The previous post argued that scientific objects are not primitive entities, but perspectival actualisations within structured systems of possibility. If that is so, then experiments cannot be understood as neutral probes of pre-existing things.
They must be doing something else.
This post asks what that is.
The passive picture
Experiments are often described as ways of finding out how things already are.
On this picture:
the world possesses determinate properties,
experiments reveal those properties,
and measurement records values that exist independently of the act of measuring.
This passive picture is deeply intuitive — and deeply misleading.
It cannot account for why experimental design matters so much, why results depend on apparatus, or why the same “object” can yield different outcomes under different conditions without error or deception.
These are not peripheral complications. They are central features of scientific practice.
Experiments as cuts
From within a relational ontology, an experiment is not a window onto reality.
It is a cut.
An experiment imposes constraints that force a system to actualise one region of its possibility space rather than another. It does not merely observe what happens; it makes something happen under controlled conditions.
What is produced is not raw data, but a phenomenon — a first-order meaning stabilised by the experimental configuration.
There is no phenomenon “behind” the experiment waiting to be revealed. The phenomenon is the result of the cut.
Measurement produces phenomena
This is most visible in measurement.
Measurement is often treated as a technical step: the translation of a property into a number. But this presupposes that the property is already there to be translated.
In practice, measurement:
defines what counts as a relevant distinction,
fixes scales and thresholds,
excludes alternative actualisations,
and coordinates construal across observers.
In doing so, it produces the phenomenon it reports.
This does not make measurement arbitrary or subjective. On the contrary, it explains why measurement must be so carefully standardised: only through tightly constrained cuts can phenomena recur.
Reproducibility without revelation
Reproducibility is often taken as evidence that experiments reveal mind-independent reality.
But reproducibility can be understood more precisely.
What is reproduced is not access to the same underlying object, but the stability of a configuration of constraints. When the same cut is made again, the same region of possibility is actualised.
Reproducibility is therefore a property of systems and practices, not of isolated things.
This is why reproducibility fails when conditions drift, apparatus changes, or background assumptions shift — not because reality has changed, but because the cut no longer holds.
Objectivity as coordination
Once experiments are understood as cuts that produce phenomena, objectivity must be rethought.
Objectivity is not achieved by removing observers, intentions, or contexts. It is achieved by coordinating construal.
Protocols, standards, instruments, and shared languages function to align cuts across time and space. They ensure that different researchers make the same cut, so that the same phenomenon can be actualised again.
Objectivity is therefore relational rather than transcendental: it arises from shared constraint, not from access to a view from nowhere.
Why this matters
Understanding experiments as ontological acts rather than epistemic probes clarifies several persistent tensions:
why experiments are creative without being fictional,
why intervention is inseparable from observation,
why scientific practice cannot be reduced to theory alone,
and why disputes about what is “really there” often miss the point.
Science does not discover phenomena by stepping back from the world.
It brings phenomena into being by holding certain possibilities steady.
Toward incompleteness
If experiments are cuts that actualise regions of possibility, then no finite set of experiments can exhaust a system.
There will always be further cuts that could be made, further configurations that could be stabilised, further phenomena that could emerge.
The next post turns to this openness directly, asking why scientific laws and models cannot close their own domains — and why this is not a failure, but a necessity.
No comments:
Post a Comment