Two husbands sit in a counselor's office. The counselor asks each one how things are going. Husband A says: "Our marriage is in trouble." Husband B says: "The marriage is in trouble."1 [POPULAR SOURCE]
Same content. Different sentence. Different marriage.
Husband A is in the marriage even while reporting the trouble — our, we, the bond is still claimed even in failure. Husband B has rotated the marriage out of personal possession entirely. The marriage is now an entity outside of himself, the way one might refer to the building or the project. He has not yet filed for divorce. Linguistically, he has already left.
This is the diagnostic core of Lieberman's we-pronoun framework. Cooperative pronouns (we, our, us) signal that the speaker is psychologically inside the relationship. Individualized pronouns (I, me, you, your) signal partial or full exit. The diagnostic runs across marriages, families, workplaces, sports fan loyalties, and the sharpest forensic application of all — crime victim narration.1
The framework's most dramatic empirical finding comes from law enforcement. Victims of violent crimes — abduction, assault — rarely use the word "we" when describing the events.1
A genuine kidnapping victim does not say "we got into the car" or "we stopped for gas." The genuine victim says "he put me in the car" and "he stopped for gas." The grammar separates the victim from the aggressor. He did things; me had things done to me. The pronouns mark the absence of relationship between the two parties.1
When investigators see we peppered through a crime narration, the diagnostic flag fires. We got into the car implies cooperation, association, or relationship. In a genuine kidnapping, no such relationship exists. The grammar is incompatible with the underlying event-structure. Either the victim and aggressor had a prior relationship that the report is concealing, or the report is itself a fabrication.
The forensic anchor is not metaphorical. Pronoun-based statement analysis is part of standard interview protocols in police departments and prosecutorial offices. The Sapir SCAN methodology (cited later in the chapter) is contested in academic forensic linguistics, but the underlying we-vs-I observation has more robust support — cooperative pronouns reliably correlate with cooperative relationship-states.
The forensic primitive scales directly to relationship analysis. Married couples who use cooperative language (we, our, us) more often than individualized language (I, me, you) have lower divorce rates and report greater marital satisfaction.1
The pronoun-shift signature in conflict resolution is even more diagnostic:1
"You need to figure this out" — me versus you. Enmity. Polarization.
"We need to figure this out" — us versus the problem. Cooperation. Shared responsibility.
The same disagreement, two different relational postures. The first frames the partner as the obstacle. The second frames the obstacle as something the partnership is facing together. Couples that habitually run the second register survive disagreements; couples that run the first accumulate enmity until the relational structure fails.
The diagnostic extends to family-of-origin language:1
The diagnostic note is critical: any single utterance can simply reflect momentary frustration, not relationship state. "Your son just spilled juice on the rug" is a single annoyed parent venting, not a marriage breakdown. "Your son did this. Your daughter did that. Your mother is calling. Your friends are coming over." — sustained across topics, weeks, contexts — is the signature. Lieberman's repeated rule: a single, casual reference does not mean anything, but a consistent pattern of syntax reveals everything.1
The framework has one important inversion the forensic-origin section flags by absence. Stockholm syndrome — when victims develop psychological bonds with their captors — produces we-pronoun use in genuine victims. The victim under sustained captivity who has begun to identify with the captor will produce we got into the car not because the report is false but because the underlying relational state has shifted under coercive bonding.
This means the we-in-crime-narration diagnostic produces false positives against Stockholm-affected victims. The framework's correct deployment requires considering capture duration and prior signs of bonding before treating cooperative-pronoun use as evidence of fabrication.
The same primitive scales to non-romantic relationships.1
Workplace. Firms where workers refer to their workplace as "the company" or "that company" rather than "my company" or "our company," and to coworkers as "they" rather than "my coworkers," are likely to have low morale and high turnover.1 The pronoun signature predicts the affective state. The departing employee has often pronoun-departed months before they file resignation paperwork.
Sports fans. The fair-weather fan is identified through pronoun-shift. "We won" when the team wins; "They lost" when the team loses. Genuine identification with the team carries through the loss. The fair-weather fan retreats to they the moment the affiliation costs something.
The marriage check-in. Friday dinner conversation with friends about their week. The spouse describes a difficult parenting situation. "Our daughter had a really hard time at school today." Five minutes later: "My son's teacher emailed me about his behavior." Notice the pronoun-shift across topics. Our daughter (positive context); my son's teacher (negative context, with disownership of the spouse from the email-recipient role). The shift is the diagnostic. The spouse is producing the I-vs-our primitive across the conversation. Neither utterance alone is conclusive. The pattern across the dinner is.
The job interview red flag. A candidate describes their last role. "The team had a really difficult quarter." "Their leadership made some questionable choices." "The company didn't really know where it was going." Three sentences. Zero cooperative pronouns. The candidate has linguistically exited the prior employer well before the interview. Reading: this candidate joined the company, became a fan, then became disengaged, then began searching. The interview is the formalization of an exit that happened linguistically months ago.
The sports-fan disclosure. A friend talks about their team's season. "We had such a great run in the first half — I really thought we were going to make it." Then later: "They've just been awful since the trade." The pronoun-shift fires. Either the friend has accepted that the team's current iteration is no longer us (a real reorganization-driven distancing), or the friend is producing fair-weather signature (loss-driven distancing). Distinguishing the two requires knowing what changed.
Evidence:
[POPULAR SOURCE].Tensions:
Cultural register confounds. Some languages and cultures encode relational pronouns differently. East Asian honorific systems may not produce the same we-vs-I primitives in the same way English does. Reading the framework cross-culturally without baseline-recalibration produces systematic error.
Workplace cultural baseline. Some corporate cultures aggressively train employees to use the company / the team in formal communication. Reading those speakers as disengaged misses the cultural-register confound.
Stockholm and trauma-bonding inversions. As flagged, victims under coercive bonding produce we genuinely. The framework misfires against this population.
Single-utterance reading. The cardinal misuse. Your son spilled juice is venting, not a divorce signal. The framework requires sustained-pattern observation.
Open Questions:
The framework's empirical anchors are scattered across multiple research streams. Seider, Hirschberger, Nelson, and Levenson's research on relational pronouns in marital communication is the foundational study Lieberman cites. The forensic-application research is largely in police-training literature rather than peer-reviewed academic work — making the forensic diagnostic less academically validated than the marital-communication finding.
Lieberman's contribution is the integrated diagnostic that shows the same primitive (cooperative-vs-individualized pronouns) operating across forensic, marital, family, workplace, and sports-fan domains. The cross-domain replication of the same primitive is itself evidence the underlying construct is robust. The cost: the academic literature is uneven across the domains. The marital research is solid; the forensic research is operationally accepted but academically contested; the workplace and sports-fan applications are largely anecdotal in the popular literature.
Behavioral Mechanics — Pennebaker Pronoun Diagnostic Framework: Pennebaker Pronoun Diagnostic Framework runs the broader function-word ownership diagnostic. The We-Pronoun framework focuses specifically on the first-person plural primitive that the Pennebaker framework treats as one element among many. Read together: Pennebaker tells you whether the speaker is in the agent slot at all; we-pronoun tells you whether the speaker locates themselves inside or outside the relationships they describe. The two diagnostics work as complementary axes — the speaker who runs high I-density (Pennebaker engagement signal) but low we-density (relational disengagement signal) is showing the linguistic profile of someone fully present to themselves but not to their partnerships. The combined two-axis read produces tighter inference than either alone — and surfaces the specific case (engaged-self-disengaged-relationships) that neither axis alone captures.
Behavioral Mechanics — Metanoia Grammar and Pronoun Architecture: Metanoia Grammar and Pronoun Architecture documents the they/we/I shift as the linguistic spine of propaganda construction. The we-pronoun-as-relationship-diagnostic page operates the same primitive at micro-relational scale that propaganda operates at mass scale. A movement constructs a we by naming a they in opposition; a marriage constructs a we by naming the we-couple-vs-the-problem in opposition. The same primitive — collective-pronoun construction as identity-architecture — operates across scales. The structural insight neither generates alone: the linguistic mechanism by which propaganda dissolves the I into the we is the same mechanism by which a healthy marriage dissolves two Is into a shared we. The mechanism is value-neutral. Whether it produces freedom or coercion depends on whether the we is voluntarily entered and continues to be voluntarily maintained. Reading the same mechanism across the two scales — micro-relational and mass-political — surfaces the deeper claim: pronoun-architecture is identity-architecture, and identity-architecture is exactly what most reliably distinguishes consenting cooperation from coercive merger.
The Sharpest Implication
The forensic-origin finding (crime victims rarely use we) is the framework's strongest empirical claim and also its most dangerous to operationalize. Police investigators trained on the diagnostic may treat we-density in crime narration as evidence of fabrication, missing the Stockholm-affected victims who produce we genuinely under coercive bonding. The framework therefore requires explicit teaching of the inversion alongside the primary finding — and the inversion is structurally less memorable than the primary finding. The popular reader who absorbs "crime victims don't say we" without the Stockholm caveat will produce false-positive accusations against the most vulnerable victims.
The marriage corollary is gentler but operationally significant: the pronoun-pattern across a relationship can be observed by either partner with relatively low effort. The partner who notices their own shift from our to the in describing the marriage has surfaced data their conscious deliberation has not yet produced. The shift often precedes the conscious recognition that the relationship is in real trouble by months or years. Self-observation of pronoun-pattern is therefore one of the cheapest forms of relational early-warning available — and it works on yourself more reliably than on your partner, because you have continuous access to your own speech in a way you do not have to theirs.
Generative Questions