Distribution vs. Assembly

Networks distribute content.

Spaces assemble audiences.

There are hybrid forms as well:

Channels combine distribution with an assembled audience.

Tours visit a series of audience across a travel path.

I'm thinking about content, information, audiences, space, time, aggregation, and the like. This is all probably obvious, but writing it out is helpful to me.

"Tours" includes various synonyms: roadshows, travelling circus, conference series.

This musing follows on a set of earlier thoughts on the symmetry between signals and records.

Signals transmit encoded symbolic messages from a transmitter across space through a channel by variations in energy over time from a to a receiver.

Records transmit encoded symbolic messages from a writer across time through a substrate by variations in matter over space to a receiver.

Again, there are hybrid forms as well, e.g., endocrine and chemical signalling systems are based on records (the encoded chemicals) but distribute much as signals.

Show thread

@dredmorbius i guess lots of us have been thinking in the vicinity of these thoughts of yours when it comes to video conferences versus offline assemblies. all of a sudden we had a new appreciation for the combinatorial flexibility of bodies and faces assembled in a room or at a big table or in a garden. but maybe there's more ontological work to be done to make sense of those experiences at the level of abstraction you appear to be interested in.

@dredmorbius i suppose you can go on little tours at a crowded party

@poebbel Interesting.

Public speaking for a long time was limited by the carrying power of a single human voice, and was long shaped by the acoustics of spaces (outdoor arenas, indoor chambers) in which those addresses were given. If you think through the characteristics of classical oratory (rhyme, meter, often musical accompanyment), these probably served mnemonic roles (in remembering the content, for the speaker), but might also have aided the audience in hearing (something I've just now thought of ... the mnemonic role is an older realisation).

Stone cathedrals and churches, with a horn-shaped pulpit (the preacher's voice projected outwards), and a sing-song chant intonation (reverberates off the walls) likewise.

Modern audio capture, amplification, and speaker technology changed everything. German audio technology prior to and during WWII (mics, amps, speakers, magnetic audio tape) were strategic assets and game-changers. The Nueremberg Rallies and simultaneous nationwide taped addresses could not happen without those. (The CIA was actively engaged, aling with Bing Crosby, in developing tape and audio tech in the US after the war.)


@poebbel Which was a prelude to the "tours at a party" comment.

A crowded party is a high-density social gathering usually without amplification, at least for speakers (and usually cover of music or other distraction), making a set of small-n discussions (say, 2-10 people, and usually on the smaller side) possible. These are both serial (one person may speak or listen to numerous conversations) and simultaneous (there are numerous conversations happening at once).

The "cocktail party" scenario is a frequent one in various explorations (most dating to the CP's own heyday of roughly 1940 -- 1970 or so), though I haven't read anything specific to information/comms theory/studies that I recall.

Good suggestion.


@dredmorbius This is really interesting. I don't think I knew that people had studied cocktail party scenarios from an information theory perspective.

And your summary points almost directly to the issues I’ve been thinking about.

The default setup for an online video conference is presumably a single small- or large-n discussion. You can get closer to the cocktail party if you make use of the break-out room feature which various video conferencing systems provide. And depending on the software and the settings, Individual participants can go on tours, if you like. But many people I’ve talked to still find this a poor substitute for an offline meeting or cocktail party. Given appropriate software and moderation, it’s mostly fine for when people come knowing who they want to talk to about what. But at offline meetings and cocktail parties, people negotiate from a distance, often in quite subtle ways, about who is going to talk to whom. Body language, movement through the space, tone of voice, eye contact, bodily touching (surreptitious or overt, wanted or not): all of these things are used to try to establish conversations or ward off unwanted partners.

What do you mean by “CP”? Were you referring to yourself?

@dredmorbius It seems to me like a key role is often played in these negotiations by ambiguity. The one who glances at another or moves closer to them, or in a nightclub perhaps brushes past them, makes an offer of interaction that is ambiguous. The ambiguity provides something like plausible deniability. If the offer is not accepted, both parties can choose to treat it as if it was never really intended as an offer. But it’s pretty hard to log in to a break-out room ambiguously!

@poebbel "CP" == "cocktail party".

Intermediated communications (video chat, online, etc.) tend to be far more intentional --- there's a lot of set-up and establishment overhead.

At a party you can simply drift through the crowd, the process is far more serendiptious, though also constrained by the dimensions of the space, the guest list, attendees, gatekeeping, and reach of individual voices (and/or accuity of hearing).

It's not even necessarily a case of better or worse so much as different.

That said, even as someone who's not generally an extravert, I prefer in-person meeting, given a sufficiently interesting crowd.

@poebbel I've tried a few searches to see if there's any research along the lines I'd suggested (and not just the single-signal-attention problem). Nothing so far, DDG Web and Google Scholar.

@poebbel Some of my thinking on this comes from passages from Arthur C. Clarke's books. Imperial Earth (1975) in particular has a few passages describing party dynamics. They're not specifically information-theoretical descriptions, but there's a strong flavour of that going on. One gathers that there's some basis from Clarke's own lived experience.

@poebbel What I often find is:

  • There's a lot of practice that's not especially aware of theory.
  • There's almost always someone who's had thoughts similar to my own out there. A hobby has been having new-to-me "original" thoughts, then putting an effort into discovering who got there earlier. (If I'm within a few decades of first emergence, I consider myself reasonably warm-on-trail.)
  • There's a lot of writing / discussion that's very implementation-specific. I like to look beyond that.
  • Identifying the abstract notions is very useful in finding patterns, and potential new developments.

As far as I'm concerned, keep it coming. I enjoy structured thought as an end in itself. I also find that breaking things down to an abstract form helps me recognise parallels between disparate phenomena... which I guess I just enjoy as an end it itself as well!

@Austin Thanks.

One thing that fell out of my information/signal/record symmetry notion was the realisation that channels and substrates (if you can think of a better single-word term for "recording medium" as distinct from "transmission medium", please do), was that each is characterised by a predictable ground state and a generally unconstrained state space that can be mapped to it.

Erwin Schroedinger came up with the notion of "aperiodic crystals" (I encountered this in Hofstadter's Goedel, Escher, Bach). Einstein pressaged the discovery of the maser and laser working with stimulated emissions of radiation. Lasers & masers are analogues of radio-frequency transmitters. Stone, clay, papyrus, paper, punch-cards, mag-tape, spinning rust, digital media (CD/DVD/BluRay) are all state-impressionalble media. Fourier transforms apply strongly to analogue signal encoding in signals.

Mind, I kind of half-grasp most of this, though could probably work my way through it given the initiative to do so and understanding of significance.

A prediction though is that matter states which are highly regular but arbitrarily modifiable will likely prove excellent storage media. I'm looking at you, graphene.

@dredmorbius Your word 'symmetry' helped me. And 'patterns'. I've read a bunch if philosophy, and paid attention to conceptual patterns (some writers are clearly full of them) without ever explicitly considering them as a means of conceptual elaboration / invention.

@dredmorbius Do you by any chance have a name for the linguistic 'function' F()() such that F(space)(time) is a signal and F(time)(space) is a record?

@poebbel No....?

Is there such a function, or are you seeing this in what I've written?

(If the latter, then thank you, as that's ... possibly useful.)

Actually, I think what we'd be looking for might be something like:

Message: δe/δt * d

Record: δm/δd * t

(I'm pretty sure that's wrong. It's conceptually ... interesting)

That's the "change in energy with time ... multiplied by? ... space" and "change in matter over distance ... multiplied by? ... time".

I'm pretty confident of the differential bit. The operation and the third term, not so much.

There's quite a bit of physics of signals and wave propagation, and the message equations might borrow from that. Also Claude Shannon.

I'm not aware of a similar physics of records, per se, though some aspects of that, e.g., Newton's Laws of Motion, would apply.

@poebbel BTW, I'm not sure if this was answering the question you'd asked directly.

"signal function" and "record function" (or equation or dynamic or formula) might be the terms I'd use. I've not thought of that.

NB, I've tied this "distribution / assembly content / audience / information communications" thread together into a semi-sensible bloggish post here:


@dredmorbius Nice to see you have a way of giving these Masto musings a more permanent home!

Re my question, I should perhaps have written something life F(x)(y), though I'm not even sure if *that* corresponds to how mathematicians would write it. All I meant was that if you look at your definitions of signals and records and treat the concepts of space and time as variables, and abstract from them, then the difference between the two definitions disappears. Well, almost. There are a few other defining concepts (transmitter/writer, channel/substrate, energy/matter) that you’d either have to treat as variables too, or cut out. Since if you treat them as variables you don’t end up with the nice symmetry that space and time give you, and since I was interested in that symmetry (you mentioned the word), I just imagined cutting them out, producing:

Transmits encoded symbolic messages across (x) by variations over (y) to a receiver.

@dredmorbius I wanted to call this slightly more general definition a function, thinking that you could use to map different things as x and y in each case to different definienda. And in some cases, there’d be pairs of mappings/definitions that relate to each other in the way that yours did, where x in one = y in the other and vice versa, a certain kind of symmetry.

I’m not sure, but I don’t think you had just this abstraction in mind, so my question went a bit astray.

It was a bit rough and ready, in any case. But the background to it was that I was thinking about how symmetries and patterns might be ways of ordering a theoretical construction. And if you formulated a symmetry like the above, then under certain circumstances (e.g. if you were to discover that the function had applications to more cases than to signals and records) it might be useful to remember it, in which case a name would come in handy.

@poebbel In functional programming terms:

message( mass, energy, time, space, entropy, medium, encoding, ... )

That is, a message is a function of the parameters listed. The role / iteraction of the parameters would of course vary.

The key distinction between a record and a signal is in noting what changes, and what iterates.


  • A record is a change in matter/space, iterated over time.

  • A signal is a change in energy/time, iterated over space.(or distance).

(I think iteration or integration (in the calculus sense) is correct here.)

Closer to what you have in mind?

Well, I’m actually not sure if what I had in mind was so determinate, unfortunately. I was excited by the discussion and just sort leapt before I looked. I’m not sure how interesting the “line of inquiry” that I’d struck out on was (hardly a line; more of a squiggle).

But I definitely recognize what seems to be the old symmetry in your new pair here. So to that extent I follow your thought!

I notice a difference in this new formulation though. I wonder if it’s significant.

Just now you used the verb “iterate”. Earlier you had “transmit”. I assume that these two words are describing similar processes, or maybe the same process. But if so, then something in your way of speaking has changed. Because here it’s the signal that's iterated. Above the signal was the thing that transmitted, and what it trasmitted was the message.

I don’t fully know what your aims are – is this something that strikes you as in need of “clarification”, as they sometimes say?

@poebbel So, yes, I did change terminology, and I did so only after the thought that "iterating" might make more sense than "transmitting" (and be easier to express mathematically) occurred.

I'm looking for maximum clarity and expressibility in describing what's going on. There may be a summation (e.g, sigma expression) iterated over time/space. Possibly an integral (the limit of sumation).

Or an iterated / cellular computation (e.g., as in computer simulation).

I'm drawing off of ideas such as Xeno's Paradox and Newton's method of approximations. I'm not fully certain these help.

I'm also not entirely certain what my aim is, though again, coming up with a useful and symmetric description (or finding the failure points of symmetry) is much of that.

@dredmorbius Well I hope I can hear more about the work as it progresses!

Sign in to participate in the conversation

The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!