At the edge of thought, with an emerging complex of ideas, in the moment, there is fragility, diaphanousness, withdrawal, and vulnerability to overshadowing. Starting to speak out loud will, in the best case, set one rolling like a snowball down a snowy hill. But in the worse cases, it creates a clanging sound, loudens the part that got spoken, and the rest is dimmed or pushed away. This also happens when typing text.

What if one could speak from the gestalt? This could be something like a word-cloud, or like the progression from title to table of contents to extended annotated table of contents to summary to full work. But the form isn't the point, the point is to prioritize holding the whole thing. It's "speaking the book from the side, rather than front to back", hence "sidespeak".

Instead of speaking linearly, starting at the beginning and progressing through an explanation of the idea in some order (e.g. discovery order, or the order in which a listener could follow along stepwise), we follow a morphogenetic plan.

Think of the painter. Ze doesn't start with, say, the top corner, fill in all the detail bit by bit, and then go to the adjacent patch and complete that, and so forth. Rather, ze might first set a background color field; then block out areas, such as the sky and the ground, or the gross topography of a face; then shade some main features, such as mountains and a river and trees, or eyes and mouth and hair; then come fine outlines, concrete items, and finally full detail, first in the important features and then in the periphery.

In a sense, this is highly repetitive. The artist goes over some areas, some lines, many times, repeating and emphasizing and elaborating and tweaking. Also, in a sense it is highly disorganized--the focus of activity jumps around between different areas, colors, levels of detail, and items.

Or, think of the developing conceptus. At the very first, there is raw growth of DNA, in division of fully undifferentiated cells. Soon there is the main axis, and a differentiation between what will become the fetus and what will become the placenta and amniotic sac. From there, tissues grow, differentiate, interconnect, migrate, and further differentiate, in an self-coordinating self-elaborating multi-level spatial pattern.

Could we speak this way? We're strongly conditioned to track whether the listener would follow what we're saying. Which is well. But doing so requires grammar, a progressive setting up of context in order of other-understandability, transparent vocabulary, a progression of motivated transitions and backreferences, and signposting of the structure of the argument or explanation. To speak this way for more than a few monents is simply embarrassing, or disturbing (as though you've gone crazy), or an imposition (like telling someone a long dream, pregnant with hidden meaning for you, but trite contextless noise to the listener).

But hey, now you could just say "I'm going to sidespeak".

I don't know how to do it, but I imagine it might involve: saying single words; jumping around to different branches of the thought; elliptic phrases; allusions and unexplained analogies; sentence fragments; statements of emphasis and deemphasis; bits of context; redescriptions, recitations; abduction, above induction and deduction; a process of listening for aspects of the idea-complex that are being supressed and wish to reassert themselves, and then giving them some voice; pataphorical movements. More important is the meditative-like skill involved.

Ideally it would also involve rapid lexicogenesis, but unfortunately we don't widespreadly have the sort of high-speed, high-richness skill at extemporaneous lexicogenesis that we might have been able to have if we'd been building a shared craft of lexicogenesis.

Related: Focusing, probably "Thinking At the Edge"