Where Does Fact End and Interpretation Begin?

There is no single magical line between fact and interpretation

The word fact often sounds as if it names something completely raw and direct. In practice, even the simplest research sentence contains conceptual choices. If we say “the thermometer displayed 38.4 °C at 14:03,” we have selected an instrument, a unit, a time of reading and a rule by which the display counts as a temperature measurement. This is still a highly direct claim, but it is not a view from outside all methods and concepts.

Philosophy of science therefore warns that observation and theory are not perfectly separate worlds. Instrumental data are produced through procedures, models, calibrations and assumptions. Yet this does not imply that all descriptions are equally arbitrary. A good procedure is precisely one in which we can show how the datum was produced, which assumptions were needed and how an error could be detected.

For this article, a ladder of transitions is therefore more useful than a sharp border: trace or event → observation or measurement → record → source → reconstruction → interpretation → hypothesis → broader conclusion. The more transitions we make, the more additional support the conclusion requires.

Begin with the most modest description possible. On the table is a document with a date, signature and particular wording. Three people are visible in a photograph. A sensor recorded a specific value. A pattern is carved into a stone. These are claims about traces, objects or results that others can usually inspect again.

Even here we must be precise. “The document exists” is not the same as “everything written in the document is true.” “The photograph shows a person with a raised hand” is not the same as “the person is expressing a particular political intention.” “There are twelve marks on the wall” is not the same as “the marks represent twelve months.”

The first rule is therefore: a description should contain as little unspoken meaning as possible. If understanding the sentence already requires knowing intention, cause or symbolism, we have probably moved into interpretation.

Observation and measurement are not method-free

With scientific data there is a temptation to treat a number as a pure fact. But a measurement is the result of a relation among the measurand, instrument, procedure, calibration, sampling and model. NIST therefore insists that a high-quality measurement result should also include a statement of uncertainty.

If two instruments show slightly different values, reality has not thereby become subjective. It means we must understand resolution, systematic error, environmental conditions and fitness of method. Transparency about procedure restricts the room for arbitrary interpretation.

The lesson generalises: the more a “fact” depends on data processing, classification or a model, the more important it is to show the path from raw signal to published result.

In historical research, a primary source is often the closest surviving trace of an event, but primariness by itself does not guarantee truth. A memorandum may document what an official wrote; it does not automatically prove that the official’s description of events was accurate. A diary can be excellent evidence of what its author believed without being a neutral overview of the entire situation.

This is why the U.S. National Archives separates observing a document, understanding its origin and using it as historical evidence. The American Historical Association similarly stresses integrity of the historical record while acknowledging that interpretations evolve as documents are placed into wider contexts.

A useful formula is: a source proves itself first — that a particular record, object or recording was created in a particular context. What else may be inferred from it is the next question.

Provenance and context determine how much a source can carry

The same sentence has different evidential weight if it appears in an original signed minute, a later copy, a newspaper summary or a memory written decades afterwards. We therefore ask about provenance: who created the source, when, for whom, for what purpose, how it was preserved and whether we have an original, a copy or a third-hand quotation.

Context is not a device for relativising inconvenient content. It is a condition for knowing what the content means. A word, symbol or gesture can carry different meanings across periods and communities. Without temporal, linguistic and cultural placement we may project a modern meaning into the past.

For an important claim we therefore always separate what the source contains from what we think that content means in its setting.

History, archaeology and many other fields often investigate events that cannot be repeated. We know the past through documents, material remains, dating, testimony, environmental traces and later consequences. From these elements we build a reconstruction.

A reconstruction is not mere invention, but neither is it identical with any single source. Its strength depends on how many independent traces converge, how well we know their provenance, what is missing and whether rival reconstructions explain the record equally well.

Zvezdno izročilo človeštva applies this distinction very practically: a cuneiform tablet, a modern philological edition, a nineteenth-century ethnographic record, a living community tradition and an astronomical catalogue answer different questions. A catalogue can tell us which physical object we identify with a star today; by itself it does not tell us what that star meant to a particular culture.

Interpretation asks about meaning, intention and cause

When we ask “what did the author mean?”, “why did a community build this object?”, “what caused the political change?” or “what does this symbol mean?”, we are already in the interpretive part of inquiry. That is not a weakness of the humanities. Meaning, intention and social action are often precisely what we want to understand.

Hermeneutics developed an entire philosophical tradition around the fact that understanding human texts and actions is interpretive. But interpretation is not therefore without standards. A good interpretation must respect the text, historical context, language, counterexamples and rival readings.

In history, the same sequence of events may support debate about different causes. We must therefore distinguish the sentence “this happened” from “this happened primarily because of X.” The latter may be very well supported, but it requires additional argument.

This problem is especially important in astro-theology, mythology and comparative religion. Two myths may contain twelve figures; two cultures may use a similar serpent, tree or solar symbol; two rituals may occur in the same part of the year. Similarity is data for comparison, not proof of common origin.

Astro-teologija – skriti jezik neba therefore uses a foundational rule: an astronomical phenomenon, a historical tradition and a symbolic interpretation are not the same thing. An astro-theological reading should be labelled as such; it may be a well-documented ancient connection, a later tradition or a modern comparative interpretation.

For a historical connection we need additional bridges: chronological possibility of contact, a geographic route, textual or material traces of transmission, specific shared features and, where possible, intermediate stages. The larger the distance between similarity and a claim of origin, the greater the burden of evidence.

A hypothesis is an interpretation that accepts risk

An interpretation becomes stronger as research when we derive consequences from it that distinguish it from competitors. If we propose that a sequence of marks is a calendar, we should ask what would have to look different in the data if that explanation were correct and what rival explanations would predict.

This is where the difference between an interesting story and a hypothesis becomes visible. A hypothesis should permit at least some discrimination among possibilities. If every possible discovery confirms our idea and nothing could weaken it, its research power declines.

The modern debate over alleged Upper Palaeolithic lunar calendars is a useful methodological example: the issue is not whether a lunar interpretation is imaginative, but whether the data and tests actually distinguish a calendrical explanation from alternatives.

Inquiry needs ideas that extend beyond current evidence. Speculation can be the beginning of an excellent hypothesis. The problem begins when the narrative loses the label showing that the bridge between datum and conclusion is still long.

We can use a simple linguistic test. “It is possible,” “is consistent with,” “suggests,” “supports,” and “proves” are not stylistic variants of the same sentence. Each requires a different evidential strength. If we exchange them merely to make the story sound more decisive, we have replaced evidence with rhetoric.

THY-REALITY therefore does not prohibit unusual possibilities. It only requires that a possibility not receive the status of reconstruction, a reconstruction the status of fact, or an interpretation the status of a direct source without additional support.

Absence of evidence and evidence of absence are not the same thing

If we do not find a document in an archive, it does not automatically follow that the document never existed or that the event never happened. Archives are incomplete; objects decay; records disappear; some things were never recorded. This is absence of evidence.

Sometimes, however, an absence becomes informative. If a hypothesis were true and we would with high probability expect a particular trace, while a well-surveyed archive or measurement repeatedly fails to show it, the absence can become evidence against the hypothesis. The key question is: how strongly should we expect the trace if the claim were true?

Therefore “there is no evidence” is not enough. We should state what was searched, where, how complete the record is and whether the expected trace should have survived at all.

Data often do not determine a single explanation. Philosophy of science calls this underdetermination: at a given time the same body of evidence may be compatible with multiple models. This is not permission to choose our favourite story, but an invitation to seek a datum on which the models diverge.

A good researcher therefore asks not only “What supports my interpretation?” but also: “What would I expect if alternative A were correct? What about B? Which observation would actually discriminate between them?” Such discriminating evidence is often more valuable than ten additional examples compatible with every option.

Independent replication, another method, another archive or another kind of trace can reduce shared error. Convergence is strongest when different evidential routes do not share the same weak link.

Twelve steps from datum to an honest conclusion

For a demanding claim we can use a practical protocol: 1. What is the direct object, signal or experience? 2. What was actually observed or measured? 3. How was the record produced? 4. What is the provenance of the source? 5. What does the source explicitly say? 6. What must be reconstructed? 7. Which part interprets meaning, intention or cause? 8. Which alternatives explain the same datum? 9. What evidence would discriminate among them? 10. Are we making a comparison or claiming historical connection? 11. What is the uncertainty? 12. Is the language of the conclusion stronger than the evidence?

This protocol does not turn history into a laboratory or symbolism into mathematics. It does allow the reader to see where human inference enters the argument. That is enough to compare interpretations by quality without pretending interpretation can be eliminated.

This article completes the first four-article foundation of the new branch: What Do We Mean by Reality? defined levels of reality, Perception Is Not Reality: How the Brain Builds Our World the limits of perception, How Do We Know What We Know? Evidence, Belief and Certainty evidence and certainty, and this article the path from trace to explanation. Mind, Brain and Consciousness: Three Things We Often Confuse then opens a new question: what do we mean by mind, brain and consciousness — and why should we not use those three terms as synonyms?