The correct terms when discussing suicide follow a specific framework: say "died by suicide" rather than "committed suicide," use "suicide attempt" for nonfatal behavior with intent to die, and distinguish "suicidal ideation" (thinking about it) from planning and preparation. The influential mid-century shift by the American Association of Suicidology (circa 1960s) and later the World Health Organization pushed clinicians, journalists, and researchers toward terms that are precise, non-judgmental, and safe for vulnerable readers, a vocabulary now anchored in the DSM-5-TR and CDC surveillance definitions.
How the terminology framework works
The vocabulary exists to solve three problems at once: scientific precision, legal-regulatory consistency, and contagion risk. Words carry valence, and the words used about suicide measurably change how audiences respond. WHO guidance on responsible reporting advises against language that presents suicide as a solution, sensationalizes a death, or describes method in detail, and that guidance is why "died by suicide" replaced "committed" in style guides. The mechanism is not merely politeness: a 2018 systematic review by Niederkrotenthaler and colleagues in Social Science & Medicine found that harmful reporting practices, including explicit method description and dramatic framing, are associated with increased suicides. This is the Papageno versus Werther effect distinction: Werther-style coverage (naming method, romanticizing) raises risk, while Papageno-style narratives (coping, survival) appear protective. The vocabulary is therefore a safety intervention, not etiquette.
Precision matters because epidemiology depends on definitional boundaries. A "suicide attempt" requires some intent to die; a "suicidal gesture" implies communication without that intent; "self-harm" (or nonsuicidal self-injury, NSSI) refers to self-inflicted damage without suicidal intent. The CDC's surveillance definitions separate suicide (death from self-directed injury with evidence of intent) from self-harm, and coding systems such as ICD-10 and ICD-11 distinguish intentional self-harm (X60-X84 in ICD-10) from undetermined intent, accidents, and assault. Without that split, rates become incomparable across jurisdictions. Plainly: the words are part of the measurement instrument.
Why the specific terms matter
Consider what "committed suicide" implies. The verb "committed" historically attaches to crimes and sins, which is why the phrase can stigmatize survivors and deter help-seeking among people at risk. Organizations including the Associated Press stylebook and major mental-health bodies now recommend "died by suicide" or "killed himself/herself." Similarly, "successful suicide" is replaced by "died by suicide," and "failed attempt" by "suicide attempt," because success and failure language frames death as an achievement. The aim is a register that is factual and compassionate without being euphemistic. "Passed away from suicide" or "lost their battle" obscures the cause and can be coded ambiguously.
There is a counterintuitive professional caveat here: "committed suicide" does not affect a deceased person, and some survivors of loss use it about their own family member because it is the language of their era. Guidance asks professionals to gently model preferred terms rather than correct the bereaved. Language is not a purity test; it is a tool for reducing harm.
Quantified context: how often these distinctions surface
Suicide is a major public-health concern globally. WHO estimates roughly 720,000 deaths annually, and the CDC ranks suicide among the leading causes of death in the United States, particularly among young and middle-aged adults. Those headline figures are the reason terminology standards exist: reporting guidelines and research vocabulary are attached to a high-volume, high-consequence event class. The precise numbers shift year to year and are revised, so cite a current surveillance dataset rather than a remembered figure.
Exact coded definitions are equally consequential. NCHS guidance defines suicide as death resulting from self-directed injurious behavior with evidence of intent to die; an "attempt" is nonfatal self-directed potentially injurious behavior with any intent to die; "ideation" is thinking about, considering, or planning suicide. The CDC's broader self-directed violence framework in 2011 drew a taxonomy dividing suicidal from nonsuicidal behavior. A hospital admission coded as self-harm under X60-X84 is not automatically a suicide attempt, and a "suicide attempt" without intent, as some coding systems record it, is a classification tension, not a synonym.
Edge cases: where the vocabulary strains
Some deaths cannot be cleanly classified. A fatal overdose in a person with an active suicide plan and a known intent will be coded suicide; an identical overdose in a person without documented intent may be coded undetermined intent (Y10-Y34) or accident. Intent is reconstructed from messages, prior attempts, and witness accounts, which makes classification imperfect. This is where "undetermined" or "manner undetermined" terminology appears; the medical examiner vocabulary (natural, accident, suicide, homicide, undetermined) does not match clinical or research vocabulary. Professionals explicitly reconcile these framings rather than assuming equivalence.
Assisted dying introduces another edge case. "Medical aid in dying" (MAID) and "physician-assisted death" refer to a legally sanctioned practice in some jurisdictions for terminally ill adults; the practice is deliberately distinguished from suicide in the legal and medical literature, and conflating the two is both inaccurate and inflammatory. "Euthanasia" refers to a clinician administering the lethal agent, distinct from self-administration. Similarly, "suicide by cop" (officer-involved death where the person acts to provoke lethal force) is a specific forensic concept, and "murder-suicide" describes a homicide followed by the perpetrator's death.
Warning terms also have precise functions. "Warning signs" denote recognizable signals (talking about dying, giving away possessions, sudden calm after agitation), while "risk factors" describe background vulnerability (prior attempt, access to lethal means, social isolation). The two are not interchangeable, and standards distinguish "screening" tools (e.g., the Columbia Protocol, the PHQ-9 item 9) from "risk assessment" performed by clinicians.
"Died by suicide" — correct; avoids crime framing. "Committed suicide" — avoid; stigmatizing. "Suicidal ideation" — thinking about suicide; no necessary action. "Suicide attempt" — nonfatal behavior with intent to die. "Self-harm / NSSI" — self-injury without suicidal intent. "Survivors of suicide loss" — people grieving a death by suicide, not the decedent.
Misconceptions and the wrong-terms ledger
The most common error is treating "suicide" as a verb phrase with moral weight ("he committed suicide"). The second is using "successful" and "failed," which frames death as achievement. The third is collapsing ideation, self-harm, and attempt into one bucket, which corrupts both risk assessment and research. A fourth misconception is that the "right" terminology settles causation; it does not. The fifth is assuming "suicidality" is a stable trait; acute risk can fluctuate over hours, and terms like "active suicidal ideation" (with plan or intent) versus "passive ideation" (wishing to be dead without planning) capture that gradient.
A subtle point: "suicidal gesture" is now widely discouraged because it implies manipulativeness and dismisses real distress. Clinicians prefer "suicide attempt" or "suicidal behavior" and reserve "nonsuicidal self-injury" for the distinct phenomenon. Meanwhile "parasuicide" was an older term; its use varies, so contemporary writers should define it if they use it at all.
Comparison of common terms and alternatives
| Common phrase | Preferred term | Why the change matters |
|---|---|---|
| Committed suicide | Died by suicide | Removes crime or sin connotation |
| Successful suicide | Died by suicide | "Success" frames death as an achievement |
| Failed attempt | Suicide attempt / nonfatal attempt | "Failure" misrepresents a health event |
| Suicidal gesture | Suicide attempt / suicidal behavior | "Gesture" dismisses genuine distress |
| Completed suicide | Suicide death | Neutral, non-triumphal framing |
Timeline: what unfolds after a suicide death or attempt
| Phase | Time frame | Terminology in play |
|---|---|---|
| Immediate response | Hours | "Undetermined" vs "suicide" may wait on investigation |
| Investigation | Days to weeks | Medical examiner classifies manner of death |
| Postvention | Weeks to months | "Survivors of loss," contagion risk language |
| Follow-up care | Months | "Attempt survivor," "lived experience" |
| Surveillance coding | Quarterly to annual | ICD codes and death certificate cause of death |
| Research publication | Years | "Intentional self-harm," "self-directed violence" |
This sequence shows how terminology evolves as information changes. An early report may describe a "death under investigation" that later becomes a "suicide death" in the final certificate. That is normal and not contradictory.
Practical language standards in behavior
For journalists, clinicians, and organizations, several concrete rules apply. First, refer to "a death by suicide," not "a suicide" as a noun. Second, do not describe methods or locations in detail; say "a fatal overdose" or "a self-inflicted injury" without specifics. Third, do not present a single cause; avoid "caused by" language and say "associated with" or "in the context of." Fourth, include help-seeking information, such as the 988 Suicide & Crisis Lifeline in the United States, and avoid images of locations or methods. Fifth, do not imply that suicide is a solution or describe it as "inevitable." Sixth, flag content when describing a specific death to reduce identification risk.
For clinicians and researchers, add precision thresholds: use "suicide attempt" only where intent to die is present; use "self-harm" for injury without intent; document "active" versus "passive" ideation; and specify time frame ("last 12 months," "lifetime"). For a family member or friend, the practical threshold is simple: if someone is talking about wanting to die, give away possessions, or says goodbye, treat it as urgent and contact a crisis line or emergency services without delaying. Those specific signals, called warning signs, are the ones worth memorizing.
Limitations and what the vocabulary cannot do
Language is a harm-reduction tool with measured but bounded effects; it does not substitute for clinical care, means restriction, or surveillance. Classification codes disagree at the boundaries (undetermined versus suicide), and intent is often reconstructed rather than observed, so precise rates carry uncertainty. Some fields are still evolving, including terms for assisted dying and digital self-harm, and legacy terms persist in older records. The most rigorous position is to use the current standard vocabulary, define terms when audience familiarity varies, and acknowledge that wording choices are evidence-informed rather than absolute.
For medical and mental-health content, this article is general educational information, not a diagnosis or treatment plan. It does not promise any outcome and cannot replace evaluation by a qualified clinician. If you or someone near you is at immediate risk, contact local emergency services or a crisis line such as 988 in the United States.