Evidential Mediumship: What Counts as Proof
Evidential Mediumship: The Standard That Separates Comfort From Proof
Evidential mediumship is the part of mediumship that can be checked. A medium who says your grandmother loved you and is at peace has given you comfort, and comfort matters, but comfort is not evidence. A medium who says your grandmother kept a blue tin of buttons in the kitchen drawer, called you by a childhood nickname, and worried about the oak tree in the garden has given you something specific, and that specificity is what evidential mediumship is about. This page explains what counts as evidence, what does not, how the standard grew out of more than a century of investigation, and how to judge a reading you have already had, or one you are thinking of booking. Start with what is a medium on the mediumship hub if the basic terms are new to you.
The Evidence Ladder: From General to Specific
The simplest way to judge mediumship is to sort what was said into four levels. Level one is general and warm: I feel a loving presence, someone here loves you very much, your grandmother is proud of you. Almost anyone in grief would accept these statements, and almost any sitter would find them comforting. They are not evidence, because they could apply to almost anyone. Level two is general but personal in shape: I feel a mother figure, someone here had trouble breathing at the end, there was a difficult time in the family around money. These statements feel more specific, but they still apply to a large number of people, and a skilled reader can produce them by guessing from a sitter's age, appearance, and reactions.
Level three is specific and checkable: a first name, a nickname, a job, a place, a shared memory, a specific object. Your grandfather was a railway worker. Your aunt's name was Rose. You and your brother used to hide in the apple tree. The house had a green gate. These details can be right or wrong, and the sitter knows which. Level four is specific, checkable, and hard to know by normal means: a detail the sitter did not know at the time and had to check afterward, a detail about a person the sitter never met, or a detail that only one other living person could confirm. Level four is rare, and it is the level that researchers have always cared about most, because it is the hardest to explain by guessing, by reading the sitter, or by prior knowledge. A reading that stays at levels one and two is a comfort reading. A reading that reaches levels three and four, repeatedly, is an evidential reading.
The Case That Set the Standard: Leonora Piper
No case shaped the evidential standard more than that of Leonora Piper, an American trance medium who lived from 1857 to 1950. Piper was discovered almost by chance in the 1880s, after she went into a trance during a visit to a psychic healer. She was then studied for decades by the Society for Psychical Research, which had been founded in London in 1882 to investigate psychic phenomena carefully. Her case matters because of the controls the investigators used, not because of any single message. She was watched for fraud, followed by detectives, and introduced to sitters under false names. She was taken to England in 1889, where she knew no one, and studied under the supervision of a committee that included the physicist Oliver Lodge. The investigators gave her no chance to learn about the people she would meet before a sitting.
The most famous part of her case involves a young man named George Pelham, known as G.P., who died in a riding accident. Within weeks of his death, a personality claiming to be Pelham began to speak through Piper in trance. In one series of sittings, Pelham was presented with a group of 150 strangers and picked out 30 of his own personal friends from among them, without making a single mistake. He also gave details about their shared lives that Piper had no normal way of knowing. The Harvard psychologist William James, who had started out skeptical, studied Piper for years and concluded that fraud or ordinary guessing could not account for the full range of her results. Richard Hodgson, a lawyer and investigator for the Society who had built his reputation on exposing fraudulent mediums, studied her for two years, introduced more than 150 sitters to her under false names, and eventually wrote that her trance phenomena were genuine. Piper was under expert observation from 1886 until 1911, and during that whole period she was never found to make any conscious attempt to cheat.
Piper's case does not prove that mediumship is real in the way a laboratory experiment proves a chemical reaction. It shows what evidence looks like when it is collected carefully, over decades, with controls that block the obvious explanations. That is why it is still cited today, and why it sets the standard against which modern mediums are judged. A modern medium who gives level one messages and calls them evidence is not meeting the standard Piper's investigators set. A modern medium who gives level three and four details, under conditions where they could not have known them in advance, is working in the same tradition.
Gladys Osborne Leonard and the Proxy Sitting
A second case from the same tradition shows a different control, and it is worth knowing because it answers the most common objection to mediumship, which is that the medium reads the sitter. Gladys Osborne Leonard, who lived from 1882 to 1968, was one of the most thoroughly investigated mediums of the twentieth century. In her best known sittings, the person who sat with her was a proxy, someone who knew nothing about the dead person and nothing about the family, and who asked questions on behalf of a sitter who was not present. The medium could not read the proxy's reactions to learn about the dead person, because the proxy did not know the answers. The real sitter then checked the results afterward, without the medium present.
Proxy sittings are the gold standard for this reason. They remove the two most common normal explanations for a good reading: that the medium guessed from the sitter's appearance, age, clothing, and reactions, and that the medium learned the facts from the sitter's own answers during the reading. A medium who works well with proxies is doing something harder than a medium who works face to face with a sitter who is eager to help. When you choose a medium today, ask whether they have ever worked with a proxy, a blind sitting, or any condition where they could not see or question the sitter. A medium who welcomes such conditions is showing confidence in their work. A medium who refuses all controls is telling you something too.
What Counts as Evidence, in Detail
With that history in mind, here is what counts as evidence in a modern reading, and what does not. A first name counts, especially an unusual one, and it counts more if the medium gives it before the sitter says it. A nickname counts more than a first name, because nicknames are private and specific. A job counts, especially a distinctive one, like a lighthouse keeper or a piano tuner, and it counts more if the medium also gives a detail about the work. A cause of death counts, especially a specific one, like a fall from a ladder or a heart attack on a particular day, and it counts more if the sitter had not mentioned it. A shared memory counts, especially a small one, like a picnic in a specific place or a song sung on a specific occasion. A specific object counts, especially one with a story, like a watch that stopped, a ring that was lost and found, or a letter that was never sent.
What does not count is equally important. A statement that applies to most grieving people does not count. A statement that the sitter supplied first, and the medium then repeated, does not count. A statement that the medium changed after the sitter said no does not count; that is fishing, not evidence. A prediction about the future does not count as evidence about the dead, because it is about the living, and it cannot be checked until later, if ever. A message that is entirely about the sitter's feelings, however accurate, does not count as evidence about a dead person, because it is psychic material, not mediumistic material. The distinction between the two is explained in medium vs psychic, and it is the distinction that keeps a reading honest.
The Tricks an Honest Sitter Should Know
An honest medium wants an honest sitter, and an honest sitter knows the techniques that can make a weak reading look strong. The most common is cold reading, which is the skill of producing statements that sound specific but apply to many people, and then using the sitter's reactions to narrow them down. A reader who says I feel a father figure, watches your face, and then says he had trouble with his chest is not necessarily receiving anything. They are watching you, and you are helping them without knowing it. The defense is simple: give the medium nothing. Do not nod, do not say yes, do not supply names, do not correct mistakes out loud. Sit still, listen, and write down what was said. Judge it afterward, when you can check the details calmly.
A second technique is hot reading, which is learning facts about a sitter before the reading, from social media, from a booking form, or from a conversation in the waiting room. The defense is equally simple: book under a first name only if you can, do not fill in details on a booking form beyond what is needed, and do not talk about the person who died in the waiting room or on the phone beforehand. A third technique is the shotgun, where the medium gives a rapid run of names, jobs, and details and lets the sitter pick the ones that fit. The defense is to count the misses as well as the hits. A medium who gives twenty details and gets three right has not given you an evidential reading. They have given you a guessing game, and the three hits are the ones you will remember unless you write down all twenty.
How to Judge a Reading You Have Already Had
If you have already had a reading, judge it now, while the details are still fresh. Write down everything you remember, in two columns. In one column, put the specific details: names, nicknames, jobs, places, objects, memories. In the other column, put the general statements: love, pride, peace, watching over you, proud of what you are doing. Count the specific details. How many were right? How many were wrong? How many were so general that they could apply to anyone? Did the medium give the details before you said anything, or did you supply them first? Did the medium change a detail after you said no? Did the reading contain anything you did not know at the time and had to check afterward? That last question is the most important one. A detail you had to check, and that turned out to be right, is the strongest evidence a reading can produce.
If the reading was mostly general statements, that does not mean it was worthless. It may have brought you comfort, and comfort has value. But it was not an evidential reading, and it is worth knowing the difference, because the difference decides what you do next. If you want comfort, a general reading may be enough. If you want evidence, you need a medium who works at levels three and four, under conditions that block guessing, and you need to be prepared for the possibility that the evidence will not come. A medium who is honest will tell you that. A medium who promises evidence every time is promising something no one can guarantee.
If you are considering a sitting, how a medium reading works explains the practical side, how to choose a medium explains what to look for in a reader, and what to ask a medium helps you prepare questions that give the medium a fair chance to produce evidence without giving away the answers. If you are developing your own ability, how to develop mediumship explains the practice that leads, over years, to work that can be judged by this standard.
If you want to talk through a reading you have had, or prepare for one you are considering, a reader can help you think it through, and the first three minutes are free. First 3 minutes FREE — every time, for every reading, with all advisors.
For a simple next step, try a One-Card Tarot reading and reflect on one specific detail from your own experience that no stranger could have guessed.
Frequently Asked Questions
What is evidential mediumship?
Evidential mediumship is mediumship judged by specific, checkable details that identify a particular dead person, such as a name, a nickname, a job, a shared memory, or a specific object. General statements about love and peace are comfort, not evidence. A reading is evidential when it repeatedly reaches that specific level.
Who was Leonora Piper?
Leonora Piper, 1857 to 1950, was an American trance medium studied for decades by the Society for Psychical Research. She was watched for fraud, followed by detectives, and tested with sitters under false names, including a trip to England in 1889. In the George Pelham case she identified 30 of his friends from 150 strangers without a mistake, and investigators William James and Richard Hodgson concluded her results could not be explained by fraud or guessing.
What is a proxy sitting?
A proxy sitting is one where the person who sits with the medium knows nothing about the dead person and asks questions on behalf of a sitter who is not present. It blocks the medium from reading the sitter's reactions or learning facts from the sitter's answers. Gladys Osborne Leonard, 1882 to 1968, was tested this way, and it is still the strongest control in mediumship.
Can a medium prove survival after death?
No single reading can prove survival after death in the way a laboratory test proves a fact. What evidential mediumship can do is produce specific details that are hard to explain by normal means, repeatedly, under controlled conditions. That is strong evidence for the sitter, and it is the standard researchers have used since the 1880s, but it is not proof in the scientific sense.
How can I tell if a medium is fishing for information?
Watch for three signs: the medium asks leading questions and then presents your answers as messages, the medium changes a detail after you say no, or the medium gives a rapid run of names and details and lets you pick the ones that fit. A good medium gives specific details without asking, accepts a no without arguing, and lets the misses stand.
Last updated: November 11, 2026
