Working papers

Before We Say ‘Consequence’

A definition, two questions it leaves open, and why consequences so often fail to come back

Terry Cooke-Davies. Working paper. Draft 1, 22 October 2026. Comments and disagreement welcome.

Why a word matters

In ordinary speech a consequence is something that happens to someone because of something done. The word is so familiar that it seems to need no definition. Yet in any inquiry into why harm persists — why individuals, institutions and whole societies go on doing what damages them and others — it is quietly asked to do a second job. It names not only what happens, but what comes back: the difference that reaches someone, is noticed, and changes what they do next.

The two jobs are easily run together, and the running together has a cost. To say that consequences should be allowed to reach us can be heard, in the first sense, as a call to learn by suffering: passive, backward-looking, and a poor rival to the call for better choices. In the second sense it is nothing of the kind. It is a call to keep open the channels through which what our actions cause can find its way back to those able to respond, before the harm compounds. That is itself a choice, and an active one.

This note is a companion to Before We Say ‘System’, and follows the same course: a working definition, the questions it leaves undeclared, and what follows when they are declared.

The definition

A consequence is a difference in the world that follows from an act or omission and can recognised as stemming from it.

Three choices in that sentence are deliberate.

A difference, rather than an event. Many consequences are not happenings at all. They are states — a depleted aquifer, a degraded soil — or absences: the disagreement missing from the minutes, the bird no longer heard in spring, the baseline that has shifted between one generation and the next. The word ‘difference’ admits all of these, and echoes Bateson’s ‘difference that makes a difference’.

An act or omission, rather than an action. The meeting not held and the warning not given have consequences as surely as anything done. The law has long used the paired phrase for this reason.

Can be traced to it, rather than simply results from it. ‘Traced’ not in the sense of constructing a causal map, but rather meaning ‘can be recognised as stemming from’. Here the word ‘consequence’ parts company with the word ‘effect’. An earthquake has effects; we do not speak of its consequences unless we recognise them as stemming from someone’s choice — to build on the fault line, or not to strengthen the school. A consequence is an effect described as following from an act. Attribution is built into the word. This is what makes it possible for a consequence to return at all: it carries the address of an actor.

The actor may be a person, a group, or an arrangement — a market, an institution, a body of rules. When it is an arrangement, a second definition becomes necessary, because an arrangement has no body to which anything can return.

A consequence returns when it reaches, and is recognised by, someone with standing to alter the act that caused it — ideally the one who acted.

The definition separates the actor from the responder. Where they coincide, return is ordinary learning. Where they come apart, return depends on someone else carrying the consequence back — telling it — and on that telling reaching someone with standing to act. A consequence returned to ‘the institution’ means nothing unless it reaches a living person within it who can change what the institution does.

Two questions the word leaves undeclared

Every use of the word ‘consequence’ leaves two questions open, and is incomplete until they are answered.

The first question: where does it land?

On the actor, or on others: other people, other creatures, the future.

The second question: does it return?

Is it registered, and recognised as stemming from the act, by someone with standing to alter it?

The two questions are independent of one another, as the table shows.

Two of the four cells are failures, and they are not the same failure. The lower right is severance: the consequence lands outside the loop altogether, on those with no voice in it. Much of the ecological and social harm of the modern world lies in this cell. But the upper right shows that a consequence can land squarely on the actor and still fail to return, because it is never registered, or never recognised as stemming from the act that caused it. The arrangements that prevent consequences from coming back are therefore not only arrangements of displacement. They include whatever governs what is noticed and how it is explained.

The return, stage by stage

Read together, the two definitions imply that a return has stages, and that each can fail on its own.

First, the difference occurs — as an event, a state, or an absence.

Second, it reaches someone. This is where severance operates: the difference lands where no one who could respond is placed to receive it.

Third, it is noticed. What registers is governed by what the receiver is ready to notice. Absences are the hardest of all, because nothing presents itself; an absence is noticed only by someone who remembers what should be there.

Fourth, it is recognised as stemming from the act. A difference noticed but explained some other way — as bad luck, as the weather, as someone else’s doing — does not return to the act that caused it. What causal stories are available depends on the maps the receiver holds.

Fifth, someone with standing responds. Even when everything before has worked, an arrangement may still break the chain here: the person who noticed and understood has no power to change the act, and no route to anyone who does.

Implications

Appreciative settings and the maps that shape them

The third and fourth stages are governed by what Geoffrey Vickers called the appreciative system: the settings, formed by experience, that make us ready to notice some things and not others, and to value what we notice by particular standards. Vickers saw regulation as the work of keeping relationships within acceptable bounds, and the signal on which it runs is a mismatch between what is and what should be. A consequence, in the second sense of the word, is just such a signal. Appreciative settings decide whether it arrives as one.

Those settings are not formed only by what has returned. They are formed, far more, by the maps we inherit: the stories, categories and explanations that tell us what to look for and how one thing leads to another. Maps make attribution possible, and they also limit it. Where a map has no place for a causal link, the consequence that travels along that link cannot be recognised as stemming from the act, however plainly it arrives.

What travels and what stays behind

The briefing What Travels and What Stays Behind proposes why the settings we inherit drift from the consequences we now face. The learning that built human culture copies what visibly succeeds. Production is visible; successful regulation is not, because it shows itself only as nothing going wrong. Maps therefore travel without the consequences that formed them, and appreciative settings are handed on tuned to what earlier generations could see. For most of history consequences returned quickly enough to correct this. Where the path of return has lengthened, the inherited settings govern unchecked.

In the terms of this note, that is a failure at the third and fourth stages, compounded across generations, and it falls hardest on absences. Regulation that succeeds produces an absence — the harm that did not happen — and an absence returns to no one who has not been taught to look for it.

Consequences and indicators

There is a second sense in which consequence can be relocated from the world into the map. The managerial vocabulary of ‘outcomes’ and ‘impacts’ names differences that someone has chosen to measure. An indicator is a constructed return: a channel built so that some consequences come back by design. That is often valuable. But a constructed return can be gamed, and once it is, it reports what the indicator rewards rather than what happened. The indicator then stands where the consequence should be, and the real difference, unmeasured, falls into the upper right-hand cell of the table.

Ashby’s distinction applies here as it does to the drawing of boundaries. There is freedom in what we choose to count as a consequence. There is no freedom in whether reality bears the choice out.

Diagnosis and remedy

Recognition Theory is, in these terms, an account of why returns fail at each stage: consequences displaced onto those outside the loop; proxies standing in for what they were meant to measure; visible production copied while invisible regulation is lost. Answerable Realism is the corresponding stance: keep each stage of the return open; trust what cannot be gamed on paper; ask those inside sustaining relationships what they are noticing.

One further implication bears on the relation between choosing and returning. Choices shape which consequences can come back, because arrangements are chosen; and returned consequences shape the next choices. The two are mutually enactive. But they are not equally vulnerable. A choice can sever the return; a return cannot sever choice. When severance has occurred, the cycle of choosing and learning continues in appearance while its correcting strand has been cut. The choice on which all the others depend is therefore the choice not to be exempt from what one causes: that no arrangement may be exempt from the return of what it causes.

What counts as a consequence is chosen. What happens is not.

Sources

W. R. Ashby, Design for a Brain: The Origin of Adaptive Behaviour, 2nd edn (Chapman & Hall, 1960).

G. Bateson, Steps to an Ecology of Mind (Chandler, 1972).

J. Henrich, The Secret of Our Success (Princeton University Press, 2016).

D. Pauly, ‘Anecdotes and the shifting baseline syndrome of fisheries’, Trends in Ecology & Evolution 10(10) (1995), 430.

G. Vickers, The Art of Judgment (Chapman & Hall, 1965); Freedom in a Rocking Boat (Pelican, 1972).

This note is a companion to Before We Say ‘System’ (21 October 2026), whose third question — where do the consequences go? — it takes as its starting point. It draws on the briefing What Travels and What Stays Behind (third draft, 13 October 2026).

Developed in dialogue with Claude (Anthropic) as part of an ongoing AI-enhanced appreciative practice. The thinking and every word are the author’s own.