Public explainer · Change and identity

Does changing desire mean I don’t know who I am?

By Nicholas Julian Madison · Public explainer v0.1 · Structured evidence synthesis · Evidence reviewed August 13, 2026

Begin with what changed Keep identity and desire distinct Review the evidence

Not by itself.

A change in desire can be real, important, and sometimes disorienting. But the fact that something changed does not establish that you misunderstood yourself before, that your identity is false now, or that you must adopt a different identity.

It establishes that something changed.

That is where inquiry can begin. It is not where interpretation has to end.

Desire ≠ identityEvent ≠ profileDesire ≠ consent

Specify the observation

What changed?

“Desire” can sound like one clearly defined thing. In practice, people use the word for different experiences.

Someone might mean:

Wanting sexual activity generally.
Wanting sex with a particular person.
Wanting one activity but not another.
Thinking about sex more or less often.
Finding desire easier or harder to access.
Experiencing desire only under certain conditions.
Noticing a new direction, intensity, or pattern of wanting.

Researchers also do not measure desire in only one way. One influential questionnaire distinguished desire involving another person from solitary desire, while acknowledging that sexual desire is not easily reduced to a single dimension. In qualitative interviews, people could often distinguish desire from arousal when asked, but their descriptions also overlapped or reversed the expected sequence. Desire sometimes preceded arousal; sometimes it appeared responsive to a person or stimulus; sometimes the boundary between the two was unclear. DI-S01, DI-S02

This does not make the terms useless. It means we should specify them.

Desire is not automatically the same thing as attraction, fantasy, genital response, pleasure, behavior, willingness, or consent. A person may experience one without another. If the thing that changed is not identified carefully, an explanation may answer a different question from the one the person is asking.

It establishes that something changed.

Do not silently cross levels

A moment, a pattern, and a person are different levels

Desire can differ from one moment to another. It can also show recurring patterns or change over longer periods.

In three longitudinal studies, researchers measured desire across intervals ranging from moments within a week to three occasions spanning 13 years. Desire varied within both women and men over the short term. The study found some average sex differences over the longer period, but the effects were small, the distributions overlapped substantially, and short-term variability did not support the stereotype that women’s desire fluctuates while men’s remains stable. DI-S03

A 30-day diary study of 133 mixed-sex couples also found recurring fluctuations in reported desire. Partners sometimes showed similar rhythms, but their desire was not identical at every point. The study concerned a particular nonclinical, cohabiting population; it did not establish a universal rhythm or explain what any fluctuation meant to an individual. DI-S04

The canonical distinctions must remain intact:

Event ≠ profile.
Profile ≠ person-level pattern.
Category ≠ person.

This article does not redefine those terms. It uses the distinctions to limit inference: evidence at one level should not be silently treated as evidence at another.

A single event does not establish a profile. A recurring profile does not, by itself, tell us what category a person belongs to. Even a lasting change does not dictate what meaning the person must give it.

Relevant does not mean proven as the cause

Conditions matter without providing one complete explanation

Desire does not occur outside an organism, a history, a body, a relationship, or a situation. State, time, fatigue, stress, closeness, conflict, learning, health, hormones, and wider social conditions may all be relevant questions.

Relevant does not mean proven as the cause.

State and stress

Experience-sampling research has found associations between desire and momentary states such as happiness, stress, tiredness, anger, loneliness, and partner closeness. These associations did not follow one universal pattern, and they did not support the idea that only women’s desire is context-sensitive. DI-S03

In a 14-day study of 59 young, healthy, heterosexual people in relationships, higher momentary stress coincided with lower reported desire and arousal. Desire and sexual activity were also associated with lower stress at later measurements. Because the associations could operate in more than one direction—and because the sample was narrow—the researchers did not claim that stress straightforwardly caused the observed changes. DI-S07

Stress can therefore be part of the context without becoming the explanation for every person.

Relationship

Relationship conditions can be associated with partner-specific desire. In an experience-sampling study, 134 participants reported intimacy, desire, and sexual activity up to ten times a day for seven days. Intimacy and desire were associated, but the strength of that relationship changed depending on the time between measurements. DI-S05

A separate series of experimental, observational, and daily-experience studies found that perceived partner responsiveness was associated with greater desire for that partner. The results were not identical across participants, and the studies did not establish that relationship quality supplies a complete explanation for desire. DI-S06

Partner-specific findings should not be generalized to all desire. A change involving one person does not necessarily establish a global change in capacity, orientation, or identity.

Hormones

Hormones are often invoked as if they offer a direct explanation. The evidence is more bounded.

In a study of 213 university students who completed daily reports across two menstrual cycles, the average pattern included a small midcycle increase in desire. Individual patterns varied substantially: some participants showed a midcycle increase, some a perimenstrual increase, and others no cyclical change. Psychological changes were more predictive than the physical changes measured in that study. DI-S08

A month-long study of 41 young men found no significant positive relationship between day-to-day changes in baseline testosterone and day-to-day desire. That does not mean testosterone is irrelevant to every sexual process or clinical condition. It means a broad statement such as “testosterone controls desire” cannot explain every within-person change. DI-S09

Hormonal findings depend on which hormones, populations, timescales, health states, and forms of desire are measured. A group average cannot identify why one person’s desire changed, and it cannot determine what that change means for identity.

Learning and safety

Learning is another broad category. Human studies provide some evidence that experiences can alter responses to particular cues. But the evidence does not justify the claim that a person’s identity is trained into or out of existence.

For example, a small field-conditioning study paired scents with sexual or nonsexual interactions in 14 heterosexual couples. It found evidence of altered genital responding in the men tested. The study measured a narrow learned arousal response—not desire, identity, consent, or person-level meaning. DI-S10

Safety also requires specification. Physical safety, emotional safety, social safety, freedom from pressure, and trust within a relationship are not interchangeable. Nor should a change in desire be used to infer that someone is unsafe, traumatized, repressed, or healed. Those are different questions requiring different evidence.

Conditions can matter without granting an observer the authority to decide which condition explains the person.

Questions, not a mechanical diagnosis

What the Capacity–Pathway framework can clarify

The Neurosexology architecture is:

The canonical architecture remains

ORGANISM → CAPACITIES → PATHWAYS → ACCESSIBILITY → DISCOVERED → under CONDITIONS RECRUITED → EXPRESSION

Conditions act across the architecture.

If desire appears, disappears, or changes, the observed expression does not tell us everything that preceded it. A capacity may be present while access to it varies. Something may be undiscovered without being inaccessible or absent. Conditions may affect accessibility or recruitment without proving that the organism acquired an entirely new capacity.

The framework gives us places to ask questions. It does not allow us to work backward from one outcome to one cause.

A similar component across two events also does not make those events identical. In encounters involving the same person, sensation, fantasy, or activity, the organism may recruit capacities and pathways differently under different conditions. Sexual can describe a mode of recruitment, but sexual is not an additional recruitment route, a mechanism by itself, or a permanent property of the component.

This does not mean every change in desire is merely a change in recruitment. It means the outcome alone cannot settle what changed within the architecture.

Distinguishable, potentially related, not interchangeable

Identity is not a mechanical readout of desire

Research on sexual orientation often measures identity, attraction, and behavior separately because they do not always align.

A three-year study of 744 adolescents found both stability and change in self-described identity and romantic attraction. It also found patterns in which identity, attraction, and behavior did not coincide in a uniform way. The sample came from three low-income high schools in one region, so its percentages should not be treated as universal. Its more durable contribution is methodological: different dimensions of sexuality can follow different trajectories. DI-S13

Across other longitudinal samples, stability has often been more common than change, while change has remained a documented possibility. This pattern appears in young-adult data collected about six years apart, in a midlife sample followed for ten years, and in a national probability sample of sexual-minority adults followed for two years. The reported rate of change differed substantially across those studies because their populations, labels, intervals, and methods differed. DI-S12, DI-S11, DI-S15

Research involving asexual and gray-asexual participants offers another reason not to collapse identity, attraction, and desire. In one three-wave study, most participants maintained their identity between adjacent waves, while measures of attraction and solitary or dyadic desire showed related but nonidentical patterns. The sample was small and cannot represent every asexual person, but it demonstrates why these variables should be measured separately. DI-S14

Together, these studies do not show that identity is arbitrary. More narrowly, they show why researchers should not calculate sexual-orientation identity mechanically from one observation of desire, attraction, or behavior.

The broader safeguard in this article is governed rather than statistically derived: no observation authorizes an outsider to determine what a person’s identity must mean.

A person may experience changing desire while retaining an identity that remains accurate and meaningful. Another person may revise an identity. Someone else may remain uncertain, use several forms of language, or decline a label. Longitudinal research can document some of these patterns in the identity categories it measures. It cannot decide which interpretation a particular person must choose.

A later change also does not automatically make an earlier identity dishonest. People describe themselves using the experiences, meanings, language, and knowledge available to them at the time.

Finding, interpretation, and meaning

What the evidence can—and cannot—tell us

Research can sometimes tell us:

That a reported measure changed.
How frequently change appeared in a particular sample.
Whether two measured variables were associated.
Whether a pattern persisted across the period observed.
Whether different dimensions moved together or separately.

Those findings matter. But each leaves another question open.

A self-report is not the same thing as a physiological measure. An observed association is not a mechanism. A measure of genital response is not a measure of wanting. A group pattern is not an explanation of one person. A change between two survey answers cannot, by itself, tell us what the change felt like or meant.

The numbers can tell us what happened. They cannot, by themselves, tell us what the person experienced.

The goal is not to prove one explanation. The goal is to stop different kinds of things from being confused with one another.

Return without forcing a verdict

So, does changing desire mean you do not know who you are?

Not by itself.

It may mean that something about your experience is different. The difference may be brief, recurring, lasting, person-specific, activity-specific, relationship-specific, or not yet understood. It may eventually matter to identity, or it may not.

Change does not have to be denied. Identity does not have to be declared unreal.

The person is not the contradiction.

The person always outranks the framework.

Evidence reviewed August 13, 2026

Sources behind this distinction

The bounded evidence set supports claims about the measurement and variability of desire, qualified contextual associations, and the separate study of identity, attraction, behavior, and desire. It does not determine why one person changed or what the change must mean.

DI-S01
Construct and measurement
Spector, Carey, and Steinberg (1996), ‘The Sexual Desire Inventory.’ Questionnaire-development studies; revised sample included 380 participants.
Use: Supports distinguishing dyadic from solitary desire and rejecting a wholly unitary measure.
Limit: Early psychometric work; binary sex reporting; does not establish that two dimensions exhaust desire.
Open source record
DI-S02
Desire and arousal in participants’ language
Mitchell, Wellings, and Graham (2014), ‘How do men and women define sexual desire and sexual arousal?’ Thirty-two qualitative interviews in southeast England.
Use: Supports the claim that people can distinguish desire and arousal while also describing overlap, reversals, and ambiguity.
Limit: Small purposive sample; describes participants’ concepts, not universal definitions or mechanisms.
Open source record
DI-S03
Within-person variability across timescales
Harris et al. (2023), ‘Does Sexual Desire Fluctuate More Among Women than Men?’ Three longitudinal studies: a 13-year panel study and two seven-day experience-sampling studies.
Use: Central evidence that desire varies within people across multiple timescales and that short-term variability is not unique to women. Supports cautious discussion of affective and relationship-oriented correlates.
Limit: Samples and measures differed across studies; the long-term sex difference was small; does not explain individual causes.
Open source record
DI-S04
Patterns and synchrony in couples’ desire
Vowels et al. (2018), ‘Using spectral and cross-spectral analysis to identify patterns and synchrony in couples’ sexual desire.’ Thirty-day dyadic diaries from 133 cohabiting mixed-sex couples.
Use: Supports recurring within-person fluctuation and possible partner synchrony without identical desire at every point.
Limit: Nonclinical, cohabiting heterosexual sample; unusual analytic method; does not supply a universal cycle.
Open source record
DI-S05
Intimacy and sexuality in daily life
van Lankveld et al. (2018), ‘The associations of intimacy and sexuality in daily life.’ Experience sampling up to ten times daily for seven days; 134 participants.
Use: Supports a time-sensitive association among intimacy, desire, and partnered sexual activity.
Limit: Predominantly heterosexual and White sample; one partner per relationship; association weakened with time lag and does not establish a complete causal pathway.
Open source record
DI-S06
Perceived partner responsiveness
Birnbaum et al. (2016), ‘Intimately connected.’ Three studies using experimental, observational, and daily-experience methods.
Use: Supports the bounded claim that perceived partner responsiveness can be associated with partner-specific desire.
Limit: Context- and sample-dependent; stronger effects reported for women; cannot be generalized to solitary or global desire.
Open source record
DI-S07
Stress, desire, arousal, and activity
Mües et al. (2025), ‘Bidirectional associations between daily subjective stress and sexual desire, arousal, and activity.’ Fourteen-day ecological momentary assessment; 59 young, healthy, heterosexual people in relationships.
Use: Supports concurrent associations between stress and desire and the possibility of bidirectional temporal relationships.
Limit: Small, narrow sample; no causal conclusion; desire and arousal were measured as occurrence at each assessment.
Open source record
DI-S08
Menstrual-cycle and noncyclical change
Kiesner et al. (2023), ‘Menstrual cycle-driven vs noncyclical daily changes in sexual desire.’ Daily reports from 213 female university students across two menstrual cycles.
Use: Supports a small average cycle association alongside large individual differences and noncyclical psychological associations.
Limit: Desire measured with two questions; relationship status and sexual orientation not assessed; cannot explain any individual’s pattern.
Open source record
DI-S09
Daily testosterone and desire
Catena et al. (2024), ‘Day-to-day associations between testosterone, sexual desire and courtship efforts in young men.’ Daily testosterone and self-report data from 41 men over one month.
Use: Supports rejecting a simple positive day-to-day testosterone–desire rule in eugonadal young men.
Limit: Small male sample; does not address clinical hormone deficiency, treatment, other hormones, or other populations.
Open source record
DI-S10
Learned cues and a narrow response measure
Hoffmann, Peterson, and Garner (2012), ‘Field conditioning of sexual arousal in humans.’ Field-conditioning experiment with 14 heterosexual couples; genital response assessed in men.
Use: Supports acknowledging that learned cues can affect a narrow sexual-response measure.
Limit: Measures genital arousal, not desire or identity; very small sample; must not be used to claim that orientation or identity is conditioned.
Open source record
DI-S11
Adult identity stability and change
Mock and Eibach (2012), ‘Stability and change in sexual orientation identity over a 10-year period in adulthood.’ MIDUS population sample of 2,560 adults.
Use: Supports coexistence of substantial sexual-orientation identity stability with documented adult change.
Limit: Baseline sample was overwhelmingly heterosexual; limited identity categories; two measurement points cannot reveal the process between them.
Open source record
DI-S12
Young-adult identity prevalence and stability
Savin-Williams, Joyner, and Rieger (2012), ‘Prevalence and stability of self-reported sexual orientation identity during young adulthood.’ Add Health data from 12,287 young adults approximately six years apart.
Use: Supports the claim that stability was more common than change while rates differed across sexual-orientation identity categories.
Limit: Measures self-reported identity rather than desire; historical labels and category structure affect findings.
Open source record
DI-S13
Identity, attraction, and behavior trajectories
Stewart et al. (2019), ‘Developmental patterns of sexual identity, romantic attraction, and sexual behavior among adolescents over three years.’ Three annual waves; 744 adolescents from three rural southeastern U.S. schools.
Use: Supports separate measurement of sexual identity, attraction, and behavior and documents both concordant and discordant trajectories.
Limit: Regional adolescent sample; attraction was romantic rather than specifically sexual; behavior collected as lifetime history at the final wave.
Open source record
DI-S14
Asexual identity, attraction, and desire
Su and Zheng (2023), ‘Stability and Change in Asexuality.’ Three annual waves; 168 participants; measures included identity, attraction, and solitary and dyadic desire.
Use: Supports treating asexual identity, attraction, and desire as related but nonidentical longitudinal variables.
Limit: Small sample and specific identity population; results cannot define all asexual experience or other identities.
Open source record
DI-S15
Sexual-minority identity continuity and change
Krueger et al. (2024), ‘Sexual Identity Continuity and Change in a U.S. National Probability Sample of Sexual Minority Adults.’ Two-year follow-up from the Generations national probability cohort.
Use: Supports the coexistence of sexual-orientation identity continuity and change among sexual-minority adults; 91% reported continuity and 9% a change over the study interval.
Limit: Excluded transgender participants, used selected age cohorts and racial or ethnic groups, and measured identity labels rather than desire or mechanism.
Open source record

Continue with adjacent distinctions

Behavior, culture, and evidence

Record details

Record ID: NS-KNOW-CHANGE-IDENTITY-001
Record type: Public explainer
Canonical URL: https://neurosexology.org/knowledge/change-and-identity/
Publication status: Published
Evidentiary status: Structured evidence synthesis
Review status: Evidence review current as of 2026-08-13
Version: 0.1
Framework version: 1.0
Author / architect: Nicholas Julian Madison
Related record IDs: NS-KNOW-BEHAVIOR-IDENTITY-001 · NS-KNOW-CULTURE-EMBODIMENT-001 · NS-GOV-EVIDENCE-001
Source and citation record: View record sources
Last substantively revised: August 13, 2026
Evidence last reviewed: August 13, 2026