extras—it had better go that way than for more pills and more doc-
tor bills.” To her daughters, she writes, “Try to forgive me for what
The Desire for Death ● 103
I’ve done—your father would be so much better for you. It will be
harder for you for awhile—but so much easier in the long run—I’m
getting you all mixed up.”19
Another example: “I started to list the people who wouldn’t mind
if I wasn’t around. I clearly wasn’t a good wife for my ex-husband. He
wouldn’t miss me. And I never felt that comfortable in my role as a
mom—didn’t feel like I was a good mom necessarily . . . It’s like I’ll
be a burden off their backs. Clearly their lives will be enhanced be-
cause I’m not around. At that point, I honestly felt I was doing them
a favor.”20
Alvarez quoted a man who had spent most of his life in mental in-
stitutions: “If I commit suicide, it will not be to destroy myself but to
put myself back together again. Suicide will be for me only one
means of violently reconquering myself . . . By suicide I reintroduce
my design in nature, I shall for the first time give things the shape of
my will.”21 A paraphrasing of his words would be that “I feel so inef-
fective that I focus on one sure way I can have dramatic effect,
namely my own death by suicide.” The man’s words can also be seen
as another example of the blending of life and death that I believe oc-
curs in at least some instances of serious suicidal behavior. In Chap-
ter 2, I described Shneidman’s example of Ariel, who viewed things
in the cemetery as “graceful” and “gracious.” The blending seemed to
be of death by suicide with beauty and caring. In this man’s example,
by contrast, the blending seems to be of death by suicide with a lively
sense of taking charge and imposing will. One type seems to involve
seeing suicide as a means to belong or be cared for; the other type as
a means to feel effective—these types correspond to the bedrock is-
sues of thwarted belongingness and perceived burdensomeness.
Research studies also show a connection between feeling ineffec-
tive or being a burden on the one hand, and feeling suicidal on the
other. In the sections below, this work is summarized, starting with
104 ● WHY PEOPLE DIE BY SUICIDE
research on the general connection between depressive symptoms
(one of which is suicidality) and experiences of ineffectiveness, espe-
cially in social domains.
Social Skills and Depression
There is little doubt that depression and suicidality are associated
with social skills problems.22 People with depression consistently eval-
uate their own social skills more negatively than do nondepressed
people23—this result applies to depressed children as well as de-
pressed adults.24
Given the pessimistic outlook associated with depression and
suicidality, it is not too surprising that depressed people evaluate
their social skills negatively. But is there an objective basis for their
negative evaluations; that is, do others also rate their social skills neg-
atively? In fact, when observers or conversational partners rate de-
pressed people’s social skills, a common finding is that depressed
people’s skills are rated as lower than those of nondepressed con-
trols.25
A related line of research has examined the communication be-
haviors of depressed versus nondepressed people. For example, as
compared to nondepressed people, depressed people speak more slowly,
and with less volume and voice modulation; depressed people have
longer pauses in their speech patterns, and take longer to respond
when someone else addresses them.26 Voice modulation and rate ac-
count for the animated and appealing qualities of speech; because
depressed and suicidal people’s speech often lacks these qualities,
they are often negatively perceived by others.
In addition to the quality of speech, the content of speech in de-
pression has been evaluated. In studies of married couples with a de-
pressed member, themes involving dysphoric feelings and negative
self-evaluation were likely to emerge in their conversations.27 It has
been found that depressed people may be at their most expressive
The Desire for Death ● 105
with their spouses when they are discussing negative issues.28 There is
some evidence that negativity in social interactions is particularly
likely to emerge between depressed people and intimate relationship
partners (as opposed to strangers or nonintimate acquaintances).
For example, in a study of depressed and nondepressed students dis-
cussing “events of the day” with either a friend or a stranger, re-
searchers reported that depressed students tended to withhold nega-
tive verbal content when talking with strangers but were more likely
to disclose negative topics when talking with a friend.29 Other work
has found that depressed people are more likely than nondepressed
people to make unsolicited negative self-disclosures.30 This problem
may be particularly acute in relationships between depressed people
and intimate others. Depressed and suicidal people, then, are likely
to instill negative views of themselves in others, perhaps especially
those close to them.
This research shows that depressive symptoms (a key one of which
is suicidality) are associated with feelings of social ineffectiveness,
and that these may manifest particularly in close relationships. These
findings are roughly consistent with the view espoused here, that an
extreme form of ineffectiveness, perceived burdensomeness, is impli-
cated in suicidality, but it would be more satisfying if studies focused
specifically on perceived burdensomeness and suicidality. In the next
section, several such studies are summarized.
Perceived Burdensomeness Contributes to Suicidality
To my knowledge, five studies have been framed as direct tests of
the possibility that perceived burdensomeness is involved in seri-
ous suicidal behavior; all five affirm the connection. One study was
conducted to test DeCatanzaro’s model of self-preservation and self-
destruction.31 Arguing from a sociobiological or evolutionary stand-
point, this model posits that staying alive actually may reduce inclu-
sive fitness for an individual if the individual is low in reproductive
106 ● WHY PEOPLE DIE BY SUICIDE
potential and if the individual’s continuing to live poses such a bur-
den to close kin that it costs them opportunities for reproduction.
One upshot of this view is that suicidal behavior may have been se-
lected for in the course of evolution—a controversial point to which
I return later.
To test this model, Brown and colleagues conducted a question-
naire study of college students and found the predicted correlation
between feeling a burden on kin and suicidality. Burdensomeness
stood out as a unique and specific predictor of suicide-related symp-
toms even when other variables, such as the individual’s repr
oductive
potential, were accounted for.32
My students and I also conducted empirical tests of the associa-
tion between perceived burdensomeness and suicidal behavior. We
trained raters to evaluate actual suicide notes regarding the following
dimensions: perceived burdensomeness, hopelessness, and general-
ized emotional pain. The raters read each note and then made three
separate ratings on a 1-to-5 scale of the amount of perceived burden-
someness, hopelessness, and generalized emotional pain conveyed
by the note.
Unknown to the raters, half of the notes were from people who
died by suicide, and half were from people who attempted suicide
and survived. The goal of the study was to compare perceived bur-
densomeness versus Shneidman’s emphasis on emotional pain versus
Beck’s emphasis on hopelessness. In statistical analyses, the notes
from those who died by suicide contained more perceived burden-
someness than notes from attempters; no effects were found regard-
ing hopelessness and emotional pain. This study’s relatively strin-
gent comparison of notes from those who died by suicide to notes
from those who attempted and survived (which distinction was un-
known to raters), with perceived burdensomeness emerging as the
only unique predictor of death by suicide, added to our confidence in
the results.33
The Desire for Death ● 107
Moreover, a second study from this same paper on a separate col-
lection of notes took a similar approach, except that all notes were
from those who died by suicide, and perceived burdensomeness,
hopelessness, and generalized emotional pain were used as predic-
tors of lethality of suicide method (e.g., self-inflicted gunshot wound
was viewed as relatively more lethal than overdose). Here again, per-
ceived burdensomeness was a significant predictor of lethality, whereas
hopelessness and generalized emotional pain were not. The conver-
gence of the two studies made the findings more persuasive.
A survey on reproductive behavior, quality of family contacts, and
suicidal ideation on several hundred community participants as well
as on five high-suicide-risk groups (e.g., general psychiatric patients
and incarcerated psychiatric patients) found that perceived burden-
someness toward family and social isolation were especially corre-
lated with suicidal ideation.34 It is important to note that these two
variables correspond to two of the three main aspects of the present
model, burdensomeness and lack of belonging.
My students and I recently completed a study on perceived bur-
densomeness and suicidality among 343 adult outpatients of the
Florida State University Psychology Clinic. Areas of diagnosis for
these patients were represented in the following proportions: 39 per-
cent mood disorder, 14.6 percent anxiety disorder, 6 percent sub-
stance use disorder, 12.2 percent personality disorder, 9 percent ad-
justment disorder, and 18 percent other disorders.
We hypothesized that perceived burdensomeness would directly
relate to both past number of suicide attempts and an index of
current suicidal symptoms, and furthermore, that this relationship
would exist even when accounting for known risk factors such as
personality disorder status, depressive symptoms, and hopelessness.
This is similar to the “kitchen sink” approach in our paper described
in the last chapter. We also wanted to see if there was a special con-
nection between perceived burdensomeness and suicidality. So, for
108 ● WHY PEOPLE DIE BY SUICIDE
purposes of comparison, the associations of hopelessness to suicide
indices (controlling for personality disorder status, depressive symp-
toms, and perceived burdensomeness) were examined. Our reason-
ing was that if perceived burdensomeness is important in suicidality,
its associations to suicidality should be as rigorous as those regarding
the documented risk factor of hopelessness.
Here, as in other studies, the connection of perceived burden-
someness to suicidality was supported. Specifically, there was an as-
sociation between measures of perceived burdensomeness and sui-
cidality, and this association persisted even when a host of other
variables was accounted for (specifically, age, gender, hopelessness,
depressive symptoms, and personality disorder status). Furthermore,
the link between perceived burdensomeness and suicidality was at
least as strong as that between hopelessness and suicidality.
Other Research on Burdensomeness and Suicidality
Although not direct tests of the burdensomeness view of suicidality,
several other studies have reported results consistent with this per-
spective. For example, in a study on genuine suicide attempts vs.
nonsuicidal self-injury, genuine attempts were defined as those with
lethal intent and bodily injury; nonsuicidal injury involved things
like superficial cuts made without the intent to die. The researchers
reported that genuine suicide attempts were often characterized by
a desire to make others better off, whereas nonsuicidal self-injury
was often characterized by desires to express anger or punish one-
self.35 “Making others better off ” is similar in concept to perceived
burdensomeness. In this study, as in those reviewed earlier, an index
related to perceived burdensomeness was associated with relatively
severe suicidal behavior. Persuasive aspects of the study were that
suicide attempters were compared to a compelling control group
(those engaging in nonsuicidal self-injury), and that effects of burden-
someness were compared to effects of other relevant dimensions,
The Desire for Death ● 109
such as anger expression and self-punishment, with burdensomeness
emerging as among the most important.
One might imagine that feelings of burdensomeness might be
particularly acute among depressed, terminally ill people. In fact,
Filiberti and colleagues studied vulnerability factors for suicide in
five terminal cancer patients who died by suicide while they were
cared for at home by palliative care teams.36 Of all the various possi-
ble factors identified, being a burden on others was judged to be one
of the two most important. Interestingly, the other was fear of the
loss of general competence, which, according to the current model, is
a related but weaker form of perceived burdensomeness. In a study
mentioned earlier on 3,005 psychiatric patients at risk for suicide, of
whom thirty-eight died by suicide within two months of evaluation,
nine clear risk factors were identified, among which were feelings of
being a burden on others.37
O’Reilly and colleagues reached similar conclusions. They studied
psychiatrists’ reports on their patients’ suicides, and of the three vari-
ables seen as frequently present in the month preceding suicide, one
was “feeling a burden on others.” The two others, incidentally, were
related to thwarted belongingness—social withdrawal and turning
down help from others.38
If burdensomeness were involved in suicidality, one might expect
that suicidal patients’ self-views would be especially discrepant from
how they view other people. In a telling study, researchers compared
self-views as well as views of other people among suicidal patients
and psychiatric and nonpsychiatric controls. Not surprisingly, the
suicidal patients endorsed more negative self-views than the other
two groups, similar to the research reviewed earlier in which de-
pressed people evaluated their social skills negatively. Notably, the
suicidal patients rated other people more favorably than did the
other two groups.39 The conclusion is that suicidal people view
themselves in quite negative terms, and this is particularly true when
110 ● WHY PEOPLE DIE BY SUICIDE
their self-views are compared to their views of other people. This
marked discrepancy in suicidal people between self-ratings and
views of others could instill a sense of burdensomeness—the idea
that “I’m bad but I’m especially bad when compared to others, who
are good.”
A roughly similar dynamic emerged in a study of chronically ill
patients. The researchers predicted that among chronically ill pa-
tients who are dependent on a caregiving spouse (but whose health
status prevents their reciprocating the care), receipt of support may
exacerbate feelings of burdensomeness, and thus may increase sui-
cidality. This hypothesis was in “grave danger of refutation” (Popper,
1959), given the pervasive positive associations between social sup-
port and health. But the hypothesis survived the test—the study
found a positive correlation between social support and suicidal ideation among physically ill participants.40
The concept of perceived burdensomeness is fairly easy to under-
stand as applied to adults—the image of the failed breadwinner
imagining his family will be better off without him is tragic but not
hard to conceive. But what about perceived burdensomeness as ap-
plied to younger people, including children? Young people do die by
suicide, and so if perceived burdensomeness plays a role in suicide in
general, it should be applicable to youth too. In fact, researchers have
studied burdensomeness and suicidality in youth in their work on
the “expendable child.”41 These authors hypothesized that suicidal
Why People Die By Suicide Page 14