Ancient Worlds & Lost Knowledge

Myths, Symbols & Mystery Traditions

Astrology: When the Sky Became a Story

Question & DebateDISPUTED

Can Astrology Predict the Future? What Scientific Tests Actually Show

Review the Carlson double-blind experiment, a 15,000-person birth-date study, published criticism and what real evidence for astrology would require.

DarkBrain Knowledge Published

Can Astrology Predict the Future? What Scientific Tests Actually Show

What would count as a prediction?

A friend tells you that an astrologer once warned her about a major life change. Months later she left her job, moved to another city and began a new relationship. To her, the chart had seen the future. To a skeptic, the prediction may have been broad enough to fit many outcomes. Both people are reacting to the same story, but they are asking different questions.

The real research question is more demanding: Could an astrologer produce a clear prediction, before an event occurs, that succeeds more often than a relevant comparison method? And could independent investigators obtain a similar result when they repeat the test?

This is where the history, symbolism and intimate appeal of astrology meet scientific standards. The question is not whether people have long believed the heavens carry meaning. They have. It is whether the celestial positions used in particular astrological methods can identify personality traits or forecast events with measurable reliability. [S004; S017]

Which astrology would we be testing?

The word astrology describes a family of traditions. A newspaper Sun-sign column offers broad messages to millions of readers. A natal astrologer reads a chart constructed from birth data. Other practitioners consider planetary transits, relationship charts, electional timing or different tropical and sidereal conventions. Some use astrology symbolically, with no claim that a chart delivers independently verifiable facts.

One test cannot cover every one of these activities. A critique of mass-market daily predictions is not automatically a test of individualized natal readings. A personality-matching experiment does not directly measure whether a practitioner can name future events. The correct experimental design depends on the claim actually being made.

That said, a broad claim cannot evade testing forever by retreating into a new interpretation each time a prediction fails. A genuinely empirical claim must specify a relationship that can be observed. Whether a practice uses the tropical zodiac, the sidereal zodiac or a particular house system is part of its protocol and should be fixed before the result is known. [S009; S010; S011]

The Carlson experiment and the importance of blinding

In 1985, physicist Shawn Carlson published A Double-Blind Test of Astrology in Nature. The investigation used methods intended to address objections that earlier tests had misunderstood professional astrology. Participating astrologers helped develop the study design. Two procedures were used to examine whether natal-chart interpretations and personality data could be matched correctly. [S017]

A double-blind arrangement aims to prevent both the person providing a judgment and the person administering or evaluating it from using identity clues that would reveal the right answer. Such protection matters whenever ordinary conversational hints could be mistaken for a method's special accuracy.

The study's results were interpreted as failing to support the proposed astrological matching ability: performance was consistent with chance expectations under its analysis. But the scientific point is not simply that a famous journal published an unfavorable result. It is that a test tried to define what success meant before examining the answers. [S017]

The experiment is often quoted as though it ended the entire debate. That framing is too strong. One study can be informative and important while still limited by how people were recruited, what traits were measured, how outputs were evaluated and whether the tested procedure represented the practitioner's core claims.

The published criticism of a celebrated result

A later article by Suitbert Ertel reexamined Carlson's statistical analysis and criticized the treatment of different parts of the data. Ertel reported that alternative analyses produced one marginal result and another conventionally statistically significant result, while also stating that these were insufficient to establish astrology as empirically verified. [S019]

This deserves to be included rather than hidden. A credible critical account presents the original study and its published objections, even if the two reach different conclusions about the strength of a particular finding. The existence of an alternative statistical treatment is a reason to examine the methods, not a license to announce that the planets have been proved to govern character.

Statistical significance alone is not the whole story. Small effects, multiple comparisons, dependent measurements and post-hoc decisions can create apparently favorable results. Likewise, weak design or inappropriate scoring can hide a real effect. The appropriate response is to specify the competing analyses in advance and repeat the test with enough statistical power. A later critic's result and the original author's interpretation remain contributions to a methodological dispute, not a substitute for independent replication. [S017; S019]

The wider question is whether a transparent method can consistently extract useful information from natal data in circumstances where ordinary clues and retrospective interpretation are unavailable. No one experiment, including a reanalysis, should be made to carry more weight than its protocol allows.

What a much larger birth-date study found

In 2006, Peter Hartmann, Martin Reuter and Helmuth Nyborg examined relationships between dates of birth, personality and general intelligence. Their analysis used two large samples, one with more than 4,000 middle-aged men and another with more than 11,000 young adults, giving a total exceeding 15,000 people. They also tested popular Sun-sign associations and related classifications. [S018]

The authors reported no relevant relationships between birth date and the measured dimensions of personality or general intelligence and no support for associations with the tested astrological Sun-sign groupings. Unlike a personalized reading in a consultation room, this approach looked for population-level patterns that should be detectable if certain broad birth-date claims were reliable. [S018]

The result is relevant but bounded. The samples and measures do not represent every population or every claim astrologers might formulate. A result about personality and intelligence should not be described as a direct experimental failure of every prediction about marriage or career. Even so, a large null result is hard to reconcile with strong versions of the claim that ordinary zodiac signs reliably encode stable personality differences.

A further complication is season of birth. Seasonal timing can be associated with exposure patterns, school-age cutoffs, environmental conditions or other real-world processes in some research fields. Finding a seasonal effect would not by itself establish astrological causation. Researchers have to distinguish a calendar-based correlation from the proposed influence of planets and zodiac signs. [S018]

Why a good test needs a prediction recorded in advance

Imagine a proposal: astrologers believe that a certain configuration predicts an unusually high probability of changing careers during the next year. A fair test would define the configuration, the population, what counts as a career change, how to handle people already planning to leave, the expected difference from chance and the acceptable error rate. The investigators would document the prediction before collecting outcomes.

A poor version of that test would instead gather a set of life histories, highlight those who changed careers and then search their charts for a pattern that could explain it. Because a complex chart contains many components, the second approach offers countless opportunities to find a matching pattern after an event. This may be interesting for generating hypotheses, but it does not establish a successful forecast.

A strong design would use preregistration, blinded outcome assessment where practical, a suitable comparison group and a replication by a team not involved in designing the original prediction. Researchers would also account for the large number of possible astrological configurations; testing many associations increases the risk of a chance finding.

None of these principles requires a scientist to be hostile to astrology. They are exactly what a supporter would want if the goal were to demonstrate that a method contains reliable information not available by simpler means.

What if astrology works as metaphor?

Some practitioners reject the idea that astrology should be judged only as prediction. They describe a birth chart as a symbolic language, an art of correspondence or a structured conversation about life themes. A client may leave such a conversation with a useful insight. The experience can be real and meaningful without being evidence that a planet caused the insight.

But this philosophical position also sets limits. A symbolic reading does not gain authority to make medical diagnoses, pressure relationship decisions or promise financial success. If a practitioner says the system offers reflection, that is one claim. If the same practitioner promises that the stars reveal objective facts about another person's secret intentions, the claim has crossed into territory where evidence is required.

For the reader, it helps to ask what is actually being offered. Is it a story to think with? An emotional ritual? A historical tradition? A factual prediction? A consultation can move between these modes without warning. Keeping them distinct protects both intellectual honesty and personal autonomy.

From our desire for certainty to a better standard

Astrology has survived for centuries because it offers more than an isolated forecast. It connects birth, season, character and the enormous sky into a single story. It gives language to uncertainty and can make a difficult period feel patterned rather than random. The cultural and psychological appeal is not trivial. Neither is it reliable evidence of a cosmic mechanism. [S006; S012]

Current well-known empirical tests have not established the central predictive claims discussed here. The historical record, the technical intricacy of birth charts and the subjective feeling of accuracy do not replace controlled validation. Published criticism of a landmark study is worth examining, but it does not by itself verify astrology. [S017; S018; S019]

There is room for curiosity without accepting an unsupported conclusion. A serious claim deserves a careful test. An ancient story deserves accurate historical treatment. A person's reflective experience deserves respectful listening. Those obligations can coexist.

The stars do not need to be false in order for a horoscope to be unreliable. They are real astronomical objects. Our task is to identify which parts of the human story we have written onto them, and what observations, if any, show that the story knows more about tomorrow than we do today.