PrefaceGrammar as Cognitive Partition

Chapter 1 · Section 0

The Transmission Myth


The Transmission Myth

The dominant metaphor for language in Western philosophy is transmission: a speaker encodes a thought into words; the words travel across a medium; a listener decodes the words back into the same thought.

This metaphor is wrong. Not merely incomplete — structurally wrong in a way that generates endless philosophical confusion.

Why Transmission Fails

Consider the simplest possible utterance: "It's cold."

Under the transmission model, the sentence carries a proposition PP from speaker AA to listener BB. If transmission were perfect, BB receives exactly PP.

But which PP? The sentence is compatible with at least:

  • P1P_1: the ambient temperature is below some threshold
  • P2P_2: the speaker is uncomfortable
  • P3P_3: a request to close the window
  • P4P_4: an invitation to sympathy

The disambiguation doesn't come from the sentence. It comes from the shared context: the room, the relationship, the prior conversation, the facial expression. Wittgenstein noticed this. Grice formalized it. And yet the transmission metaphor persists.

The Synchronization Alternative

A better metaphor is synchronization. Language is not a truck carrying cargo; it is a tango.

In a tango, neither partner transmits choreography to the other. Instead, through a sequence of micro-signals, they co-construct a shared movement. The "information" is not in either body separately — it emerges in the coupling.

Formally: let SA(t)S_A(t) and SB(t)S_B(t) be the mental states of AA and BB at time tt. Successful communication is not SB(t1)=SA(t0)S_B(t_1) = S_A(t_0), but rather:

SA(t1)SB(t1)functional<ε\|S_A(t_1) - S_B(t_1)\|_{\text{functional}} < \varepsilon

for some functionally relevant norm. The goal is convergence, not copying.

This reframing has immediate consequences. It explains why:

  • Perfect translation is impossible (different prior states SB(t0)SA(t0)S_B(t_0) \neq S_A(t_0))
  • Metaphor is not ornament but load-bearing structure
  • Context is not background noise but the primary channel
Forward pass through a 3–4–2 neural network
Forward pass through a 3–4–2 neural network click to expand
Transmission model vs. synchronization model of communication
Transmission model vs. synchronization model of communication click to expand