervasive linguistic modality of today, that
does not just dissipate the moment the power of images is discovered. We still tend to externalize language to the space of words-in-themselves, where we use it as mere encoding of abstract,
unpictured, definitional, operable concepts.
In this modality, language is the arranger of an intellectual life, spent in an internally coherent, but rather isolated layer of experience, with little etheric resonance. Perhaps we could call this modality Language as a Service - LaaS. The idea is that language becomes
hosted in a functional, readily accessible space, but that space is not grounded in owned imagination. It resides with an external provider. It’s almost like we pay a periodic subscription to it, in order to be granted access to its mechanistic-linguistic logic, as a modular service, for the purposes of ordinary communication, and intellectual viability. Our ordinary self scampers in that space, using the ready-made, proven functionalities - the verbal links - rather than making them
anew every time, out of active soul substance. We take home the practicality on the one hand, but also pay for it, in
diminished soul substance, so that the direct connection of living concept to word is essentially lost (or never found). Maybe the metaphor of building a website with Wix or Squarespace can work. It’s convenient, it gets results, but the more I build, the more I complexify the linguistic construction within the pre-made architecture, the more I am locked in, and the smaller the chance to know what background understanding and what work it would take to steer the front-end in full autonomy of expression, away from dependence.
The other problem we face, strictly connected to the first one, when we are used to the practicality of Language as a Service - when we trade our labor-intensive first-person activity against easy, pre-assembled verbal modules - is the
disconnection with feeling. This is probably just the other side of the same disconnection. When we are not aware, or not attentive to, the responsibility to use language as a felt expression of the soul, the connection word-sound-feeling that makes the spoken word a
testimony of the individual human heart (also as part of a human group) remains unrealized. Even when it is realized, it's still very easy to lose it. This has little to do with vague feelings and etched soul pathways. It’s rather the lack of experiential knowledge of the creative connection between the heart and the word - through
sound. As a modern habit, it boils down to treating words as mere
verbal symbols. This habit misses the conscious experience of the powerful divine prerogative (made available to humans as language) to re-create the world (the concepts, the ideas,...)
out of oneself.
That’s what language can achieve for man, simultaneously to its quality of facilitating living thinking. In and with its quality of facilitating human thinking, language puts this higher prerogative into our individual and collective hands, so that we
put our unique stamp on reality, or re-form it, re-create it, in words, out of ourselves, out of our unique self, unique in its individuality of soul and also in its belonging to groups. This inevitably relies on the experience of sound in language (sound and all its inherent qualities). If this essential, feeling-nature of language is ignored, I believe language becomes like a sort of
leakage for the soul. Spiritual efforts get constantly diluted and blurred in the captivity of LaaS.
In this context, I believe that LLMs actually
are the true-size, mirror manifestation of this epochal tendency to use words-in-themselves, words that we have renounced to enliven with home-made activity, vertically connected, and imbued with both living thinking and living feeling. LLMs constitute the accomplished cultural output of our modern language use as mere, functional encoding. Though they may feel less immersive than VR, I believe LLMs are actually more immersive than VR, if we don’t restrain considerations to the purely sensory. VR, as we know it today, directly addresses the sensory spectrum of our perceptual flow, while thoughts and emotions are affected as a consequence.
However, as subscribers and habitual users of Language as a Service, all our mental pictures,
our entire perceptual spectrum is taken care of directly, in verbalized repurposing. And I think that LLMs can be seen as the neat rendering of this mode of being, to ourselves, in the technological space. It’s like a VR-at-large. VRAL?

(it's 'fun', in Swedish, vrål means something like "horrible scream"). Anyway, it goes one notch deeper into the capture of our flow of becoming. Is the next notch going to be what Levin is working at, targeting the will? I guess so.