NOTE: I submitted the following as a term paper for the course “Skeptical Approach to Parapsychology” at the Rhine Education Center.
The assignment was to take a noteworthy parapsychological study and evaluate it by apply critical thinking–being neither unduly credulous nor unduly dismissive of its claims.
(Also, since the assignment was to approach the task from a scientific, parapsychological perspective rather than a religious one, I don’t simply provide a theological analysis of what the soul is, and I consider options a non-religious researcher would need to.)
The paper received an “A.”
Can the Soul Be Weighed?
by Jimmy Akin
A minor pop culture trope holds the human soul weighs about as much as a piece of bread, or 21 grams. This trope appears various places, including the title of the 2003 Sean Penn movie 21 Grams.
The trope’s basis is a set of experiments begun in 1901 by Duncan MacDougall, M.D. His results were published in 1907 in American Medicine and the Journal of the American Society for Psychical Research (vol. 1, no. 5).
MacDougall weighed humans and dogs at the moment of death and—in the case of humans—found a measurable loss of weight coincident with death. After accounting for known, natural substances the subjects’ bodies could have released, MacDougall conjectured the loss of weight may have been due to the departing human soul.
Rationale for Experiment
MacDougall explains the basis for his experiment by stating that, if the personality survives death, it must exist as a “space occupying body.” He writes:
It is unthinkable that personality and consciousness continuing personal identity should exist, and have being, and yet not occupy space. It is impossible to represent in thought that which is not space occupying, as having personality, for that would be equivalent to thinking that nothing had become or was something, that emptiness had personality, that space itself was more than space, all of which are contradictions and absurd.
He reasons that whatever substance this personality-bearing, “space occupying body” (hereafter “soul,” for convenience) may have a measurable weight. He writes:
According to the latest conception of science, substance or space occupying material is divisible into that which is gravitative—solids, liquids, gasses, all having weight—and the ether which is non-gravitative.
MacDougall considers whether the soul might be made of normal “gravitative” matter, although he also considers two alternatives.
The first is that the soul might be made of luminiferous ether—a substance formerly believed to fill the universe and be responsible for propagating light waves through space. MacDougall thinks this option impossible, since ether was believed to be continuous throughout the universe, whereas individuals’ personalities are separate and distinct.
The second alternative is that the soul may be made of “a middle form of substance neither gravitative matter nor ether, not capable of being weighed, and yet not identical with ether.” Such a “middle form” might be non-continuous, allowing separate personalities/souls, but still not being weighable. However, MacDougall thinks it more reasonable to suppose that the soul “must be some form of gravitative matter” since it is linked organically with the body until death.
He thus proposes weighing dying individuals.
Examining the Rationale
MacDougall’s rationale is clever and worth examining in light of the history of philosophy and subsequent scientific developments.
Although he says it is “unthinkable” that the soul is not a space-filling body, many prior thinkers disagreed. In the Middle Ages, it was a commonplace for philosophers to regard spirits—including God, angels, and human souls—as entities that lacked extension in space. These spirits could be said to be “in a place” in an accommodated sense. When a spirit manifested its influence on something in the material world, the spirit could be said to be “in” that location (cf. Summa Theologiae I:52:1).
In the Early Modern period, there was renewed discussion of this subject, with Renee Descartes taking the position that the soul is non-extended and Henry More arguing that spirits must be extended. (An issue that arose as a result of this discussion was how a non-extended, immaterial entity could control a body since the two could not have physical contact. Parapsychologically, this would be “explained” in terms of psychokinesis [PK], though the basis or bases of PK remain very unclear.)
Since many thinkers consider the idea of a non-extended soul conceivable, we will include this possibility when considering MacDougall’s results.
From a scientific perspective, MacDougall’s discussion of ether has been superseded. Evidence against the existence of ether had been discovered in the famous, 1887 Michelson-Morley experiment, and the idea ceased to be commonly used in physics during the twentieth century.
However, something like MacDougall’s “middle form” of matter/energy emerged in twentieth century science—that is, things other than ether that lack mass. Current science holds that there are massless particles, such as the photon and gluon. However, these are force-carrying particles and are not thought to form structures that would be capable of sustaining a personality independent of massive particles.
A possibility MacDougall didn’t consider was that the soul might be made of a gravitative substance different than the solid, liquid, or gaseous states known in his day. While subsequent science has proposed additional states of matter, such as Bose-Einstein condensates, none of these are good candidates for the soul. (E.g., Bose-Einstein condensates can exist only close to absolute zero, and it would seem impossible for such a substance to coexist with a warm, living human body.)
In light of parapsychological research suggesting that ghosts are not electromagnetic phenomena, an interesting thought that could not have occurred to MacDougall would be the idea that souls might be made of “dark matter”—a hypothetical form of matter that does not interact with the electromagnetic force but that does interact gravitationally. The loss of such a soul could be weighable, and yet the soul would not show up on EMF detectors. (It should be immediately pointed out that current dark matter theories do not predict the existence of soul-like objects, but neither do they completely rule them out. A dark matter soul would need to interact with its body through a form of PK rather than EM.)
MacDougall’s Experiments
The design MacDougall used for his experiments was carefully thought out.
For human subjects, he arranged a large platform scale on which a bed could be set, along with a dying patient, and then balanced it. The scale was sensitive to two tenths of an ounce (5.7 grams).
With patient consent, MacDougall chose subjects dying of conditions expected to result in a peaceful passing so as not to jar the scale with death throes. (All died of tuberculosis—“consumption”—except for one in a diabetic coma.)
He tracked the subjects’ weight in the hours preceding death to account for the natural loss of moisture that occurs through perspiration and respiration when the body is not being hydrated.
He then recorded any sudden change the scale registered at the time of death—to the extent this could be determined in his day. (Since electrocardiogram [ECG] monitoring was still being pioneered, this involved observing signs such as cessation of eye and muscle movement, breathing, and heartbeat—as determined by stethoscope.)
After death, MacDougall checked if the subject’s bowels had moved and whether—and how much—urine had been discharged.
For his canine subjects, MacDougall was unable to find dogs dying in peaceful ways. He thus used healthy dogs, sedated them to keep them still, and euthanized them—while monitoring their weight on scales that were sensitive to 1/16th of an ounce (1.8 grams).
MacDougall’s Results
Six trials were done with human subjects, with the following results (all numbers converted to metric):
Subject |
Measurement at Death |
Second, Later Measurement |
| 1 |
-21 g |
|
| 2 |
-14 g |
-46 g |
| 3 |
-14 g |
-43 g |
| 4* |
-11 to -14 g |
|
| 5 |
-11 g |
|
| 6* |
-43 g |
|
The measurements at death represent sudden drops that occurred within the space of “a few seconds.”
In two cases, a second measurement was taken shortly after death:
-
- Subject 2’s initial measurement was a sudden loss coincident with the last movement of the facial muscles, and the second reading was taken after cessation of heartbeat was verified.
- Subject 3’s additional reading was taken “a few minutes” after death.
MacDougall eliminated the results of Subjects 4 and 6 (marked by asterisks) from consideration:
-
- With Subject 4, MacDougall reports that “unfortunately our scales were not finely adjusted and there was a good deal of interference by people opposed to our work”—apparently hospital employees who regarded the experiment as too morbid. However, “at death the beam sunk so that it required from three-eighths to one-half ounce to bring it back to the point preceding death.”
- With Subject 6, although MacDougall recorded the measurement at death, he rejected it since “the patient died almost within five minutes after being placed upon the bed and died while I was adjusting the beam.”
Fifteen trials were conducted with canine subjects. MacDougall reports:
The same experiments were carried out on fifteen dogs, surrounded by every precaution to obtain accuracy and the results were uniformly negative, no loss of weight at death.
Eliminating Naturalistic Explanations
MacDougall sought to account for conventional material substances released by the body—moisture in the form of respiration and perspiration, as well as evaporation from urine and feces.
He tracked a slow, steady loss of weight before death due to moisture loss through respiration and perspiration, so this could not be responsible for the sudden drops coincident with the moment of death.
His subjects did not suddenly expel 11-21 grams of moisture with their last breaths. Neither did they suddenly release this amount of perspiration, which would have remained in contact with their bodies and the bedclothes and only evaporate slowly, meaning it still would have been weighed by the scale.
MacDougall did not report the subjects experiencing bowel movements upon death, though if they had, the feces “would still have remained upon the bed except for a slow loss by the evaporation of moisture depending of course, upon the fluidity of the feces.”
He reported some subjects releasing urine upon death (due to the relaxation of the urinary sphincter), however, “the urine remained upon the bed and could not have evaporated enough through the thick bed clothing to have influenced the result.”
Having eliminated semi-solid and liquid substances released by the body, MacDougall sought to account for gas that could be suddenly released at death—i.e., air in the lungs.
Physics indicates this should not matter. At ground level, the Earth’s atmosphere is pressing downward on objects, including the scale, and it does not matter whether the air is in the subject’s lungs or above the chest. The scale should not be materially affected, which is what MacDougall found:
Getting upon the bed myself, my colleague put the beam at actual balance. Inspiration and expiration of air as forcibly as possible by me had no effect upon the beam. My colleague got upon the bed, and I placed the beam at balance. Forcible inspiration and expiration of air on his part had no effect.
Alternative Naturalistic Explanations
Alternative explanations for MacDougall’s results have been proposed. A selection is considered and critiqued by Masayoshi Ishida in the Journal of Scientific Exploration (vol. 24, no. 1), though what follows here are principally my own thoughts.
Since the human body begins to cool at death, could the loss of heat be responsible for the observed loss in weight—either directly or due to a change in air currents (as proposed by Len Fisher)?
Neither would be plausible. Heat is produced by the small-scale motion of atoms, and the fact these vibrate less after death does not change their weight. Only a large-scale removal of atoms from the bed would produce the observed readings.
Similarly, while convection currents caused by the heat of a living body might lightly press down on the bed—if such currents existed in these cases—they would not dissipate at the moment of death. The coldness of death—known as algor mortis—takes hours to occur and is frequently used to determine time of death in criminal investigations. There would be no sudden loss of weight.
What about heartbeat or breathing? These produce vibrations that could affect a scale, and they cease suddenly at death. However, they would cause a living, prone patient to slightly oscillate up and down on the bed, and if the scale were visibly at balance when the patient was alive then it should remain even more steadily (and likely sub-perceptually) at balance upon death. There would not be a sudden drop of 11-21 grams.
It could be proposed that there was something wrong with MacDougall’s scales, that the measurements he took were botched, or that he committed fraud. However, there does not appear to be evidence supporting these hypotheses.
Paranormal Speculations
Lacking a good naturalistic explanation for MacDougall’s results, it is reasonable to consider paranormal explanations. These can only be speculative due to the limited data his experiment returned. Replication and new types of experiments would be needed to test individual hypotheses.
The first possibility is MacDougall’s own conjecture—that the loss of weight may be due to the departure of the soul, conceived of as a space-filling entity capable of being weighed.
If so, the soul might be a very fine structure made of conventional matter/energy recognized by the Standard Model of particle physics. Alternately, it might be made of an undiscovered form of matter that interacts gravitationally.
Questions that might be asked are what would account for the variance in numbers MacDougall saw upon death, what was responsible for the additional weight loss in the second readings, and why there was no weight loss observed with dogs.
All the readings were within a factor of ~4 (11-46g), and the readings at the moment of death were within a factor of 2 (11-21g). Given the small sample size (4-6, depending on which are counted) and the sensitivity threshold of the scale (5.7 grams), these differences might simply be due to normal variation in taking measurements.
However, it also is possible that—just as some humans have heavier bodies—some humans have heavier souls.
If further experiments showed that the second, greater readings taken in two cases represent a real, second post-mortem weight loss, it might be proposed that there is more than one paranormal “thing” that detaches at death.
This idea may correspond to certain religious conceptions. In ancient Egypt, the human was thought to consist not only of the physical body but also several soul-like entities referred to as the ba, the ka, the shut, etc. Similarly, some Christians have understood humans as being tripartite, consisting of body, soul, and spirit. Even body/spirit dualists like John Duns Scotus have held that humans have multiple intangible “substantial forms” when alive.
Such claims, in light of MacDougall’s second readings, should alert us to the possibility that the death process may involve more than the departure of a single soul-like entity.
When it comes to dogs, MacDougall’s results would be equally consistent with the hypotheses that dogs do not have souls that survive death or that their souls produce results below the sensitivity threshold of the scale used (1.8g).
Attention should be paid to how MacDougall’s results might be explained if the soul is not spatially extended, as various philosophers have proposed. Why would the departure of such an entity result in an observed loss of weight?
It seems difficult to imagine a non-extended entity having intrinsic mass, but the soul could still interact with weighable matter. This would seem to be a form of PK, and two possibilities for the loss of weight spring to mind.
First, the soul would seem to have a tight psychokinetic association with the body during life, as illustrated by the ease of producing voluntary motions (e.g., lifting an arm) and the difficulty in psychokinetically moving objects outside the body. This tight association might not instantaneously vanish upon death. The soul might retain a PK “grip” on particles or atoms within the body, and as the soul detaches during the death process, enough of these might be pulled along with it to explain the loss of weight.
Second, there may be an explanation in line with the super-psi hypothesis that psychic functioning is part of people’s activity in their everyday environments. People use their bodies to steady themselves as they navigate their surroundings, resulting in them shifting their weight as they move body parts. They might use PK to assist this process. They might even continuously, psychokinetically cause their bodies to slightly sink down as part of steadying themselves in their environment, and if this PK ceased upon the departure of the soul, it could result in the observed loss of a number of grams.
In both this and the previous case, MacDougall’s variant readings might be explained by differences in the strength of the subjects’ PK. Depending on how the death process works from the soul’s perspective, it might also explain the larger, apparently postmortem readings he obtained in two cases—as the soul detached or the PK ceased functioning in stages.
Conclusion
MacDougall’s 1907 paper remains intriguing, and a good naturalistic explanation for his results has not been found.
Unfortunately, the small sample size he was able to achieve greatly limits the paper’s evidential value. MacDougall wanted to perform many more experiments with human subjects, but opposition to the project made this impossible.
Thus far, it appears no one has attempted to replicate his experiment with dying humans. However, there have been attempts to do so with animals. In 1907 the Los Angeles Herald reported on an animal replication effort by H. La Verne Twining, which produced mixed results.
Unfortunately, until human replications are attempted with substantially larger sample sizes—as well as modern measurement and control methods—MacDougall’s paper remains only a fascinating, suggestive study.
