A book rests on a shelf. Nothing visibly moves, yet its remaining there depends on the shelf, its attachments, and the structure supporting them. A message changes someone’s plans. A repair restores a damaged connection. Each example invites a causal question: how did this contribution make the relevant difference?
Changism treats causation as a mode of productive dependency within organized activity. Its starting point is concrete involvement: how something is made, transmitted, transformed, sustained, or prevented through actual processes.
This approach also separates questions that ordinary “because” language can bring together. How something was produced, what constitutes it, and what justifies a belief about it can require different explanations. Their connections matter, but one answer cannot automatically do every job.
What it means to produce a difference
Return to the shelf. Under suitable loading, contact and the distribution of stress through the shelf and its attachments sustain the book’s position. The relevant difference is between that maintained support and what would happen if the supporting connection failed under otherwise specified conditions.
Here, producing a difference includes maintaining an outcome. The account does not require the book to keep moving or the shelf to deform more and more. Static support also does not require a continuous energy supply merely to keep the book at rest. The organization and its response under loading do the explanatory work.
To call the shelf productive in this respect is to identify its actual contribution. Its colour might help a reader locate the book, while its material connections bear the load. The source must be considered under the aspect that matters to the outcome.
The word “productive” does not mean beneficial or intentional. Damage, inhibition, and decay can belong to productive histories. Nor does it add a special force between cause and effect. It names the concrete interaction through which the specified contribution occurs.
Changism calls a system’s capacity to contribute productive efficacy. The shelf may be capable of carrying another book even when no second book is placed on it. Productive dependency concerns a contribution actually made, such as the support already sustaining the first book. Capacity and exercise need separate accounts.
How this differs from correlation or sequence
Suppose two displays show matching numbers. One might copy the other. Both might receive a common signal. A test routine might supply the same value to each. The agreement leaves their productive relationship unsettled.
A causal inquiry must identify the relevant route: how the proposed source enters the process, what carries or transforms its contribution, and how that contribution reaches the outcome. Simply occurring earlier is insufficient. So is finding a regular association without distinguishing the routes that could explain it.
A route need not carry one unchanged object from beginning to end. A sensor reading can become a comparison, then a control instruction, then an adjustment. What matters is the relevant inheritance through those transformations.
The connection can also break at a join. Imagine a sensor value reaching a controller’s memory, only to be overwritten by a test value before the controller uses it. The sensor affected the memory, and the controller produced an output. Those two facts do not establish that the sensor’s value determined that output. The later operation used something else.
Measurements, interventions, and comparisons can help discriminate these alternatives. Their success depends on the conditions of the test: a backup may replace a disconnected input, or the chosen measurement may miss the contribution. A failed test of one proposed route is different from a demonstration that no relevant route exists.
The occurrence itself does not require a permanent record. An interaction can make a difference even when no durable trace survives. Records help us investigate productive histories. Losing a record does not undo the interaction it recorded.
Why a contribution is not always the cause we select
Many conditions can contribute to an outcome. Identifying them still leaves a further question: which difference are we trying to explain?
An engineer investigating a shelf’s collapse may focus on its attachments under a particular load. Someone asking why those attachments were installed may instead investigate the designer’s assumptions and how they entered the drawing and construction. These inquiries select different contributions within a connected history.
Changism therefore distinguishes establishing a productive contribution from selecting it as the cause under a particular explanatory question. That selection needs a justified outcome, comparison, and level of description. The research has not supplied a completed universal rule for making every such selection.
Background conditions belong to this choice of focus. Holding the wall’s support fixed while investigating a bracket does not make the wall irrelevant to the physical arrangement. It identifies what this inquiry is taking as given. Explanatory interests can change the question; they cannot make an unsupported account of what happened correct.
How causation differs from explanation and grounding
Explanation is the broader activity of making something intelligible. A causal explanation follows how an outcome was produced or sustained. Other explanations concern constitution, possibilities, truth, evidence, or what follows from premises.
Grounding, in the metaphysical sense, asks in virtue of what something obtains. What makes an arrangement a functioning instrument? What makes someone the treasurer of an association under its adopted rules? Such questions concern the basis of a thing, capacity, or status. They need not describe another event that produced it.
The distinctions can be set out through the questions each explanation answers.
| Explanatory concern | Question | Example |
|---|---|---|
| Production | How was a difference made or sustained? | Contact and connected supports carry a book’s load |
| Constitution | What makes this a particular thing or status? | An association’s rules and qualifying appointment establish who holds an office |
| Capacity and constraint | What does this organization permit under specified conditions? | A shelf can withstand a stated load even if it never carries it |
| Truth | What makes the claim true? | A claim that an appointment occurred depends on whether that appointment occurred |
| Evidence | What warrants believing the claim? | Assessed records and checks support an account of the appointment |
| Logical consequence | What follows from the premises? | A valid inference preserves truth if its premises are true |
These are distinctions among claims about reality, rather than six separate worlds or an exhaustive classification of every dependence.
A false notice can make people leave a meeting. Suppose it falsely states that the organizer had already cancelled it. People’s departure can be an actual consequence of encountering and believing the notice, while its claim about the organizer’s earlier action remains false. Success in producing an effect does not establish truth.
Likewise, explaining the procedure that constitutes an appointment does not establish that the required acts occurred on this occasion. That requires a history. Establishing the history requires evidence. The existence of a document, its causal effects, and the warrant it provides are connected achievements with different conditions.
For Changism, grounding can therefore be a legitimate part of explanation. The problem arises when “grounds” silently changes its job: a statement about what constitutes something is treated as if it had already explained how a particular difference was produced. Moving between those claims requires an additional reason connecting them.
What prevention and missing actions reveal
A stop catches a sliding book before it falls. Its effect can be described negatively—the book did not fall—while the explanation follows a concrete interaction. Removing the stop would also be an actual intervention, changing the arrangement the book encounters.
Now suppose someone was meant to install the stop but never did. The missing installation can matter to an explanation comparing what happened with what a properly fitted stop would have prevented. Yet the non-occurrence is not itself a physical participant sending an influence towards the book.
The Changist research proposes separating the actual processes, the relevant alternative, and any responsibility attached to the omitted task. This preserves the practical importance of omissions without treating an absence as an additional producer. The proposal still has to account adequately for the ordinary causal claims people make about failures to act.
Negative wording alone cannot settle the issue. Active suppression or maintaining a barrier can involve concrete work. We must examine what happens in the case, rather than infer an empty process from a sentence containing “not”.
Why removing a cause may leave the outcome unchanged
Imagine two supports sharing a load, each capable of carrying it alone. Removing one may leave the book at the same height while changing the load carried by the other. Both contributed before removal. Measuring only whether the book fell would miss that contribution.
A backup introduces a different history. It may remain idle until the first support fails. The same final position can therefore accompany either shared support or replacement by a previously unused support. Understanding the organization lets us distinguish them.
This is why productive dependency is not simply the rule “without this, the outcome would not have occurred”. Such comparisons are useful, but redundancy, replacement, and interference require a more detailed account of the actual route.
Reciprocal influence also needs precision. The book’s load contributes to the shelf’s deformation; the shelf’s response sustains the book’s position. These are directed contributions under different aspects. Their coexistence does not make every dependence reciprocal or settle how all dependencies in reality are arranged.
Whether change requires a changeless bearer
Changism’s answer begins with a different order of explanation: productive activity is fundamental, and particular things are understood through its organization. A person, stone, or instrument remains real. Its reality does not require an additional, fully independent bearer beneath its properties, relationships, and history.
A candle flame helps make the proposal concrete. Fuel is supplied, combustion occurs, and heat helps sustain the conditions for continued burning. The flame persists through this connected activity. Wax, the wick, and surrounding air are real material conditions. Accounting for them does not require a further hidden flame that first exists complete in itself and then acquires burning as an attachment.
Changism calls its starting commitment productive act: concrete reality understood as actually making, undergoing, sustaining, and constraining differences. “Act” does not imply an intention or a cosmic agent. It also contributes no extra ingredient to the physical account of combustion.
The proposal must accommodate stability as well. A stone does not have to behave like a flame or continually replace its material. Its organized structure and capacities can persist while particular responses remain unexercised. Treating activity as fundamental is not permission to deny that ordinary things can rest.
Explanations can still go deeper. We may investigate the material processes composing an organism or the medium carrying a signal. Changism allows a substrate in this sense: an underlying organization that realizes or supports another. It questions the further promotion of that basis into a relation-free, changeless bearer.
Nor does the argument demand infinitely smaller components. Being elementary within a physical theory does not automatically mean being independent of every relation. Size, composition, explanatory priority, and independence are different questions.
What would justify this starting point
A strong substance theory can give its beings genuine powers. A structural account can represent detailed interactions and histories. Changism must compare itself with those alternatives, rather than assume that every rival begins with powerless objects.
Its challenge is that an account beginning from a wholly nonproductive basis must explain how concrete production obtains. If the explanation introduces manifestation, transmission, or realization, it must say what work those relations perform. The requirement is for an adequate account, not a demand that every kind of explanation secretly be a cause.
The same demand applies to Changism. Calling an interaction productive must help distinguish its source, route, outcome, and relevant alternatives. If another account preserves those distinctions equally well, the examples alone establish no Changist advantage. Evidence for a particular causal route does not deductively prove that productive act is fundamental.
The proposal is thus a philosophical interpretation with concrete obligations. It must explain exercised contributions and unused capacities, accommodate the difficult cases, and keep causal production distinct from the other kinds of dependence that explanation requires.
Read further
Begin with Changism, Volume I, Part II. Chapter 12, “Activity Without a Hidden Actor”, develops the philosophical starting point. Chapter 13, “Where Things Meet”, examines interaction and registration. Chapter 14, “How One Thing Makes a Difference”, follows productive routes and different kinds of explanation. Chapter 15, “Why Change Has a Shape”, develops capacities and constraints.
For the technical arguments, visit Research. “The Productive Core of Causation” specifies efficacy, productive routes, and difficult cases. “Dependence Across Registers” examines the different meanings of dependence. “Act Without a Subject” and “Against the Final Brick” develop the comparative arguments about what an account of concrete reality should take as fundamental.
Continue exploring: Things and Identity examines the organization through which things persist. Truth, Records, and Knowledge separates what happened from its records and our reasons for believing it. Time and Change distinguishes productive order from temporal measurement.
Chapter numbers refer to the second edition of Changism, Volume I.