Habermas's Communicative Action: The Only Serious Theory of Why Feedback Cultures Actually Work

60-Second Summary

Habermas (Theory of Communicative Action, 1981): human speech aims at either strategic effect (get you to do something) or communicative understanding (reach a shared, reasoned agreement). The two require different conditions and behave very differently.
Every serious speech act raises four validity claims a listener can challenge: truth (the facts), rightness (norms/appropriateness), sincerity (the speaker means it), comprehensibility (it can be understood).
Genuine communicative action needs an 'ideal speech situation': symmetrical access to speak, freedom from coercion, all four validity claims open to challenge. Most workplace 'feedback' violates at least two.
Feedback fails when strategic action wears the clothes of communicative action — 'this is developmental' while a rating is being decided, 'we want honest input' inside a power gradient with no protection.
Design implication: separate the forums where strategic action is honest from the ones where communicative action is genuinely possible, protect symmetry, and let all four validity claims be challenged.

A company launches a 360-review programme with the tagline 'radical candour, real growth'. Six months later, HR analyses the free-text responses and finds that 80% of comments are either bland-positive or coded-negative (soft phrases everyone in the org knows how to translate). Written directly, the same feedback would end careers. The programme did not fail because people are cowardly. It failed because it asked communicative action out of a setting that was actually strategic. Habermas predicted this in 1981.

What Habermas actually argued

“In communicative action, participants are not primarily oriented to their own success; they pursue their individual goals under the condition that they can harmonize their plans of action on the basis of common situation definitions.”
— Jürgen Habermas, Theory of Communicative Action, Vol. 1 (1981)

Habermas, working out of the Frankfurt School tradition, made a distinction that most workplace communication training ignores. Speech aims at one of two things:

Two orientations of speech

1
Strategic action
Speech oriented to success. The speaker's goal is to influence the listener toward a desired outcome. Negotiation, persuasion, spin, sales, most performance reviews with comp attached. Not illegitimate — but different rules apply.
2
Communicative action
Speech oriented to mutual understanding. The speakers accept the possibility of being wrong and coordinate action through the force of the better argument. Requires very specific conditions to be possible at all.

Every serious speech act, Habermas argues, raises four validity claims that a listener can rationally challenge:

The four validity claims

1
Truth
Is what you say factually correct? Challenge: 'that isn't actually what happened'.
2
Rightness (normative)
Is what you say appropriate given our shared norms and context? Challenge: 'you shouldn't have raised that this way / in this forum'.
3
Sincerity
Do you actually mean what you say? Challenge: 'I don't believe you're saying this in good faith'.
4
Comprehensibility
Can what you say be understood? Challenge: 'I don't know what you actually mean by that'.

For communicative action to be genuinely possible, Habermas describes an 'ideal speech situation': every participant has equal opportunity to speak, initiate, question, and challenge; freedom from coercion; and all four validity claims are open to challenge without penalty. He knew this is an ideal — an approximation to work toward, not a state a real organisation achieves. But the closer the setting is to the ideal, the more communicative action is possible; the further from it, the more everything defaults to strategic action while pretending otherwise.

How workplace feedback violates the conditions

The claim vs the reality in most feedback forums

The claim ('communicative action')

'Radical candour — say the hard thing'
'This is developmental, not evaluative'
'All voices matter equally in this room'
'We want your honest input'
'No one will be penalised for what they say here'

The reality (strategic action underneath)

Speaker's comp, promotion, or standing depends on the listener's next decision
Same conversation is upstream of a rating being decided next week
Power gradient is steep and unnamed; junior voices carry a fraction of the weight
'Honest input' is remembered and correlated with future decisions
No structural protection exists — the 'no penalty' promise cannot be enforced

Six common feedback failures, in Habermasian terms

1
Sincerity claim quietly untrue
'This is developmental' said in a forum where a rating decision is imminent. Listeners correctly detect the strategic frame and respond strategically.
2
Rightness claim contested but not surfaced
Speaker breaks a norm (raises a personal issue in a group forum, escalates over the manager's head) — the group knows, the speaker doesn't. Feedback dies because the norm violation is never addressed.
3
Truth claim untested
Feedback is delivered as fact ('you're not strategic enough') without evidence, and challenge to the truth claim is treated as defensiveness. Communicative action requires that truth claims be challengeable.
4
Comprehensibility failure treated as capability failure
The listener says 'I don't understand what specifically you mean'; the speaker hears 'they don't get it'. Both leave. The instrument reads 'not coachable'.
5
Coercion masked as invitation
'You can decline the 360, of course' — inside a culture where declining is career-limiting. The invitation is nominally free and structurally coerced.
6
Asymmetric access to speak
Skip-levels, town halls, and All Hands where senior leaders speak for 45 minutes and take 3 questions from a queue. Formally participatory, structurally strategic.

Designing feedback that can actually work

Seven design moves that create real Habermasian space

1
Separate strategic forums from communicative forums — and name them
The performance conversation attached to a rating is strategic; that's fine, don't pretend otherwise. The quarterly growth conversation with no rating attached can be communicative if designed well. Different time, different frame, different rules.
2
Make all four validity claims explicitly challengeable
Build the norm: any feedback receiver can ask 'what specifically did you observe' (truth), 'is this the right forum for this' (rightness), 'is there something else going on here' (sincerity), 'help me understand what you mean by X' (comprehensibility) — without being labelled defensive.
3
Protect symmetry of access
In any forum claiming communicative action, junior participants speak first, senior participants respond. Time to speak is tracked and equalised. Structural asymmetry is named openly at the start.
4
Insulate consequence from disclosure
Communicative-action forums cannot double as evidence-gathering for decisions. Anything said in a growth conversation cannot appear in a calibration deck. If this can't be true, the forum isn't communicative.
5
Prefer specific observations over character claims
'Here is what I saw in Tuesday's review meeting' is challengeable on truth. 'You're not strategic enough' is not — it's an identity claim disguised as feedback. Ban the second in real feedback conversations.
6
Publish who can end a conversation and how
In a real communicative-action space, either party can pause and re-frame without penalty. If the junior party cannot pause the conversation, it isn't communicative — it's strategic wearing developmental clothes.
7
Invest in comprehensibility
Give people the language for the four validity challenges. Coach managers to expect them and to respond to them without escalation. This is the practical version of Habermasian 'ideal speech situation' training.

The honesty test for any feedback programme

Ask: 'What happens to someone junior who tells the most senior person in this room they are wrong, on the record, with reasoning?' If the honest answer is 'nothing bad', you have a communicative-action space. If the honest answer is 'it depends on how they say it, and their standing, and the mood', you have a strategic space in developmental clothes. Design accordingly.

FAQ

Frequently asked questions

Is strategic action bad?

No. Negotiation, performance evaluation, sales, and most decision-making conversations are legitimately strategic. Habermas's point is not to purge it — it is to stop laundering it as communicative action, which produces cynicism at scale.

How does this relate to psychological safety?

Amy Edmondson's psychological-safety construct is roughly the empirical operationalisation of Habermas's conditions for communicative action in team settings. Safety is what it feels like when the four validity claims can be raised without penalty.

How does this relate to Foucault?

Foucault tells you the disciplinary conditions that make honest speech dangerous. Habermas tells you what honest speech would require if the conditions were repaired. They are complementary — Foucault the diagnosis, Habermas the design brief.

Doesn't real leadership sometimes require just telling people?

Yes — and that's fine, if it's named as strategic action. The failure mode is telling people while claiming to have a dialogue. Say 'this is a decision, here is the reasoning, questions welcome' rather than performing a dialogue that has already been decided.

Takeaways

Speech is either oriented to success (strategic) or to mutual understanding (communicative). The two require different conditions and behave differently.
Every serious feedback exchange raises four validity claims — truth, rightness, sincerity, comprehensibility — and each must be challengeable for communicative action to be real.
Most workplace feedback fails because strategic action is dressed as communicative action inside power gradients with no structural protection.
Design feedback by separating strategic and communicative forums, protecting symmetry, insulating consequence from disclosure, and coaching people to raise the four challenges without escalation.
The test for any feedback programme: can the most junior person tell the most senior person they are wrong, on the record, with reasoning, and nothing bad happens? If not, you have strategic action in developmental clothes.