ArticlePDF Available

Can Death-Related Dreams Predict Future Deaths? Evidence from a Dream Journal Comprising Nearly 12,000 Dreams

Authors:

Abstract

Dreams that appear to predict future events that could not have been anticipated through any known inferential processes have been reported for centuries, and dreams that appear to anticipate the death of an acquaintance or loved one are particularly common. Such reports become more suggestive of genuine precognition if there are no natural cues (such as an illness) to an impending death and if the time interval between the dream and the subsequent death is brief. Most reports are diffi cult to evaluate because we dream many times each night but typically remember and report only a salient subset of our dreams. Thus we cannot assess whether the time interval between a death-related dream and the death of the dream character is brief or lengthy because we have no control set of non-death–related dreams to which its time interval can be compared. The study reported here provides just such a control set by comparing deathrelated and non-death–related dreams featuring the same set of dream characters who died after the dreams occurred. These were drawn from the author’s own dream journal in which he has recorded his nightly dreams for nearly twenty-fi ve years. The mean time interval between death-related dreams and the person’s subsequent death was signifi cantly shorter than the time interval between non-death-related dreams and his or her death, t(11) = 3.30, p = .004, one-tailed. Cases in which death-related dreams occurred after the characters had died are also considered. Seven of the cases are discussed in detail.
RESEARCH ARTICLE
Can Death-Related Dreams Predict Future Deaths?
Evidence from a Dream Journal Comprising Nearly 12,000
Dreams
ANDREW PAQUETTE
NHTV University of Applied Sciences, Breda, The Netherlands
Paqart@gmail.com
Submitted March 1, 2014; Accepted April 4, 2015; Published September 15, 2015
Abstract—Dreams that appear to predict future events that could not
have been anticipated through any known inferential processes have been
reported for centuries, and dreams that appear to anticipate the death of an
acquaintance or loved one are particularly common. Such reports become
more suggestive of genuine precognition if there are no natural cues (such
as an illness) to an impending death and if the time interval between the
dream and the subsequent death is brief. Most reports are di cult to
evaluate because we dream many times each night but typically remember
and report only a salient subset of our dreams. Thus we cannot assess
whether the time interval between a death-related dream and the death of
the dream character is brief or lengthy because we have no control set of
non-death–related dreams to which its time interval can be compared. The
study reported here provides just such a control set by comparing death-
related and non-death–related dreams featuring the same set of dream
characters who died after the dreams occurred. These were drawn from the
author’s own dream journal in which he has recorded his nightly dreams
for nearly twenty- ve years. The mean time interval between death-related
dreams and the person’s subsequent death was signi cantly shorter than the
time interval between non-death-related dreams and his or her death, t(11) =
3.30, p = .004, one-tailed. Cases in which death-related dreams occurred after
the characters had died are also considered. Seven of the cases are discussed
in detail.
Keywords: deathdreamspsiprecognition
Introduction
Precognitive dreams—dreams that appear to predict future events that
could not have been anticipated through any known inferential processes—
have been reported cross-culturally and historically for centuries (Myers
1920). Among such dreams, those that appear to anticipate the death of an
Journal of Scienti c Exploration, Vol. 29, No. 3, pp. 411–423, 2015 0892-3310/15
412 Andrew Paquette
acquaintance or loved one are particularly common (Stevenson 1970). Such
dreams provide stronger evidence for a psi-mediated or paranormal process
of precognition if there are no clues to the impending death—such as an
illness, threats of suicide, or participation in a life-threatening activity—and
if the time interval between the dream and the subsequent death is relatively
brief.
A major problem with most reports of such dreams, however, is that
people are most likely to recall and report dreams that are salient precisely
because subsequent events make them appear predictive and to forget or
fail to report dreams that have no subsequent related events. This is a real-
life variant of the “ le-drawer” problem in experimental work, in which
studies with positive results are more likely to be published than studies
with null results, thereby biasing the known database toward false positive
conclusions (Scargle 2000). Thus we cannot assess whether the time interval
between a death-related dream and the death of the character in the dream
is brief or lengthy because we have no control set of non-death–related
dreams to which its time interval can be compared. In the absence of such a
control set, we are left with the reductio ad absurdum that all death-related
dreams about living persons will eventually be predictive of their deaths:
All we have to do is wait long enough!
The study reported here provides just such a control set by retrospectively
analyzing nearly 12,000 nightly dream scenes from my own journal, in
which I have recorded my nightly dreams nearly continuously for twenty-
ve years (Paquette 2011). For this study, I rst used a word search to locate
all dreams containing death-related content and to identify the characters in
the dream. Next, I located all non-death–related dreams for that same set of
dream characters. Finally, I attempted to determine whether the character
in each dream was still living or had died. If the character had died, I noted
the date of death. For many of the characters, I had already recorded this
information in the journal itself, but in several cases I had to check other
sources to determine whether the dream character was living or deceased.
For this report, I computed for each dream character the mean time interval
between death-related dreams and the date of death with the mean time
interval between non-death–related dreams. A death-related dream is
considered to provide stronger evidence for precognitive anticipation of a
subsequent death to the extent that it occurs closer in time to the death than
do “control” dreams that are not death-related.
In the psi literature, there are also reports of post-death communications,
apparitions that occur after the featured character has already died but before
the percipient had been made aware of the death through normal non-psi
channels (Stevenson 1995). Some of these post-death communications take
Can Death-Related Dreams Predict Future Deaths? 413
the form of what are called “leave-taking” dreams, in which the spirit of
a deceased person purports to communicate to the dreamer from within a
dream (Barrett 1991/1992). Although I will discuss some of my post-death
dreams, most of them were discarded as evidence for a psi-related process
because I was aware of the character’s death at the time the dream occurred.
Method
The Dream Journal
The dream journal from which the dreams in this study are drawn was
begun on September 15, 1989. Ironically, I started it to demonstrate to my
wife that I was not having precognitive dreams but that memory errors
and faulty reasoning could easily produce the apparent correspondences
between dream content and subsequent events. Instead, several persuasive
correspondences accumulated (Paquette 2012).
The cutoff date for the study reported here was February 2, 2014. On
that date the journal contained reports of 11,779 individual dreams, an
average of 3.1 dreams per night. For purposes of this study, all dreams
from a single night are combined into a single dream record, yielding 3,732
dated dream records to be searched for death-related content. Each record
in the journal had been entered by hand immediately after waking, and
notes of events that appeared to con rm or discon rm points in the dream
were added and dated only after the journal entry itself was committed to
writing. These were recorded in the margin of the journals on the date of
the veri cation attempt or any followup attempts. Beginning on October 12,
2012, all dreams were recorded on an iPad and emailed to myself to create
an independent time-stamped record at a cloud storage network.
Although dream recordkeeping was fairly constant across the 25 years
covered by the journal, there is a block of 1,980 days, from February 9,
1993, to July 13, 1998, during which there were only 32 entries. This was
due to my life/work situation at the time and was not due to differences in
the number or types of dreams I was having during this period. These 32
entries were included in the search for death-related content.
The level of detail in each entry varies widely depending on dream
recall and waking circumstances. If I do not recall much, then the entry
is brief. On the opposite end of the spectrum, I may remember far more
than can be recorded in the time I have before I must go to work or deal
with other obligations. When that happens, I will write as much of what I
consider to be the most salient information before I run out of time. Some
entries run to more than 20 handwritten pages and took as much as three
hours to write. The average record word count is 201, with a maximum
414 Andrew Paquette
value of 3,018. Many entries are accompanied by illustrations, sometimes
multiple pages of them. These have not been counted in full but to date
1,019 of these pages have been scanned. In some records, the word count
is close to zero because illustrations arranged as a storyboard were used
instead of written descriptions.
Identifying Death-Related Dreams
The database of dream records was searched for death-related content using
the following 10 search terms: Dead, Death, Ghost, Spirit, Deceased, Dying,
Die, Murder, Kill, Suicide. Dreams containing any of the search words were
de ned as death-related even if the words were used symbolically and did
not appear to refer to an actual death. For example, one dream contained
the following sentence “I had twenty-four hours before a wrongly decided
death sentence was carried out.” Later that day, the person I identi ed in my
journal had his position terminated at my of ce (to my genuine surprise):
The ‘death’ in the dream was termination of employment, not an actual
death. On the other hand, dream reports were excluded if (a) they included
a keyword in an innocuous context, such as in the sentence “an extremely
small chameleon . . . has something like a death grip on my nger”; (b)
appeared in the context of a threat to cause harm “[Colonel] Klink is
convinced that Hogan is out to kill him . . . ”; or (c) the dream characters
involved were ctional—as in the example given in (b).
Of the 11,667 individual dreams recorded in the journals, 545 (4.7%)
were identi ed as containing death-related content. To be retained as a
death-related dream for the study itself, it also had to pass the following
tests:
1. The dream character must be identi able.
2. The dream either mentions the death of the character, the possibility
that the character is dead, or contains details characteristic of
deceased characters.
3. It was possible to verify whether the character is living or, if
deceased, the date of death.
4. In the case of post-death dreams, I must not have been aware of the
character’s death when the dream occurred.
After applying these criteria, 87 dreams featuring 50 unique identi able
dream characters remained for analysis. They include both living and
deceased dream characters, but this report concerns only the dream
characters who are deceased.
Can Death-Related Dreams Predict Future Deaths? 415
Results
Pre-Death Dreams
Of the 50 dream characters identi ed in the search, 12 of them had died after
I had experienced one or more death-related dreams about them. Table 1
displays for each character the mean number of days between death-related
dreams and the date of death; the mean number of days between control
(non-death–related) dreams and the date of death; and the percentage of
total days contributed by the death-dream intervals. For example, for Noah,
there was a mean interval of 165 days between my death-related dreams and
his death and a mean interval of 6,612 days between my non-death–related
dreams and his death. So the percentage of total days contributed by the
former is 165/(165 + 6612) = .024, or 2.4%.
The cases are listed in order of increasing percentages, with percentages
less than 50% indicating that the death-related dreams occurred, on average,
closer to the date of death than did control dreams. As shown in Table 1, this
was true for 9 of the 12 characters.
Because there were no control dreams featuring Hannah or Ken, two
data analyses were conducted. The rst analysis substituted the mean
number of control days for the remaining 9 characters as an estimate for
the 2 missing control-dream intervals—displayed in parentheses. With this
substitution, the death-dream percentages are signi cantly less than 50%
using a 1-sample, one-tailed t test across the 12 cases: t(11) = 3.30, p =
.004. In the second analysis, the records for Hannah and Ken were simply
omitted. When this is done, the death-dream percentage is still signi cantly
lower than 50%; t(9) = 2.45, p = .018.
Case Descriptions for Pre-Death Dreams
In this section I provide details of my dreams and the circumstances of the
subsequent deaths for the rst 3 characters listed in Table 1: Harmony, Noah,
and Hannah. These are the cases in which the relative intervals between the
dream and the subsequent death are the shortest.
Harmony. Harmony is a relative of mine whose case is one of the most
remarkable in the dream journal. As can be seen in the rst row of Table
1, I had a death-related dream about her on the same day that her death
occurred, July 2, 2003. There had been 24 earlier dreams about her, but this
was the only one that mentioned her possible death. When I awoke from the
dream, I was certain she had died, telling this to both my wife and daughter
the next morning. The day after I had entered the dream in the journal, I
received an email telling me that she had, indeed, died.
416 Andrew Paquette
The death-related dream about Harmony is a classic leave-taking dream,
a dream in which a spirit of the deceased appears as an announcement of his
or her death. In the earlier control dreams, Harmony appeared thin and frail,
but she looked very different in my nal dream. Here, in full, is my written
record of the dream from my journal:
[Harmony] is here brie y. She is literally shining, like sunlight through dia-
monds. She is very beautiful, radiant. Has she died?
TABLE 1
Pre-Death Dreams: Mean Time Intervals (Days)
between Dreams and Subsequent Death
Pseudonym aRelationship # death
dreams
# control
dreams
Mean # days
between
death
dreams and
death
Mean # days
between
control
dreams and
death
% days con-
tributed by
death-dream
to death
days b
1 Harmony Relative 1 24 0 2479 0.0
2 Noah Relative 4 14 165 6612 2.4
3 Hannah Friend 1 0 120 (4297) c 2.7
4 Ken Stranger 1 0 491 (4297) c 10.3
5 Isador Friend 2 17 2255 5973 27.4
6 Lim Relative 1 17 1389 3335 29.4
7 Chris Relative 1 9 2466 5037 32.9
8 Yi Relative 5 28 2614 4033 39.3
9 Jonathan Celebrity 2 6 3723 5515 40.3
110 Astrid Relative 1 1 2836 2831 50.0
111 Dominik Colleague 2 5 4695 3702 55.9
112 Mostafa Celebrity 3 6 5742 3456 62.4
Total 24 127 - - -
Mean
(SD)
-
-
-
-
2207.87
(1873.75)
4297.28
(1268.77)
0.559
(0.528)
a Names of dream characters are pseudonyms generated by the online Random Name Generator:
http://www.behindthename.com/random/
b If <50%, then death-related dreams occurred closer to the date of death than did control dreams.
c No control dreams for these 2 dream characters. Mean value from the other 9 cases used as an estimate.
Can Death-Related Dreams Predict Future Deaths? 417
I had not been not in contact with Harmony or other members of this
branch of my family for more than a year and I did not know Harmony
well. Based on emails I received several years before she died, I knew that
she had Alzheimer’s Disease and had been in a nursing home. But I did
not know at the time that the disease shortened one’s lifespan (Burns &
Iliffe 2009), so my knowledge of her illness did not serve as a cue that her
death was imminent—and certainly not on the exact day of her death. I
believe this dream provides good evidence for a psi-mediated process of
information retrieval.
Noah. Noah is the rst of my three stepfathers. My mother divorced
him in 1971, but remained in contact with him until at least 1980, when I
saw him with my mother in Santa Barbara. That was the last time I saw him.
On September 29, 2010, I dreamed of Noah. Here is the relevant portion
of that dream:
Mom, Karina, Tomoyo, and my  rst stepfather Noah. It feels like I’m in Los
Angeles, walking around with the above-named group. It is puzzling that
Noah is present because I haven’t seen or heard from him in so long. It is
nice to see him again regardless because I’ve always liked him. . . . After
what seems like several hours with Noah, he is gone. I tell my mom that I
wished I had said more while he was here, but she said not to worry, that
he’d just wanted to see me “One more time. Does this mean he’s died? Must
check with mom.
After this dream I asked my mother about Noah, but she didn’t know
where he was living or how to contact him. A little over a month later, on
October 19, 2010, I dreamed of Noah again:
. . . I had been with Peter and Karina, all of us talking about Peter’s printing
press, when Noah appears to say hello. . . . This is [the] second ‘Noah’ appear-
ance in these pages, making me wonder if he is alive or not. This dream may
not be interesting in any other respect, but two appearances in two weeks
(? Or so) is a lot from someone I haven’t seen in over 20 years and have no
reason to be thinking about.
On January 8, 2011, I had a lucid dream that Noah has given me a
wristwatch that has something to do with the future. On February 23, 2011,
I dreamed I was with Noah again in the context of a party on Memorial Day.
Checking on Noah’s status proved dif cult because we had both moved
multiple times and had not made any attempt to remain in communication.
Nor did Noah have a presence on the Internet. My mother was eventually
able to turn up an obituary listing Noah’s death date as May 19, 2011.
Dreams about Noah appeared in the journal 18 times prior to his death. As
418 Andrew Paquette
shown in Table 1, these four death-related dreams occurred much closer to
the date of his death than the control dreams.
The rst dream of September 29, 2010, sounds very much like a leave-
taking dream. Noah wants to see me “one last time,” as if he is about to
leave forever. The second dream of October 19 is only interesting because
of the rst, but the idea of his suddenly wanting to say hello, a theme that
does not appear until the “one last time” dream, is interesting and suggestive
of a second attempt. The dream of January 8, 2011, appears to be symbolic.
Noah gives me a wristwatch that is somehow important to the future, to
something that will happen soon. In the dream closest to the date of his
death, I dream of him in the context of Memorial Day, a day to remember
the dead. Collectively, the set of dreams can be interpreted as a last-goodbye
and a message to me that Noah will die soon. If this interpretation is correct,
it would relate to Noah’s actual circumstances as he prepared for death. In
that sense he is not saying good-bye because he has already died, but in
preparation for death.
These are the only four dreams of Noah to appear within eight years, no
dreams outside of this group suggest his death in any way, and two of the
dreams explicitly suggest his death while one implies it (Memorial Day)—
an interesting coincidence in itself because Noah died nine days before
Memorial Day.
Hannah. Hannah was the mother of my best boyhood friend, Donovan.
In 1976, my family moved from San Jose, California, to Las Vegas, Nevada,
the last time I would see him until decades later. Sometime after I moved, I
learned that his family had also moved. I made occasional attempts to nd
them over the next few years, but had given up by 1981, when my family
moved to Santa Barbara, California.
On 24 September 1999 I reported the following dream in my journal:
. . . next, I realize that Donovan is standing directly in front of me. I am quite
happy to see him again. Donovan doesn’t seem to know where he is or what
is going on. After realizing who he is, I grab him, pull him over, and give him
a big hug. I then give him a business card so that he can get in touch. After
that he faded out and disappeared.
A little over a week later, on October 3, 1999, I wrote the following:
[My Uncle Luke] tells me that Donovan [see dream a few nights ago] . . .
died in 1995 after getting sick. My impression was he died of asthma.1 Luke
also tells me that several others have had dreams of Donovan recently and
that each had a little piece of the story. I think we found out about his death
when we were thinking about him and his mom [Hannah] just appeared.
Can Death-Related Dreams Predict Future Deaths? 419
I did not write it in the journal, but that morning I told my wife that
when Hannah appeared, she said something that was different from what
Luke said, but I didn’t remember how it was different, just that she was
correcting Luke because he had made a mistake in what he told me. Four
months later, on February 20, 2000, I had the next dream in this series. In
it, I observe an older man and woman described in the journal as “devout
Christians.” The focus in the dream is on the wife, who stands near the man
as he sits in a chair in a darkened living room, looking bereft. I write “Is this
man a ghost?” in reference to the man. His wife talks to me in a friendly
way, while the man seems completely oblivious to our presence. “Her aura
is on plain view here, and it’s a good one,” I write. Although not written
in the journal, this dream character reminded me of Hannah and made me
think of Donovan and their house.
Because of these three dreams, I was concerned that Donovan may
have died. I made several searches on the Internet, looking for Donovan’s
obituary, but found nothing. A year later, I was re-reading the dreams when
I noticed the detail about Hannah appearing in the dream with Luke and that
she had disagreed with what he had said. I then noticed the February 20th
dream and realized that the mistake Luke made was identifying Donovan
rather than Hannah as the person who had died. Pursuing this hypothesis,
I conducted a new Internet search looking for Hannah’s obituary. I learned
that she had died of cancer at the age of 64 on January 31, 2000—only 120
days after my rst dream about her. I also had a dream about Hannah 20
days after her death, well before I had learned of it (See Table 2, below).
Post-Death Dreams
The dream journal contained 68 death-related dreams and 25 control dreams
that occurred after the death of the death character. When I eliminate all
dreams in which I knew of the character’s death at the time of the dream,
only 8 death-dream characters remain, each one appearing in a single death-
related dream. These are shown in Table 2 and include both Hannah and
Lim, who appeared also in Table 1. There were no control dreams for any of
these 8 dream characters, so it is not possible to conduct an analysis parallel
to the one for pre-death dreams. Here I describe the cases of Lim, Robert,
Wynona, and Isabel.
Lim. As in the case of Harmony, the case of Lim is persuasive as an
example of a psi-mediated dream because there is such a short interval
between my dream and her death; in this case, the dream occurs one day
after the death, but before I knew about it. On the morning of June 24, 2003,
I wrote the following about Lim in my journal:
420 Andrew Paquette
At Leontine’s, she complains that I never follow directions properly. As an
example she describes a recent wake where she asked people to sit by the
“lapidary bench” (what is that?).
A few hours later, Leontine called to pass on the information that Lim
had died the previous night. The purpose behind the call was to inquire
whether my wife and I would attend the wake, which she was organizing.
I didn’t know if it was relevant, but asked my wife if she knew of any
connection between something called a “lapidary bench” and either
Leontine or Lim. She told me that Leontine was an amateur gemmologist,
that a lapidary bench has something to do with making jewelery, and that
Leontine owned a lapidary bench.
Although Lim does not appear explicitly in this dream, the dream
appears to refer to her wake and thus is counted as a death-related dream
with her as the dream character. The only death-related dream about Lim
that appears in the journal occurred about 4 years earlier.
TABLE 2
Post-Death Dreams: Mean Time Intervals (Days)
between a Death and a Subsequent Dream
Pseudonym Relationship # Death dreams # of days between
death and death
dreams
1 Lim aRelative 1 1
2 Robert Acquain tance1 2
3 Rikard Stranger 1 4
4 Therese Acquaintance 1 18
5 Hannah aFriend 1 20
6 Isabel Stranger 1 26
7 Leo Relative 1 44
8 Wynona Acquaintance 1 340
Mean 56.9
a Lim and Hannah also appear in Table 1.
Can Death-Related Dreams Predict Future Deaths? 421
Robert. As with the case with Lim, my post-death dream about Robert
is notable for its temporal proximity to his death. On the morning of
February 2, 2014, I wrote the following in my journal:
Someone comes to tell me that “Bobby ” has died. I ask for more clari cation,
get the idea it is Robert W. Green.
Thirteen hours after emailing the dream journal entry to myself, I
received an email from someone named “Bobbie” to whom I had spoken
on perhaps three other occasions in my life. The email referred to a mutual
acquaintance named “Robert H.” or “Bob,” who had died the night before.
Like Bobbie, I did not know Bob well. The last occasion I had spoken
to him was about 2012, and before that, 2011 and 2010. In total, we had
conversed on between three and ve occasions, usually for about an hour or
two on each occasion. When I met him for the rst and only time in person
in 2011, he had just had a cancerous growth removed from his nose and he
appeared to be quite ill. As with Harmony, this information about his health
could hardly have been the basis for an accurate prediction of the date of his
death three years later. This is the rst and only mention of this character in
the journal. It is also interesting that the name “Bobby” or “Bobbie” in the
dream journal and my email history occurs only in reference to Robert H.
Wynona. The case of Wynona provides an interesting contrast to that
of Harmony, whose death occurred the same day as my dream. In Wynona’s
case, the time interval between her death and my subsequent dream is 340
days, the longest in Table 2.
Wynona was a friend of my mother’s in the late 1960s through about
1977. The last time I saw her was in 1977, when I was living in San Jose,
California. My family moved away and both my mother and I lost contact
with her. On December 9, 2010, more than thirty years after I had last seen
her, I dreamed of Wynona. Here is the relevant part of my dream report:
A little later, I hear that Wynona has a gift for me also. I see her go by to
see Karina. She says hello as she passes, and then is gone. I notice that a
woman has been drawing in Karina’s sketchbook . . . I retrieve it for Karina
and then go to a booth at a Wolfgang Puck’s restaurant to  nd her. She is in
a booth with Wynona, who looks better than I’ve ever known her to look.
Is she dead?
Checking on Wynona’s death proved to be dif cult. A Google search
yielded nothing, and my mother had no idea whether Wynona was alive
or dead. A year later when I started work on this study, I asked my mother
to make another attempt to determine Wynona’s status. She discovered
422 Andrew Paquette
that Wynona had died in 2010, the same year as my dream. The only other
dream in the journal that mentioned Wynona was a pre-death, control dream
from 19 years earlier. In other words, the only dream recorded in the journal
to mention Wynona’s death over a 19-year span occurred in the same year
as her death.
Unlike the dream of Harmony, the dream of Wynona does not depict her
at what appears to be the moment of death. Instead, it leaves the impression
of a spirit that is already dead, coming by to say hello and bestow a gift
of some kind. In this sense, the impression is accurate. She did not die on
the date of the dream, but was deceased at the time of the dream. It seems
unlikely that somehow I had learned about Wynona’s death through non-psi
channels but then suppressed that knowledge. We had never lived in the
same city again. She was not a well-known person and had no presence on
the Internet. The possibility that I could have somehow learned her status
normally and then suppressed the information seems quite remote.
Isabel. Isabel was a stranger to me at the time she died. The dream oc-
curred on August 12, 2003. Here is the relevant section of my dream report:
. . . I am talking with someone . . . when I am told that one of the employees
of this place may be about to die. I think it was Matthew from Arizona Floor-
ing.2 Someone in his family was very sick or had died, so he started taking
drugs, drinking, not taking care of himself. He was going to die, or had died
also. I am very worried about him here. There is a pervasive feeling of the
unnecessary here, that he doesn’t have to die at this time but he will if he
doesn’t shape up.
After waking from the dream, I wrote down the details and then
discussed it with my wife. The person I was talking to in the dream was
a youngish-looking woman, and she had urged me to contact Matthew to
get him to stop drinking and taking drugs. She said that if he didn’t, he
would die. My wife wanted me to follow through and tell Matthew about
the dream. I was less happy about the idea because I barely knew Matthew,
having spoken to him on only a handful of occasions when I bought supplies
at the store he worked at. My wife persuaded me, however, so I made the
forty- ve minute drive from North Phoenix to Tempe and entered the store.
After introducing the fact that I had dreamed of him, I asked Matthew
if a close relative had recently died and if he had started abusing alcohol
because of it. He af rmed that this was true. His sister-in-law, Isabel, had
died three weeks earlier, on July 27, 2003. A police cruiser collided with
her car from behind while she was stopped at a stoplight. I assumed that
this meant that the young woman in my dream was his sister-in-law and
told him as she had told me in a dream that he had to stop abusing alcohol
Can Death-Related Dreams Predict Future Deaths? 423
or he would die. Matthew admitted to being self-abusive and to drinking
heavily after Isabel’s death, and then told me that he had had a dream of
Isabel earlier that same week. In it, she gave him exactly the same message
I had received in my dream. This dream is the only one in the journal that
mentions Matthew and the only one that plausibly contains a character that
can be identi ed as Isabel. What is remarkable is that we both got “the same
message” from Isabel.
Conclusion
This article has presented 9 cases of dreams that appear to anticipate the
deaths of the dream characters through a paranormal or psi-mediated
process. A le-drawer control was provided by comparing the time intervals
between the death-related dreams and the subsequent deaths against a set of
control (non-death–related) dreams. Some post-death dreams also appeared
to provide evidence suggestive of psi-mediated information retrieval.
The 20 test cases presented were culled from a database containing
3,732 nightly dream records that were compiled over a nearly 25-year span.
This reveals how dif cult it would be to conduct a replication study that
similarly controls for a le-drawer effect.
Notes
1 This is a subjective thought on my part, probably a deduction based on
my knowledge that Donovan had asthma.
2 This company name is a pseudonym.
Acknowledgments
I am grateful to JSE Associate Editor Daryl Bem for advice on the exposition
and to Giuseppe Maggiore for technical advice.
References Cited
Barrett, D. (1991/1992). Through a glass darkly: Images of the dead in dreams. OMEGA—Journal
of Death and Dying, 24(2).
Burns, A., & Ili e, S. (2009). Alzheimers Disease. British Medical Journal, 338(7962), 467–471.
Myers, F. W. H. (1920). Human Personality and Its Survival of Bodily Death. Longsmans, Green.
Paquette, A. (2011). Dreamer: 20 Years of Psychic Dreams and How They Changed My Life. 6th Books.
Paquette, A. (2012). A new approach to veridicality in dream psi studies. Journal of Scienti c
Exploration, 26(3), 589–610.
Scargle, J. D. (2000). Publication bias: The “ le-drawer” problem in scienti c inference. Journal of
Scienti c Exploration, 14(1), 91–106.
Stevenson, I. (1970). Telepathic Impressions. Charlottesville, VA: University Press of Virginia. pp. 170–173.
Stevenson, I. (1995). Six modern apparitional experiences. Journal of Scienti c Exploration, 9(3),
351–366.
Article
Full-text available
Article
Full-text available
Many people are persuaded of the existence of psychic phenomena by their own spontaneous experiences of apparent psi. However, without some measure of how often psi-suggestive experiences can be expected to occur purely by chance, it is difficult to determine the epistemic import of these cases. While methods have been developed to find statistical baselines for some spontaneous cases—specifically, ones in which cases of interest can be identified before any verification of their supposedly psychic content has been obtained—many spontaneous cases of purported psi are not identified as such until after some degree of spontaneous verification occurs: for instance, when a person notices a striking correspondence between their mental state and another event to which that state appears to have no physical causal connection. This paper develops a method applicable to these cases—the time-slice method for calculating baseline correspondence potential—and thus enables individuals to determine the epistemic import of their own spontaneous psi experiences.
Article
Full-text available
This study examined dreams about the dead. There were four categories of activities reported in these: the deceased described the state of death, delivered messages to the living, sought to change the circumstances of their death, or gave loved ones a chance to say ‘goodbye.’ Some of these categories occurred at a particular point in the grief process; others occurred at any time after the death. A remarkably large number of the dead telephoned. These dreams are discussed in terms of what they illuminate about attitudes toward mortality and loss.
Article
Full-text available
A problem in spontaneous dream psi studies is validation of purported psi elements. Dreams rarely have enough specifi city to satisfy critics that they point to paranormal knowledge of a specifi c event. This creates evidential hurdles to overcome when evaluating whether a dream contains psi-derived information such as scenes of future events or physically distant locations. In this study, the goal is to arrive at a reasonable method to establish that information derived from spontaneous dream experiences can be established as veridical. To accomplish this, a method for fi nding the equivalent of a target within a spontaneous experience is used to fi x a specifi c locale and time for comparison. Adverse scoring criteria are used to address complaints about confi rmation bias. The result is a psi-adverse method for scoring spontaneous experiences that are anchored to a specifi c locale and time. This method regularly produced signifi cant p-values when used to assess 20 consecutive dream records, comprising 598 individual line items. These records were taken as a sample from a group of 3,305 dream records made over the past 22 years by the author.
Article
Publication bias arises whenever the probability that a study is published depends on the statistical significance of its results. This bias, often called the file-drawer effect since the unpublished results are imagined to be tucked away in researchers' file cabinets, is potentially a severe impediment to combining the statistical results of studies collected from the literature. With almost any reasonable quantitative model for publication bias, only a small number of studies lost in the file-drawer will produce a significant bias. This result contradicts the well known Fail Safe File Drawer (FSFD) method for setting limits on the potential harm of publication bias, widely used in social, medical and psychic research. This method incorrectly treats the file drawer as unbiased, and almost always miss-estimates the seriousness of publication bias. A large body of not only psychic research, but medical and social science studies, has mistakenly relied on this method to validate claimed discoveries. Statistical combination can be trusted only if it is known with certainty that all studies that have been carried out are included. Such certainty is virtually impossible to achieve in literature surveys.
Dreamer: 20 Years of Psychic Dreams and How They Changed My Life. 6 th Books
  • A Paquette
Paquette, A. (2011). Dreamer: 20 Years of Psychic Dreams and How They Changed My Life. 6 th Books.
Telepathic Impressions
  • I Stevenson
Stevenson, I. (1970). Telepathic Impressions. Charlottesville, VA: University Press of Virginia. pp. 170-173.
Six modern apparitional experiences
  • I Stevenson
Stevenson, I. (1995). Six modern apparitional experiences. Journal of Scientifi c Exploration, 9(3), 351-366.