What Is Communication? A Clear Framework for How Ideas Transfer

In the spring of 1847, the physician Ignaz Semmelweis discovered that deaths from childbed fever in Vienna's First Obstetrical Clinic were five times higher than in the Second Clinic, and the only difference he could identify was that the First Clinic was staffed by doctors and medical students who moved directly from autopsies to deliveries, while the Second was staffed by midwives who did not. After a colleague died of an identical infection following a scalpel wound from a student during an autopsy, Semmelweis concluded that "cadaverous particles" transmitted on the hands of physicians were killing the mothers. He introduced mandatory handwashing with chlorinated lime solution and mortality plummeted.

He then tried to communicate this finding to the medical establishment. He failed. He wrote letters, gave talks, published a book in 1861. The medical community largely rejected his finding. Semmelweis died in 1865, institutionalized, his discovery still not widely accepted. It was not until Pasteur's germ theory provided the mechanism for why handwashing worked that the discovery was integrated into medical practice.

The Semmelweis case illustrates what a theory of communication must account for: the gap between what a sender knows and intends to transmit, and what receivers actually receive and accept, is often vast. And the size of that gap is not primarily determined by the accuracy of the sender's information or the intelligence of the receivers. It is determined by the structural features of the communication process itself: the encoding choices made by the sender, the interpretive frameworks available to receivers, the channel through which the message travels, and the feedback mechanisms (or their absence) that allow sender and receiver to discover and correct misalignments.

A useful framework for understanding communication must account for all of these factors. What follows is a six-stage model that integrates the classical insights of Shannon, Weaver, and Schramm with contemporary cognitive science on meaning construction, mental models, and communication failure.

Stage 1: The Idea (Internal Representation)

Communication begins with something the sender wants to transmit: a concept, a fact, an instruction, an emotional state, a perspective, a plan. This content exists initially as an internal representation -- a structure in the sender's mind that includes the information itself, the relationships between its elements, the context in which it should be interpreted, and the conclusions the sender draws from it.

The sender's internal representation is almost always richer and more contextually situated than anything that can be transmitted. A speaker who has spent years thinking about a topic has an internal representation that includes the mistakes they considered and rejected, the counterarguments they found unpersuasive, the evidence they weighed and found compelling, and the qualifications they believe are important. That entire background is part of the sender's representation of the idea -- but most of it is invisible to receivers, who receive only what the sender chooses to transmit.

This creates the first source of communication failure: selection error. The sender must choose what to transmit from their full representation, and they systematically choose based on what seems important to them, given their knowledge. What is most important to transmit for the receiver may be different -- and the sender, cursed by their own knowledge, often cannot reliably identify the gap.

Stage 2: Encoding (Translating Thought to Signal)

Encoding is the process of translating an internal representation into a transmittable form: words, gestures, images, numbers, facial expressions, written symbols. The encoder must select among the representational resources available (language, visual elements, formal notation, example-based illustration) and make choices about how to use them.

Every encoding choice involves trade-offs:

  • Precision vs. accessibility: Technical language can encode precise distinctions that ordinary language cannot, but at the cost of requiring that the receiver share the technical vocabulary
  • Completeness vs. processability: More complete encoding captures more of the sender's representation, but dense encoding exceeds receiver processing capacity
  • Explicitness vs. implicitness: Explicit encoding states everything; implicit encoding relies on shared context to fill gaps (efficient when context is shared; catastrophically ambiguous when it is not)

Encoding failure occurs when the sender's choices produce a signal that cannot accurately reconstruct the intended representation on the receiver's end. This includes:

  • Jargon without translation: Using technical terms that the intended receiver does not possess
  • Assumed context: Relying on context that the receiver does not have
  • Inappropriate abstraction level: Encoding at a level of abstraction different from what the receiver needs
  • Structure mismatch: Organizing the encoding in a way that does not help the receiver reconstruct the sender's organization

*Example*: In 1999, NASA lost the Mars Climate Orbiter -- a $327 million spacecraft -- because one team encoded trajectory data in imperial units (pound-force seconds) and another team's software decoded it assuming metric units (newton-seconds). The encoding was accurate; the signal was transmitted without error. The failure was an encoding/decoding mismatch: the same signal carried different meaning for sender and receiver. The result was a trajectory error that destroyed the spacecraft.

Stage 3: Transmission (Moving the Signal Through a Channel)

The encoded signal travels through a channel from sender to receiver. The channel determines:

  • Bandwidth: How much information can be transmitted simultaneously (face-to-face conversation carries words, tone, facial expression, body language simultaneously; email carries text only)
  • Fidelity: How accurately the signal is preserved during transmission
  • Delay: The time between encoding and receipt (affecting the possibility of real-time feedback)
  • Noise characteristics: What kinds of interference distort the signal in this particular channel

Shannon and Weaver's original model identified this transmission stage as the primary location of communication problems. In electrical systems, this is largely correct: the primary failure is signal degradation during transmission. In human communication, transmission failure is real but less frequently the primary problem than encoding/decoding failures. Most organizational and interpersonal communication failures are not failures of transmission; they are failures of encoding, decoding, or feedback.

The choice of channel is itself an encoding decision with significant consequences. An instruction that should be unambiguous ("turn off the oven at 6pm") requires a different channel than a nuanced conversation about organizational strategy. Matching channel to content -- choosing channels with sufficient bandwidth and appropriate noise characteristics for the content being transmitted -- is an often-underappreciated communication skill.

*Example*: Research by Albert Mehrabian (1967, often misquoted but real in its core finding) found that the emotional component of communication between people is heavily carried through nonverbal channels -- tone, facial expression, posture -- that are absent in written text. This has practical implications for channel selection: communicating emotionally significant content through text removes the channels through which emotional meaning is primarily transmitted, creating high potential for misinterpretation. The choice to deliver a performance review in writing versus in person is a channel choice with substantial implications for what actually gets transmitted.

Stage 4: Decoding (Reconstructing Meaning from Signal)

Decoding is not the reverse of encoding. Encoding goes from internal representation to signal; decoding goes from signal to internal representation. But the receiver's internal representation is built from different materials than the sender's: it is built from the signal received, the receiver's existing knowledge and interpretive frameworks, the receiver's contextual assumptions, and the receiver's expectations about what senders like this sender typically mean by signals like this signal.

Meaning is constructed, not extracted. The receiver does not simply unpack a pre-existing meaning from the signal; they construct a meaning that they infer is consistent with the signal. This is why the same signal can produce different meanings for different receivers -- they are constructing from different knowledge bases and interpretive frameworks.

The reconstruction process is largely automatic and unconscious. Receivers typically experience comprehension as immediate and transparent -- they feel they understand what was said, not that they constructed a plausible interpretation. The automaticity of decoding conceals its inferential character and prevents receivers from noticing when their inference diverges from the sender's intended meaning.

Decoding failure occurs when the receiver's reconstruction does not correspond to the sender's intended representation. This happens when:

  • Vocabulary mismatch: The receiver does not know the meaning the sender attached to a term
  • Contextual mismatch: The receiver fills implicit gaps with incorrect contextual assumptions
  • Schema mismatch: The receiver interprets the signal through a different organizational framework than the sender intended
  • Inference error: The receiver draws incorrect inferences from explicit content using their background knowledge

*Example*: In organizational communication, the instruction "make this a priority" is decoded differently by different receivers because "priority" is contextually defined differently in different organizational cultures. For one receiver, "priority" means "work on this today, deprioritizing other tasks if necessary." For another, it means "keep it on your task list and address it within the week." The sender may not know that these decoding differences exist, and receivers may not know they are decoding differently from each other or from the sender's intent. The misalignment becomes visible only when behaviors diverge from expectations.

Stage 5: Reconstructed Meaning (What Was Actually Received)

At the end of the decoding stage, the receiver holds a reconstructed meaning: their best inference about what the sender intended to transmit. This reconstructed meaning is what actually influences the receiver's beliefs, decisions, and actions -- not the sender's intended meaning, and not the signal. The signal is only an intermediate.

The most important characteristic of reconstructed meaning is that the receiver typically cannot distinguish it from the sender's intended meaning. The receiver who has decoded a message holds a belief about what the message meant; they experience that belief as understanding of the sender's intention, not as a reconstruction that might diverge from it. Revealing the gap between reconstructed meaning and intended meaning requires external intervention -- feedback, testing, or explicit comparison.

This stage also reveals why "I said it clearly" is not a defense against miscommunication: the sender's assessment of encoding quality is irrelevant to the receiver's actual reconstructed meaning. What matters is what the receiver received, which the sender typically does not know without feedback.

Stage 6: Feedback (Completing the Loop)

Feedback is information that flows from receiver back to sender about how the message was received, understood, and interpreted. Feedback is what transforms communication from a one-way transmission into a two-way process capable of self-correction.

The feedback loops in communication that actually improve understanding require two components: (1) the receiver must send information back to the sender about their reconstructed meaning, and (2) the sender must use that information to adjust subsequent transmission. Either component alone is insufficient. Receivers who send feedback that senders ignore have no functional feedback loop. Senders who want feedback but never receive it because organizational norms suppress it also have no functional feedback loop.

Feedback operates at two levels:

Signal-level feedback: Confirmation that the message was received ("I got your email," "I heard what you said"). Signal-level feedback tells the sender that transmission occurred but provides no information about whether decoding was accurate.

Meaning-level feedback: Information about how the receiver decoded the message ("I understood you to mean X" or "I'm not sure what you meant by Y"). Meaning-level feedback enables the sender to identify decoding errors and correct them, which is the actual communication-improvement function of feedback.

Most organizational feedback is signal-level only. Acknowledgment of receipt is routine; demonstration of accurate decoding is rare. This is one reason that communication breakdowns are so common in organizations even when everyone believes they are communicating clearly: all the signal acknowledgments were there, but the meaning feedback that would have revealed decoding errors was absent.

The Full Model: What Makes Communication Succeed

Integrated across all six stages, communication succeeds when:

  1. The sender's encoding accurately represents the intended idea in transmittable form
  2. The encoding is matched to the receiver's vocabulary, contextual knowledge, and interpretive frameworks
  3. The channel carries the encoded signal with sufficient fidelity and bandwidth
  4. The receiver's decoding produces a reconstructed meaning that corresponds to the sender's intended meaning
  5. Feedback reveals any misalignment between intended and reconstructed meaning
  6. The sender adjusts subsequent encoding based on that feedback

Failure at any stage produces communication failure, and the stages are not independent: encoding choices constrain what decoding can recover; channel choice constrains what encoding can be used; feedback quality constrains how quickly and accurately encoding adjustments can be made.

The model reveals why the intuition "I said it, so they should understand it" is wrong: transmission is not understanding. The sender's encoding is the beginning of a process, not its end. What matters is what arrives in the receiver's reconstruction -- and that is determined by factors only partly within the sender's control.

Why This Model Is Useful

The six-stage model is practically useful because it localizes failure. When communication does not produce the desired result, the model suggests specific diagnostic questions:

  • Was the idea clearly formed in the sender's mind? (Stage 1 -- can the sender accurately describe their own representation?)
  • Did the encoding choices match the receiver's resources? (Stage 2 -- did the receiver have the vocabulary, context, and background to decode correctly?)
  • Did the channel carry the signal accurately? (Stage 3 -- was there information lost in transmission?)
  • Did the receiver decode accurately? (Stage 4 -- what did the receiver actually understand?)
  • What was the reconstructed meaning? (Stage 5 -- what belief did the receiver end up with?)
  • Was there feedback? (Stage 6 -- did the sender have any information about the accuracy of reception?)

Miscommunication that is attributed to receiver failure ("they didn't understand") can often be more accurately attributed to encoding failure (the encoding did not support accurate decoding) or feedback failure (nobody knew a mismatch had occurred until it was too late to correct).

This reframing has significant implications for how organizations approach communication improvement: rather than training receivers to comprehend better (which addresses the symptom), the model suggests investing in encoding quality and feedback mechanisms (which addresses the causes).

Implications for Semmelweis

Returning to the opening example: Semmelweis's communication failure was not a failure of transmission. He sent the signals; they were received. It was primarily a decoding failure created by an encoding-context mismatch. His receivers -- contemporary physicians -- decoded his claim through a framework (miasma theory, the dominant medical paradigm) that made "handwashing prevents death" incoherent. His encoding assumed a germ-theory framework that his receivers did not have.

The tragedy is that Semmelweis's evidence was compelling; the mortality data was not in question. But compelling evidence transmitted through a framework the receiver cannot access produces reconstructed meaning that does not correspond to the intended meaning. Pasteur later provided the missing framework -- germ theory -- that enabled accurate decoding of the same evidence Semmelweis had presented years earlier.

This suggests the communication lesson that extends far beyond medicine: accurate evidence, clearly stated, is often not enough. The receiver's decoding framework must also be in place. Providing evidence without establishing the interpretive framework the evidence requires produces miscommunication regardless of the evidence quality. The framework is not noise -- it is the prerequisite for correct decoding of the signal.

References

Tags

communication information transfer clear communication mental models understanding

Frequently Asked Questions

What is communication in simple terms?

Communication is the process of transferring meaning from one mind to another through signals such as language, symbols, or behavior.

Why does communication often fail?

Communication fails when meaning is distorted by assumptions, missing context, noise, or differences in interpretation.

Is communication just about speaking clearly?

No. Clear communication depends on shared understanding, context, and how information is interpreted, not just wording.

What role does the receiver play in communication?

The receiver actively interprets the message, which means communication depends as much on interpretation as on transmission.

Why is communication considered a system?

Because it involves inputs, outputs, feedback, and adaptation rather than one-way information delivery.

Share this article