In the summer of 1681, the English astrologer William Lilly fell seriously ill. He was seventy-nine years old, and the illness was enough to alarm his friends and household. But Lilly himself, when informed of the severity of his condition, is reported to have been curiously calm.
He had already seen it coming.
Lilly had been the most celebrated astrologer in England for forty years. He had predicted the Great Fire of London in 1666 — in a pamphlet published fifteen years earlier, he had included an illustration of what appears to be a city in flames beneath a configuration of heavenly bodies that matched the configuration in the sky on the night of September 2, 1666, the night the fire began. He had been brought before a parliamentary committee to explain this. He had predicted the death of King Charles I with enough accuracy to be arrested after the king’s execution and questioned about foreknowledge of the plot. He had survived these investigations partly through political skill and partly because his defenders were more powerful than his accusers.
He had, over the course of his career, calculated the deaths of hundreds of clients. He was good at it, by the standards of the practice. He understood the indicators — the Saturn configurations, the dangerous eclipse positions, the transits of malefic planets over the Ascendant — that practitioners associated with mortal crises.
In 1681, he turned those same indicators on himself.
What he saw in his own chart, he did not record publicly in the way he recorded other predictions. What he appears to have concluded is that his time had arrived. He died in June of 1681, shortly after falling ill, at the age of seventy-nine.
Whether Lilly’s death was predicted by his chart, or whether Lilly — an expert in reading the signs of terminal illness and a man of considerable intelligence — simply recognized that his body was giving out and chose to frame that recognition in astrological terms, is exactly the kind of question that makes these stories interesting.
The Mechanics of Death Prediction in Astrology
Before examining specific cases, it is worth understanding what practitioners were actually doing when they “predicted death” — because the practice is considerably more sophisticated than the popular image of an astrologer staring at a chart and announcing “you will die on Tuesday.”
Classical Western astrology, particularly the tradition flowing from the Hellenistic period through the Arabic transmission into medieval and Renaissance Europe, had developed specific techniques for identifying what practitioners called “aphetic places” — the sensitive points in a natal chart that governed vitality and longevity. These techniques included:
The Hyleg and Alcocoden: The Hyleg (from the Persian haylaj, “governor”) is the planet or point in the natal chart that represents the seat of vital force. The Alcocoden is the “dispositor” of the Hyleg — the planet that governs the Hyleg’s sign. The Alcocoden’s quality, strength, and condition were used to estimate the native’s overall vitality and expected lifespan.
Primary Directions: A sophisticated technique for advancing the natal chart forward in time. The Ascendant (or the Hyleg) is “directed” through the zodiac at a specific rate — roughly one degree per year — and when the directed point encounters malefic planets or their aspects, the period is identified as dangerous.
Solar Returns and Profections: Annual charts calculated for the exact moment the sun returns to its natal position (the Solar Return) and a system of annual “profections” advancing the houses by one sign per year. Dangerous combinations of Solar Return and natal factors identified vulnerable years.
These techniques are internally consistent and technically demanding — they require significant astronomical calculation and interpretive skill. The practitioners who used them were not guessing. They were applying a systematic framework with specific rules, derived from centuries of accumulated practice.
The question is whether the framework tracked something real, or whether the hits were remembered and the misses forgotten.
Johannes Stoeffler and the Great Flood
The most dramatic failure of death prediction in the historical record also involves a prediction of mass death — and it illuminates the psychology of the practice as much as any successful prediction.
Johannes Stoeffler was a German mathematician, astronomer, and astrologer who lived from 1452 to 1531. He was a distinguished figure — professor at the University of Tübingen, maker of fine astronomical instruments, author of texts that were widely consulted across Europe. He was not a fool or a charlatan.
In 1499, Stoeffler published a prediction that a catastrophic flood would occur in February 1524, when multiple planets would converge in the sign of Pisces (the fish, associated with water). The prediction circulated widely across Europe over the following twenty-five years. By 1524, it was one of the most widely known astrological predictions in history. Hundreds of pamphlets had discussed it. Count von Iggleheim of the Rhine had built a three-story ark and stocked it with provisions. Crowds gathered along riverbanks in Germany on the predicted date.
February 1524 brought light rain and unremarkable weather across most of Europe. The flood did not come. The crowds dispersed, some furiously. Count von Iggleheim was reportedly stoned by the mob that had gathered to watch the flood — they were angry at having been made to look foolish, and he was a convenient target for that anger.
Stoeffler continued his career undamaged. This is the interesting part. The failure of his most famous prediction did not destroy his reputation. He continued to be consulted, continued to publish, continued to be taken seriously until his death in 1531.
There is a lesson here about the sociology of prediction: failed predictions damage the prophet less than they should. The emotional investment in the prediction, combined with the tendency to find explanations for failures rather than updating beliefs, protects the reputation of predictors across a wide range of failure rates. This is the confirmation bias literature applied to historical institutions rather than individual psychology.
Stoeffler also, incidentally, predicted his own death accurately — or nearly so. He calculated that the configuration in the sky around his sixty-ninth year would be extremely dangerous for him. He died at the age of seventy-nine, which is ten years later than he predicted. He attributed this discrepancy, when alive to comment on it, to protective factors in his chart that had mitigated the danger.
Jerome Cardan’s Strange Prediction
Girolamo Cardano — known in English as Jerome Cardan — was a sixteenth-century Italian polymath whose contributions to mathematics, medicine, and philosophy were significant. He was also an avid astrologer who cast horoscopes for many of the most significant figures of his time, including Edward VI of England and various Italian princes.
Cardan cast his own horoscope and determined from it that he would die at the age of seventy-five, in 1576. The sources that record this story — and they are not wholly reliable sources, writing decades or centuries after the events — claim that when Cardan reached 1576 in good health with no obvious terminal illness, he made a decision that has puzzled historians ever since. He simply stopped eating.
He died in September 1576, at the age of seventy-four, from — in the most common reconstruction — voluntary starvation.
The story is almost certainly embellished. The sources are not contemporary and have obvious incentives to construct a dramatic narrative. But some version of the story was circulating within a generation of Cardan’s death, and it captures something genuine about the relationship between prediction and action in the divinatory tradition.
If Cardan really did stop eating in 1576 to ensure his prediction came true, he was not demonstrating astrology’s power to predict death. He was demonstrating something stranger: that a sufficiently committed believer in a prediction may take actions that produce the predicted outcome. The self-fulfilling prophecy operating at its most extreme — not “I expect X and therefore I perceive X,” but “I predict X and therefore I arrange X.”
This is not the comforting narrative that validates divinatory practice. It is a more disturbing one: the prediction that becomes self-fulfilling not because the cosmos arranged it, but because the person believed it strongly enough to arrange it themselves.
William Lilly and the Great Fire
The most well-documented case of an astrologer predicting his own death is, somewhat paradoxically, less historically interesting than the Great Fire of London prediction that preceded it — because the Great Fire prediction is the one that raises the genuinely difficult questions.
Lilly’s 1651 pamphlet Merlini Anglici Ephemeris contained a woodcut illustration showing a city in flames. The illustration depicted a configuration of heavenly bodies that matched the actual sky on September 2, 1666, the night the fire began in a bakery on Pudding Lane and spread to consume most of the City of London.
The fit between the woodcut and the actual event is not disputed. The questions are: when was the woodcut produced? Was it produced in 1651 as the pamphlet suggests, or was it inserted later into reprints? Could it have been produced from knowledge of London’s fire vulnerabilities — the city was predominantly timber construction, densely packed, with inadequate water supplies — rather than from astrological calculation? Could the astrological configuration have been calculated backward after the fire to match what had occurred?
The parliamentary committee that summoned Lilly in 1666 asked essentially these questions. He had reasonable answers to most of them. The pamphlet was published before the fire; the woodcut was in the original edition. He had identified the configuration as dangerous for London; London had experienced the predicted danger.
He was not accused of any wrongdoing. He was not found to have foreknowledge of the specific arson or accident that started the fire. He was released and his reputation, if anything, enhanced.
What the Great Fire case illustrates is the genuine epistemological difficulty of assessing astrological claims: even when a prediction appears accurate, the mechanisms by which the accuracy was produced — genuine foresight, skilled extrapolation from structural vulnerabilities, coincidence remembered and dissimilar cases forgotten, or retroactive adjustment of dates — are almost always difficult to definitively establish. The evidence is almost always ambiguous. Ambiguous evidence, as we have seen, tends to be interpreted in the direction of the observer’s prior beliefs.
The Pattern Across the Cases
Reading through the death predictions and their outcomes, a pattern emerges that is different from either “astrology predicts death accurately” or “astrology is fraud.”
The practitioners who made death predictions were doing something that drew on real skills: observational skill (identifying signs of deteriorating health), pattern recognition (correlating natal chart configurations with the timing of deaths they had observed in their practice), and actuarial reasoning (estimating longevity from the intersection of multiple factors). These are not magical skills. They are forms of practical intelligence that, applied consistently over decades, could produce above-chance accuracy.
The death predictions that turned out to be accurate often look, in retrospect, like sophisticated recognition of deteriorating conditions expressed in the vocabulary available to the practitioner. When Lilly recognized that his own illness in 1681 matched the configurations he had associated with fatal crises in hundreds of other cases, he may have been drawing on genuine pattern recognition rather than cosmic consultation.
The death predictions that turned out to be wrong — Stoeffler’s flood, Cardan’s apparent ten-year overestimate — are remembered for their failures but did not destroy the practitioners’ reputations, which tells us about the sociology of astrological credibility rather than about the astrology itself.
And the death predictions that may have been self-fulfilling — Cardan’s possible voluntary starvation — raise the most interesting question of all: what is the relationship between believing a prediction and enacting it?
This question has no clean answer. But it is worth sitting with, because it applies not only to predictions of death but to all forms of oracular guidance. The relationship between what the oracle says and what happens is mediated by the person who receives the reading and acts — or doesn’t act — in response. That mediation is part of the oracle’s effect, not external to it.
The astrologer who predicted his own death may have been right because he was perceptive. Or he may have been right because he believed he was right and arranged his circumstances accordingly. Or the prediction may have been so ambiguously stated that whatever happened would have counted as fulfillment.
Each of these possibilities is interesting. None of them is as simple as either the believers or the skeptics usually acknowledge.