A press conference. A politician at the podium. "Mistakes were made. The truth had some deficits. The people deserve better." By the third sentence, the speaker has rotated himself out of the agent position of every claim. Mistakes were made — by no one in particular. The truth had deficits — a property of the truth, not of anyone speaking it. The people deserve better — a sentiment, floating, with no one staking themselves to it.1 [POPULAR SOURCE]
Three sentences. Zero ownership. The grammar is doing the apologizing's work — and refusing it at the same time.
This is Lieberman's broader distancing-mechanism framework. Where the Pennebaker pronoun diagnostic reads which speaker is in the agent slot, the distancing framework reads how speakers get out of the agent slot when they need to. The two main routes: voice rotation (active to passive) and emotional outsourcing (cliché and metaphor borrowing).
Active voice puts the speaker in the chair. "I gave her the pen." Passive voice rotates the chair away. "The pen was given to her by me."1 Same fact, different angle of incidence with the speaker.
Children produce the rotation spontaneously. Younger sibling cries. Older sibling is asked what happened. "He fell." "She got hurt." "He banged his head." The grammar drops responsibility automatically because the older sibling does not want to hold it. Almost no child says "I pushed him into the wall, and he hit his head."1
Politicians produce the same construction at scale. Mistakes were made. The grammar is doing exactly what the political moment requires — appearing to acknowledge while declining to own. The contrasting honesty signature: "I made a mistake on your hem" from your tailor. Same admission. Active voice. Direct first-person. The tailor operates with greater integrity, and the grammar tells you so before he stops talking.1
The diagnostic note Lieberman flags: passive voice can also soften ill-received or confrontational messages without dishonesty. "We won the game!" and "The game was won" differ less in honesty than in celebration register. Active voice carries solidarity and pride; passive voice mutes them. So passive voice should not be read as deception per se, but as withdrawal from the agent position — which has multiple possible reasons, dishonesty being one of them.
Lieberman's four register-pairs for catching distancing through ornamental syntax:1
"I stand in awe" versus "I'm in awe."
"I find myself filled with pride" versus "I am so proud."
"I, for one, am glad" versus "I'm so glad."
"I am a great admirer" versus "I greatly admire."
The first phrasing in each pair is florid; the second is plain. Florid syntax is paradoxically less sincere because two structural give-aways:1
First, heightened emotional states produce simplified grammatical structures, not florid ones. "Help!" "I love you." Real emotion compresses syntax. When the syntax expands — clauses, observation-of-self constructions, register-elevation — the speaker is composing rather than feeling.
Second, the florid phrasing creates a separation between the I and the emotion. "I find myself filled with pride" puts the I in observer position, watching the I-being-filled. "I am so proud" collapses observer and feeler into one position. Distancing produces the gap.
The missing-wife pleas demonstrate the diagnostic in extreme form:1
STATEMENT A: "I'm so grateful that my wife was found alive. I'm indebted to all of the rescue workers."
STATEMENT B: "I, for one, am so grateful that my wife was found alive. I find myself indebted to all of the rescue workers."
Statement A reads as a man speaking. Statement B reads as a press release. The wedge — I, for one — and the observation construction — I find myself indebted — are exactly the florid moves the four register-pairs flag. Heard in an impromptu emotionally-charged situation, the florid version registers as inauthentic. Heard in a composed press appearance, the same version may simply be careful PR diction. Context governs the read.1
The second distancing route. A speaker trying to project an impassioned response can either generate the affective content from inside themselves or borrow it from a stock phrase. Manufacturing emotion takes mental energy. Borrowed phrases let the speaker hand the affective work over to a slogan and spectate while it does the lifting.1
Lieberman's diagnostic: ask any trauma victim about what happened, and you will not get a Nietzschean quote such as "To live is to suffer; to survive is to find some meaning in the suffering" or a cliché such as "That's the way the cookie crumbles."1 Real trauma narration is rough, ungrammatical, halting, peppered with sensory detail and broken syntax. Borrowed phrases are smooth and quotable. The smoothness is the give-away.
The amygdala-borrowing tell: anyone who tells you a traumatic experience is "indelibly in my amygdala" is performing trauma rather than describing it. The phrase is too composed for the affective state it claims to describe. Emotional congruence — Lieberman's phrase — requires the linguistic register to match the affective intensity. Florid syntax over real emotion produces incongruence; plain syntax over claimed emotion can also produce incongruence in the opposite direction (a polished speaker performing flat affect over claimed grief).1
The empirical anchor Lieberman cites is Whelan, Wagstaff, and Wheatcroft (2014) on real-life public appeals for help with missing relatives. Genuine pleas contained more verbal expressions of hope, more positive emotion toward the relative, and avoidance of brutal/harsh language.1 Faked pleas were loaded with mottos and slogans peppered with negativity. The pattern survives forensic analysis at corpus scale.
The press appearance read. A father stands at a press conference asking for information about his missing daughter. Listen for compressed first-person urgency vs florid observation-syntax. "Please, anyone who has seen Sarah, call the number" tracks differently from "I find myself reaching out to the public, hoping that someone, somewhere, has the information that will bring our beloved daughter home." The florid second version may simply be a man who is composed under stress; it may also be the linguistic signature of someone who has had time to rehearse the appeal. The Lieberman read is probabilistic — the diagnostic shifts confidence, it does not produce conviction.
The corporate apology audit. Friday afternoon. The CEO publishes a statement about a recent failure. "Mistakes were made. The truth had some deficits in our communication. The community deserves better." All three sentences are passive-or-floating constructions. None of them holds an agent. Compare with "I made the call to delay the announcement, and that was wrong" — same admission, opposite distancing posture. The apology that owns reads as integrity; the apology that distances reads as legal hedging dressed in regret.
The trauma-narration check. Someone is telling you about a traumatic experience. Listen for borrowed phrases. "What doesn't kill you makes you stronger." "Time heals all wounds." "That's just life." Each borrowed phrase is a moment where the speaker has handed off affective content to the cliché. This does not mean the trauma is fake. It means the speaker is at a distance from the affect right now — possibly because dissociation is doing protective work, possibly because the speaker has rehearsed the narration so many times the original affect has been replaced with a stable verbal artifact. Either way: the borrowed-phrase density is data about current state, not about whether the underlying event was real.
Evidence:
[POPULAR SOURCE]; clinical-practitioner synthesis of the underlying linguistic-pragmatics research.Tensions:
Professional register confounds. Lawyers, diplomats, executives, and PR-trained speakers produce florid distancing as professional baseline. Reading their press-conference speech against a personal-conversation baseline produces systematic false positives. The framework requires same-register comparison — does this speaker's florid construction exceed the baseline for their professional context, or merely match it?
Time-since-event compounds. Florid syntax is most diagnostic when applied to recent, raw emotional content. The same speaker, six months later, may legitimately produce "I find myself reflecting on what we lost that day" without distancing — the cognitive distance between the speaker and the event has had time to develop. Reading old grief through the same lens as fresh grief produces error.
Cliché borrowing as cultural register. Some communities produce high cliché-density as ordinary speech — religious communities with shared scripture, fan communities with shared catchphrases, professional communities with stock phrases. Reading any cliché use as outsourced-emotion misreads cultural-register users.
Open Questions:
Pennebaker's research on the LIWC instrument provides the empirical floor. Whelan, Wagstaff, and Wheatcroft (2014) at the University of Liverpool ran the missing-relative pleas study that Lieberman cites — coding genuine vs faked pleas across forensic outcome categories. The research is methodologically careful: large samples, downstream verification.
Lieberman's contribution is the integrated four-register-pair demonstration plus the cliché-as-outsourced-emotion translation. He compresses what Pennebaker handled at LIWC corpus scale into recognizable grammatical patterns a civilian can spot in real time. The cost of compression is the same one running through Mindreader: research-register findings get presented in operational form that sometimes overstates single-instance reliability. The diagnostic is statistical. Single-utterance reads should produce probability shifts, not classifications.
Behavioral Mechanics — Pennebaker Pronoun Diagnostic Framework: Pennebaker Pronoun Diagnostic Framework is the sister page on the pronoun-as-ownership primitive. The Linguistic Distancing page covers the voice and cliché primitives that the pronoun framework alone does not capture. Read together: pronouns tell you who is in the agent slot; voice rotation tells you whether the agent slot has been preserved or rotated out; cliché-density tells you whether the affective content is coming from inside the speaker or from a stock phrase. The three primitives form a unified surface-distancing diagnostic. The structural insight neither generates alone: distancing is multi-channel. A skilled speaker can produce confident pronouns while running passive voice, or active voice while running cliché-heavy affect. Reading any single primitive misses the integrated distancing pattern. The combined three-channel read is what catches the politician's mistakes were made register, the rehearsed grief at the press conference, and the trauma-narration that has had its rough edges sanded off by repetition.
Psychology — Dissociation and Cognitive Freeze: Dissociation and Cognitive Freeze documents the autonomic substrate of dissociative responses to overwhelming threat. The cliché-borrowing route to distancing has its psychological substrate in this territory. A trauma survivor producing high cliché-density during narration is not performing inauthenticity — they are running a protective mechanism that has substituted stable verbal artifacts for raw affective content. The Lieberman framework reads the surface; the dissociation framework explains why the surface is what it is. Read together, the two pages produce a clinically usable distinction: the deceptive florid speaker uses borrowed phrases to manufacture affect they don't feel; the dissociative florid speaker uses borrowed phrases to manage affect they cannot bear to feel directly. The behavioral surface looks similar; the underlying state is opposite. The framework's correct deployment requires holding both possibilities simultaneously rather than collapsing to the deception-only read.
The Sharpest Implication
The active-voice / passive-voice rotation is one of the most robust honesty signals available, and it is also one of the most easily faked once the speaker knows it is being read. Politicians and PR professionals have known this for decades. They produce active voice for ownership claims that benefit them and passive voice for liability claims that don't. The honest individual who has not learned to manage their grammar produces the same primitive unconsciously — passive voice when distancing, active voice when committing. This means the framework is increasingly most reliable against untrained speakers and increasingly less reliable against trained ones. Public-figure speech becomes harder to read with each additional decade of media training. Civilian speech remains relatively transparent.
The corollary the cliché diagnostic forces: anyone trying to convey deep emotion through borrowed phrases has already declined to convey deep emotion through original ones. The decision was made before the speech began. The reader of cliché-heavy speech is therefore not reading the underlying affect — the reader is reading a verbal artifact the speaker constructed to substitute for the affect. The artifact may be honest about what the speaker can produce in this moment; it is not honest about what the underlying state is.
Generative Questions