Appearance and Observation

T3 is the third independent theory document in programme T1–T9. It examines how something appears, is registered, bounded as an observation, and carried forward into interpretation, inference, and claim.

T1 established that actuality is not created by representation. T2 established conditions for knowledge, traceability, and epistemic status. T3 locates the access event between these levels: observation is an actual event within actuality but is not automatically identical with the observed state of affairs.

The document does not by itself determine what a particular observation proves, what cause lies behind a registration, or what professional, legal, or institutional conclusion should follow. It establishes distinctions that such assessments must keep visible.

A state of affairs may obtain without anyone observing it. Observation therefore does not create the observed as ontological ground. At the same time, the observation is itself something that occurs: an observer, instrument, medium, or system is affected, registers, and bounds something within actuality.

The observer does not stand outside the world and look into it from a neutral exterior. Body, position, time, instrument, language, and purpose are themselves actual conditions within the access event. This makes observation situated, not arbitrary.

T3 rejects both the unqualified identification of appearance with the whole of actuality and the sceptical confinement of the human being to a closed inner image. Access is real but bounded and reviewable through the conditions under which it occurs.

Appearance is the way something becomes manifest, affects an observer, is experienced, or leaves traces under particular conditions. It is not nothing and not merely a linguistic invention; appearance is an actual event or state in the relation among a state of affairs, a medium, and access.

That something appears does not by itself determine what appears, what caused the appearance, or whether the observer has assigned the correct object, property, or context. A shadow-form may be real as a visual experience while the first explanation of it is wrong.

T3 therefore distinguishes that something appears, how it appears, what is taken to appear, what actually grounds it, and what claim the observer later makes.

Every human observation occurs from a particular place and situation. Direction, distance, duration, body, attention, training, language, expectation, role, and purpose may affect what becomes available and what is overlooked.

Situatedness is not relativism. It gives reason to document observational conditions and compare different positions of access. Some positions may be more relevant, precise, or suitable than others for the question at issue.

When the observer’s position is concealed, a bounded view may be presented as complete actuality. When the position is visible, others can examine what the view supports and what lies outside it.

• Place, direction, and distance.
• Time, duration, and frequency.
• Sensory or instrumental access.
• Competence, expectation, and attention.
• Role, mandate, interest, and purpose.
• Known blind spots and conditions not examined.

Sensory impact means that a sense organ is physically affected. Sensation is the biological and experiential registration of that impact. Attention selects something as relevant. Observation identifies, bounds, or documents something as observational content.

Much may be sensed without being observed, and much may be observed without being understood. A correct registration may be overlooked, while a strongly expected detail may be added to the observation despite not having been registered.

Observation is therefore not merely passive reception, but neither is it free construction. It is a bounded act in which something appearing is selected and made available for further recognition.

Instruments may make conditions available that cannot be sensed directly, and may stabilise, amplify, count, or preserve what would otherwise disappear. Microscopes, cameras, sensors, thermometers, medical imaging, and digital logs are forms of mediated access.

The instrument also introduces conditions of its own: calibration, range, resolution, threshold, noise, sampling, compression, filtering, software, and display format. The result must therefore be traced through both the instrument and the processing that made the signal readable.

This does not weaken instrumental knowledge as such. It establishes that the instrument must not disappear from the account as if the result were an unmediated copy of actuality.

Registration is a physical, biological, digital, or documentary trace of an impact, event, or state. Data are registrations selected, structured, or encoded so that they may be stored, compared, or processed.

So-called raw data are not necessarily free of presupposition. Sensor choice, measurement arrangement, timing, thresholds, file format, and storage system have already bounded what can be registered and how it is preserved.

T3 distinguishes signal, registration, raw data, processed data, visualisation, result, and interpretation. Each link may preserve, amplify, lose, or transform information. Epistemic strength depends on the transitions remaining visible.

Observation and interpretation often operate together but have different functions. Observation bounds what was registered or experienced. Interpretation assigns meaning, category, cause, or context to that content.

A description such as ‘the measured value changed from X to Y’ remains closer to the registration than ‘the change was caused by object Z’. The claim ‘object Z has property P’ is a further inference, and ‘this proves theory Q’ adds another level of claim.

That observation may be concept- and expectation-laden does not make interpretation arbitrary. Interpretation is bounded by what was registered, must be compared with alternatives, and must remain corrigible by new or better observations.

• Observation: what was actually seen, heard, measured, or registered?
• Interpretation: what meaning, category, or cause was assigned?
• Inference: what conclusion was drawn from material and interpretation?
• Claim: what is asserted beyond the particular observation?

A reviewable observation report must make both content and conditions visible. It should allow the reader to distinguish what was registered from what was interpreted, and what lay within the field of observation or measurement from what lay outside it.

The requirement varies with the matter. An ordinary observation need not carry the same documentation as a medical measurement, a legal scene examination, or a research result. But the strength of the claim must correspond to the strength and traceability of the observation.

Where relevant, the report should identify the observer, time, place, duration, position, medium, instrument, settings, conditions, registered content, unregistered area, and known sources of error.

Error may arise in registration, identification of what was registered, interpretation of cause, or the inference drawn. T3 requires the error to be located in the correct link rather than treating the whole experience as simply true or unreal.

In illusion there is an actual appearance or experience assigned to the wrong object, property, or cause. In hallucination an experience arises without the external state of affairs that it is taken to refer to. In fabrication material is deliberately created or altered and presented as authentic registration.

In every case the experience, registration, or act of fabrication is itself something that occurs. It does not follow that its content, causal assignment, or representation is true.

An observation may be strengthened when repeated under comparable conditions, registered by independent observers or instruments, or supported by different access paths converging on the same state of affairs.

Repetition is not automatic confirmation. The same systematic error may recur when observers use the same calibration, data source, software, or mistaken category.

Triangulation is strongest when access paths are genuinely independent and possess different failure modes. Multiple copies of the same origin are not multiple independent observations.

A witness may make a personal observation available to others as mediated access. The account must then distinguish what the witness experienced, consciously observed, remembers, interpreted, and later learned from others.

Time between event and record, questioning, repeated narration, and social influence may change an account without deliberate lying. Credibility is relevant but cannot replace examination of content, access, and correspondence.

Legal, journalistic, historical, and medical work should make contemporaneous notes, independent traces, precise attribution, and the distinction between personal and reported observation visible.

The statement ‘I did not see X’ first reports an absence of registration in a particular observation. It is not automatically identical with ‘X was not present’.

Negative evidence becomes stronger when X would reasonably have been expected to register if present. The field, direction, duration, sensitivity, threshold, coverage, and possible concealment must then be sufficiently clarified.

Absence of findings may be important but must not be made stronger than the probability of detection supports. An instrument incapable of registering a phenomenon cannot refute it through silence.

Machines may register signals, classify patterns, estimate probabilities, and generate representations. These functions may extend and stabilise human access but must be distinguished from understanding, recognition, and responsibility.

An AI system’s output depends on data sources, representations, architecture, training, thresholds, instructions, and use context. A detection, classification, or generated explanation is not automatically an observation of the state of affairs about which the user takes the system to speak.

T3 does not settle whether machines can in principle possess consciousness or recognition. It establishes only that machine output in institutional use must remain traceable through registration, model, processing, and human decision, and that responsibility cannot be transferred to an undefined system concept.

Observational status states how a condition became available, not how certain the final claim is. It must therefore be distinguished from the epistemic statuses in T2.

A condition may be directly observed and still misinterpreted. Another may be inferred and nevertheless strongly established. The status identifies the form of access; epistemic assessment determines what conclusion that access can support.

Institutions should use the statuses consistently so that reported, reconstructed, simulated, and directly observed material do not merge into one linguistic category.

• Directly observed — the observer personally encountered the event or condition.
• Instrumentally registered — access proceeded through an identified and documented instrument.
• Documented — a preserved trace or contemporaneous account exists.
• Reported — the information comes from another agent.
• Reconstructed — the event is assembled from several traces and inferences.
• Inferred — the condition is concluded through valid inference.
• Simulated — the result is produced within a model or calculation.
• Unobserved — no identified observation of the condition is available.
• Unresolved — the material does not support a defensible observational status.

T3 must remain answerable to positions that make observation too direct, too indirect, or too theory-dependent. The replies below do not close the questions but state what the document must be able to support.

T3 enters discussion with direct and indirect realism, critical realism, empiricism, phenomenology, philosophy of perception, observation theory, measurement theory, and scientific realism. It shares the problem of how access can be both grounded in actuality and fallible.

In relation to direct realism, T3 maintains that the observer encounters actuality while denying that appearance always exhausts the object. In relation to representational theories, it acknowledges intermediaries and processing while denying that these automatically confine the observer within a self-sufficient representational space.

Hanson and Kuhn made theory-dependence of observation central in philosophy of science. T3 treats this as a genuine challenge but requires theory-dependence itself to be documented as a relation among observer, category, practice, and material. The fuller comparison and literature criticism belong to T9.

T3 requires later theory and method documents to keep observation, registration, interpretation, inference, and representation distinct. They must also document position, instrument, processing, uncertainty, and plausible alternatives.

If T3 is substantively revised, T4, T5, R1, and the operative audit types in particular must be reviewed because they use observation and documentation as grounds for linguistic claims, models, findings, and correction.

• T4: Observation reports must be distinguished from linguistic interpretation and claim.
• T5: Data, visualisation, and model must be distinguished from the observed condition.
• T6: Self-experience and identity representation must be examined without dismissing experience.
• T7: Institutions must document how observation is registered, transformed, and made into a decision ground.
• T8: Practice must build on visible observational conditions and marked uncertainty.
• T9: The theory of observation must be compared, bounded, and criticised.
• R1: The methodology must distinguish material, observation, interpretation, finding, and decision.
• A1–A7: Every audit type must document conditions of access and intermediaries.
• R7: Worksheets must keep observation and inference in separate fields.