Communication style

Communication style is the most heavily commercialised phrase in workplace psychology, and the part of it that survives scrutiny is smaller than the industry around it — but it is not nothing, and the useful move is to say exactly where the line falls rather than to defend or dismiss the whole category.

Here is the line. Specific communication behaviours have been measured carefully for fifty years: how much speaking frightens you, how readily you argue, how you handle a disagreement, how direct you are. Those instruments exist, they have normal psychometric credentials, and they predict things.

What does not hold up as well is the step from behaviours to a type — one label, four boxes, worn into every room you enter. And the single finding that most undermines that step is not an obscure one. It is that the same person communicates differently depending on who is across the table, by a margin large enough that the field's best conflict instrument was built with three separate versions to account for it.

Almost every communication-style course opens with the claim that words are about seven per cent of communication and tone and body language are the rest. It is not true, and knowing where it came from is a decent inoculation against the rest of the genre.

Albert Mehrabian ran two small studies in 1967. One had listeners judge the feeling behind a single spoken word — a word like maybe, recorded in different tones. The other pitted a recorded word against a photograph of a face. The percentages come from combining those two results, and Mehrabian himself said for decades afterwards that they applied only to situations where someone is communicating about feelings and the channels contradict each other.

That is a genuine and interesting finding: when the words and the face disagree, people believe the face. It is not a statement about how much of any conversation is nonverbal, and it cannot be, because the studies contained no conversations. A field that has repeated it for half a century without checking is a field to read slowly.

Communication research has real instruments. They are narrower than a style and better behaved.

Communication apprehension. James McCroskey's work through the 1970s produced the PRCA, probably the most used self-report measure in the discipline, and it holds up: apprehension about speaking is stable, it varies by context in predictable ways, and it predicts occupational choice and participation. Much of what gets called a quiet style is this, and it is more tractable than a style because it responds to the standard treatments for anxiety.

Argumentativeness and verbal aggressiveness. Infante and Rancer separated these in the early 1980s, and the separation is the single most useful idea in the field. Argumentativeness is willingness to take on a position; verbal aggressiveness is attacking the person — their character, competence, background. They are independently measurable, and the same person can be high in one and low in the other. That is why "we argue a lot" says almost nothing on its own, and why the pair that never argues can be in far more trouble than the pair that does. Conflict and repair works through what the difference does over time.

Conflict handling. Rahim's 1983 inventory sorted responses to conflict along two dimensions — concern for self, concern for others — producing five approaches: integrating, obliging, dominating, avoiding and compromising. It has held up across decades of organisational research.

Directness, and the trap in it. Edward Hall's high-context and low-context distinction is the standard vocabulary for how much meaning sits in the words versus the surrounding situation. It is worth knowing and it is worth handling carefully, because Hall's claim was about cultures, and applying a group-level average to the individual in front of you is a textbook ecological fallacy. Gudykunst and colleagues addressed this in 1996 by measuring the individual-level variables directly and found that people's self-construals — how independent or interdependent they take themselves to be — did the predictive work, with nationality accounting for much less than the cultural framing implies. The person, not the passport.

Now the finding this entry is built on, and it is hiding in plain sight in the design of the instruments themselves.

Rahim's conflict inventory does not come as one questionnaire. It comes in three forms — one for how you handle conflict with your boss, one for subordinates, one for peers — because the scores differ, systematically and substantially, across the three. The person who obliges upward dominates downward often enough that measuring them once would be measuring nothing.

Follow that through and the four-box picture starts to look like the wrong unit of analysis. Your directness moves with how safe the room is. Your willingness to argue moves with how much you expect to be punished for it. Your apprehension moves with the size of the audience and whether you are being evaluated. None of this means there is no person underneath — apprehension and verbal aggressiveness are reasonably stable traits. It means the stable core is narrow and the situational range around it is wide, and a report that gives you one label for all of it has quietly assigned the whole range to you.

This is why the most common experience of a communication-style result — that is exactly me at work and not at all me at home — is not a failure of the test. It is the test measuring one referent and being read as though it measured all of them.

The four-quadrant style models — Merrill and Reid's social styles from the 1960s and 1970s, and the Marston-descended family that includes DISC — deserve a specific verdict rather than a general one, and it has three parts.

The two axes are real. Assertiveness, whether you tend to tell or ask, is a well-established dimension; it is literally a facet of extraversion in the Big Five. Expressiveness, whether feeling shows or stays out of it, is also a genuine dimension of behaviour. Jones and Hartley compared a four-quadrant instrument against the five-factor model in 2013 and found the expected correspondences, which is the reassuring result: the quadrant instruments are measuring things, and the things are ones the Big Five already had names for.

The quadrant is a cutoff, not a discovery. Both axes are continuously distributed with most people bunched near the middle. Dividing at the midpoint produces four named categories out of a cloud with no gaps in it, which is how a person at the 52nd percentile and a person at the 48th end up in different boxes and different paragraphs. The cutoff problem is the general case. It also makes the descriptions unusually easy to agree with, since a description built to fit a quarter of the population reads as insight for most of it — the Barnum effect doing its usual work.

The headline claim is untested. These models say friction between people is usually a style mismatch, and that adjusting to the other person's style resolves it. That is a matching claim, and matching claims need a specific design to test — pair people across styles and within them, vary whether each adjusts, and show the interaction. This is the design that the learning-styles literature never produced, with the result described in the entry on work style. For communication styles it has not been produced either. The claim is plausible. It is not established, and it is sold as established.

None of that makes the vocabulary useless. A shared language for "you close decisions fast and I need to think out loud" is genuinely valuable in a team, and it is valuable for a reason that does not depend on the model being true: it gives two people a way to name a friction as form rather than substance, which lowers the temperature enough for the actual disagreement to be discussed. That is a real effect of having words, and it is most of what these frameworks deliver in practice.

So use it as a vocabulary and not as an identity. Three practical readings.

Read the percentiles, not the box. If you take the Communication Style test, the two axis scores are the measurement and the named style is a convenience laid over them — and how far from the middle you sit tells you far more about how much friction to expect than which of the four you landed in.

Attach the result to a relationship. Take it twice with two different people in mind, your manager and your partner, and the difference between the two is more informative than either result alone, for exactly the reason Rahim printed three forms.

Treat the dial you dislike as a behaviour, not a trait. The person who never says the hard thing is usually not low in some fixed quality of directness; they are managing an expected cost, and the useful work is on the cost — which is boundaries territory, and is specific to particular people rather than general to you.