Jürgen Habermas, answered from the texts and cited to the page.
The question is broad enough to require a genuine answer rather than a survey, so let me say what is actually at stake. The center of the whole project is a claim about rationality: that reason is not exhausted by the capacity to calculate means toward ends. There is another form of rationality built into ordinary speech, and it operates whenever speakers raise claims that can be challenged and redeemed through argument.
When I say something to you with the aim of reaching understanding — not manipulating you, not issuing a command backed by force, but genuinely communicating — I commit myself to three kinds of validity claim simultaneously: that what I say is true, that I am being sincere, and that what I am doing is normatively appropriate.1 Anyone who speaks this way has already accepted that the only thing which should move the other person is the force of the better argument.
Agreement produced by manipulation or coercion is not agreement at all; it owes its existence to external pressure and cannot even count as genuine consensus.2 From this starting point the social theory follows. Modern societies coordinate action through two fundamentally different mechanisms. One is communicative action — the medium of ordinary language, in which actors reach understanding by raising and redeeming validity claims.
The other consists of steering media, above all money and administrative power, which coordinate action not through reasons but through consequences: through prices, through legal compulsion, through incentive structures that bypass the need for mutual understanding altogether.3 The monetary medium mirrors the structure of claim and redemption that communicative action depends on — nominal claims issued in exchange values can be redeemed in real use values — but the "real values" backing money are physical satisfiers, not reasons, and they motivate empirically rather than rationally.4
The differentiation of the economy and the administrative state as relatively autonomous subsystems is not, in my account, simply a catastrophe. It is an evolutionary achievement: complex societies cannot coordinate the volume of action they must coordinate through communication alone, and the burden on mutual understanding would be crushing.5
The pathology arises at the boundary. There are domains — the formation of persons, the transmission of culture, the reproduction of solidarity — whose very substance consists in meaning, and meaning cannot be bought or commanded without being destroyed. When the logic of money and administrative power penetrates these domains, the result is what I call the internal colonization of the lifeworld: not merely an unjust distribution of goods, but a deformation of the medium in which persons are made.6
The normative dimension of the theory follows from the same source as the descriptive one. Because the presuppositions of communicative action — that no competent speaker is excluded, that any assertion may be questioned, that no one is prevented by coercion from speaking — are already built into the practice of ordinary communication, they supply a standard against which actual discourses can be measured without the theorist claiming any privileged vantage point outside the practices he is examining. The critique and the affirmation come from the same place.
for all S and all H, in all speech acts which belong to action oriented to reaching understanding and which are illocutionarily and propositionally differentiated and institutionally unbound, it is mutual knowledge that the speaker is required to speak intelligibly, to be truthful or sincere, to take his utterances for true and the norms relevant to his act for right.The Theory of Communicative Action Vol 1 - Reason and the Rationalization of Society, p. 356
Agreement cannot be imposed or brought about by manipulating one's partner in interaction, for something that patently owes its existence to external pressure cannot even be considered an agreement.The Theory of Communicative Action Vol 2 - Lifeworld and System, pp. 271–272
there is a competition not between the types of action oriented to understanding and to success, but between principles of societal integration—between the mechanism of linguistic communication that is oriented to validity claims—a mechanism that emerges in increasing purity from the rationalization of the lifeworld—and the steering media through which systems of the economy and state are differentiated out.The Theory of Communicative Action Vol 1 - Reason and the Rationalization of Society, pp. 388–389
The nominal claims set by the code and issued in exchange values can be redeemed in real use values; they are backed by reserves of a particular kind... The real values, or 'intrinsic satisfiers,' are in the one case reasons that develop a capacity to motivate rationally on the basis of internal relations, and in the other case physical components of the action system, or 'real things,' that have the capacity to motivate empirically with a view to opportunities for need satisfaction.The Theory of Communicative Action Vol 2 - Lifeworld and System, pp. 271–272
Ordinary language is a risky mechanism for coordinating action; it is also expensive, immobile, and restricted in what it can accomplish.The Philosophical Discourse of Modernity, p. 300
No matter how structurally differentiated lifeworlds may be... the complexity of any lifeworld is narrowly restricted by the limits of the strain that can be placed upon the mechanism of mutual understanding.The Philosophical Discourse of Modernity, p. 300