Long before computers, people used omens to make uncertainty actionable: they treated events such as eclipses, dreams, births, and animal organs as signs, interpreted them through established rules, and used the result to guide decisions. Mesopotamian omen traditions were organized bodies of knowledge, not just isolated guesses. Their comparison with algorithms is useful—but only in a limited sense: both can turn observations into recommendations, while omens were not algorithms in the modern computational sense.
What did people use before algorithms to make decisions under uncertainty?
They used several methods, including divination: practices for interpreting signs and deciding what those signs might mean for the future. The surviving written evidence is not a record of one universal system. It includes traditions across Mesopotamia and the wider ancient Near East, as well as Egypt, Greece, Italy, and China.
Oxford Bibliographies places written omen evidence in Mesopotamia in the early second millennium BCE and describes divination as a way to handle uncertainty—to warn or reassure an individual or a whole community. Lorenzo Verderame reports several hundred cuneiform tablets documenting celestial divination, with evidence spanning from the late third millennium BCE to the end of the first millennium BCE. These dates describe different bodies of evidence, not a single starting point for every kind of omen practice.
The appeal was practical as well as religious. When a person or ruler faced multiple possible actions but could not know their consequences, an omen offered a procedure: observe a sign, consult an interpretation, then consider a response. It did not make the future certain. It gave people a framework for acting despite not knowing it.
#1 Best Overall
How did Babylonian diviners interpret signs?
They drew on many kinds of observations
Mesopotamian divination included celestial events, dreams, births, animals, atmospheric phenomena, and the inspection of organs from sacrificed animals. A Metropolitan Museum of Art record describes a clay tablet fragment containing prognostications based on the condition of a sacrificial sheep’s liver. The museum explains that the practice facilitated communication between the divine and human worlds.
Interpreting such signs was a specialist activity. The Cambridge History of Science identifies distinct roles in the Mesopotamian scholarly repertoire, including celestial scholar, liver diviner, exorcist, physician, and lamentation priest. A sign did not interpret itself: training and inherited knowledge shaped what an expert saw in it and what advice followed.
Written compendia organized signs into patterns
A 2024 article by Andrew George and Junko Taniguchi presents four Old Babylonian tablets as the oldest known compendia of lunar-eclipse omens. Their contents organize interpretations around features such as the time of night, the movement of the shadow, the eclipse’s duration, and the date. This is evidence of systematic classification: observations were sorted into categories that readers could consult, rather than treated as wholly unprecedented events.
Francesca Rochberg’s analysis of Babylonian omen statements cautions against reducing them to simple claims that one physical event invariably causes another. Their structure often works conditionally: if a particular sign occurs, a possible consequence may follow. The Cambridge History of Science describes this as inferential reasoning expressed through conditional form. An omen therefore linked observation to a possibility; it was not a modern physical law or a guarantee.
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →Were omens an early form of data analysis?
There is a limited structural resemblance. Both omen systems and algorithmic systems can classify inputs, apply rules, and produce an interpretation or recommendation. But that similarity does not make ancient divination a precursor equivalent to a computational model. The evidence supports a comparison of how people organized inference and authority under uncertainty—not a claim that omens were algorithms in the modern technical sense.
| Comparison point | Omen systems in the evidence | Algorithmic systems, in general |
|---|---|---|
| Input | Observed signs such as celestial events, dreams, or animal organs | Data selected and supplied to a model or procedure |
| Rule | Inherited interpretive categories, often expressed as conditional associations | A computational procedure or model transforms inputs into outputs |
| Interpreter | A trained diviner or other specialist | People who design, operate, or use a model, sometimes through a platform |
| Output | A possible warning, reassurance, ritual response, or decision recommendation | An estimate, classification, ranking, or recommendation, depending on the system |
| Authority | Could be connected to temples, courts, scholarly archives, or specialist roles | May depend on the institution deploying the system and users’ trust in it |
| Potential failure | Human error, corruption, motivated interpretation, or political legitimation are concerns raised in scholarship on Near Eastern augury | Selective data, opaque expertise, motivated interpretation, and retrospective validation are useful risks to examine; this comparison does not establish how often they occur |
The table is an analytical comparison, not evidence that the two systems are interchangeable. Omen interpreters worked within religious and political worlds whose assumptions differ from those behind contemporary computing. And an algorithm’s output depends on its data, design, and use in ways that cannot be inferred from ancient omen texts.
Did Romans and Chinese thinkers believe in divination?
There was no single, uniform answer. A 2025 comparative study, “Beyond Doubt,” examines Cicero’s De Divinatione alongside the Chinese thinkers Dong Zhongshu and Xunzi. It finds tension between speculative claims about signs and the practical political or ritual reasons societies might preserve divinatory practices.
Cicero made disagreement part of the discussion
De Divinatione both considers arguments in favor of divination and questions whether it is valid. That makes Cicero a useful example of internal skepticism: a society could maintain divinatory customs while debating the evidence and credibility behind them. The historical record does not support the idea that every participant simply accepted every omen as true.
Best Value
Chinese debates also involved more than a yes-or-no belief
The comparison with Dong Zhongshu and Xunzi likewise points to tensions between claims about signs and their roles in political or ritual life. The traditions are not identical, and the comparison does not make Roman and Chinese views interchangeable. It does show why the question “Did they believe?” can be too blunt: people could dispute how signs worked while continuing to use practices with social or political significance.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What can occult history teach us about bias and prediction?
Omen history is useful for asking who gets to interpret ambiguous evidence, which rules shape an interpretation, and what happens when a prediction supports someone’s interests. Scholarship on Near Eastern augury highlights insecurity, the possibility of human error or corruption, and the political use of omens to legitimate authority. Those are documented concerns, not measurements of how frequently any particular diviner or institution was wrong.
For a modern prediction system, the parallel is a set of questions rather than a claim of equivalence:
- What counts as evidence? Identify which observations or data enter the system and which are left out.
- Who sets the rules? Ask how categories and interpretations were chosen, and who can explain them.
- Who benefits from the output? Check whether a recommendation could reinforce an institution’s or decision-maker’s existing interests.
- How can an error be recognized? Look for a way to compare predictions with outcomes rather than accepting a persuasive interpretation after the fact.
- What action follows? Treat a prediction as input to judgment, not as proof that an outcome is inevitable.
These questions apply whether a prediction comes from an ancient specialist or a modern technical institution. The historical lesson is not that people have always used the same tools; it is that uncertainty creates demand for interpreters, and interpretation can carry authority as well as information.
Free tools Windows power users keep installed
One-click scans. No signup required.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
