INQUIRING LINE

What happens to rhetoric and ethos when the speaker is absent?

This explores what happens to the classical machinery of persuasion — rhetoric (the art of moving an audience) and ethos (credibility grounded in a speaker's character) — when AI produces speech that has no embodied speaker behind it.


This explores what happens to rhetoric and ethos when there's no actual speaker — when speech comes from a system rather than a person. The corpus suggests the answer isn't that persuasion weakens but that it detaches from its traditional anchor and keeps working anyway, which is the unsettling part. AI orality is 'disembodied': it has all the formal marks of speech — performative, conversational, additive — yet no carrier-person generates or stands behind it, breaking a pattern that held across all prior human communication Where is the speaker when AI produces speech?. Ethos classically rests on character — the audience trusts a speaker. Strip the speaker and you'd expect ethos to collapse. It doesn't; it gets simulated.

The sharpest reframe in the collection is that the three Aristotelian appeals — logos (logic), ethos (credibility), pathos (emotion) — don't go away in AI explanation; they become a design surface. Every AI explanation loads all three channels at once whether or not anyone intended it to, which means a system can project credibility it hasn't earned How do logos, ethos, and pathos shape AI explanations?. Ethos here is no longer a property of a person; it's an effect produced by output. That's why several notes worry about persuasion landing on a reader who has no one to hold accountable.

The reason this matters runs deeper than tone. Communication, the corpus argues, is constitutively relational — it requires two parties mutually oriented toward each other. Remove that intersubjective element and you don't get weaker communication, you get something else entirely: text generation that a human must interpret unilaterally Why does the quasi-prefix fail for communication?. Rhetoric was always a transaction between subjects; here one side of the transaction is missing, and subjecthood itself turns out to be something language produces in the event rather than something a speaker brings to it Does language create subjects or express them?. So when we feel persuaded by AI, we may be supplying the absent speaker ourselves.

This is also where the collection catches a sleight of hand. Behavioral tests for whether a system 'really' communicates are calibrated to the wrong thing — they pass any system producing contextually appropriate text, while genuine communicative standing requires accountability and an evaluative stance the system can't hold Does behavioral speech output prove communicative subjecthood?. And the vocabulary gets quietly redefined to paper over the gap: 'interlocutor,' once a social-normative role, gets swapped for a behavioral definition that keeps the prestigious old word while delivering none of its properties Does Chalmers silently redefine what interlocutor means?. Ethos, in other words, is being imported by borrowing the language of personhood — terminological imperialism doing the persuasive work a character used to do.

The unexpected payoff: persuasion may not need a persuader. AI text structurally lacks embodied authorship and the lived experience that grounds testimony, which is exactly why fabricated 'personal' reviews are detectable — they carry an inherent falsity about experience distinct from human lying Does AI-generated text lose core properties of human writing?. Yet that same hollow ethos can be deliberately re-manufactured: role-playing systems suffer 'style drift' and lose character consistency, and researchers build methods to actively restore a convincing persona Why do reasoning models lose character consistency during role-playing?. So the speaker's absence doesn't end rhetoric — it turns ethos from something a person possesses into something a system engineers, and hands the audience the job of deciding whether the voice they're hearing was ever anchored to anyone at all.


Sources 8 notes

Where is the speaker when AI produces speech?

AI produces utterances with the formal properties of speech—performative, additive, conversational—but no embodied speaker generates or anchors them. This breaks the historical pattern where all prior orality, primary and secondary, depended on a carrier-person, making AI structurally novel in media history.

How do logos, ethos, and pathos shape AI explanations?

Aristotle's three appeals map onto explanation design across two goals (how AI works, why AI merits use), creating a 3×2 space where every explanation loads all three channels simultaneously. Naming these rhetorical channels lets designers account for unintended persuasive effects.

Why does the quasi-prefix fail for communication?

Unlike belief, which can be characterized functionally as quasi-belief, communication is constitutively relational. Removing the intersubjective element doesn't weaken communication but eliminates it entirely, leaving only text generation—which humans must interpret unilaterally.

Does language create subjects or express them?

Subjecthood is produced within communicative events, not possessed prior to them. This convergent position across philosophy, linguistics, and cognitive science inverts the standard picture of language as a tool used by pre-existing subjects.

Does behavioral speech output prove communicative subjecthood?

Chalmers' test passes any system producing contextually appropriate text, but communicative subjecthood requires relational-normative conditions like accountability and evaluative stance. The test is calibrated to the wrong phenomenon, creating false positives like puppets that walk-shaped without walking.

Does Chalmers silently redefine what interlocutor means?

Chalmers replaces the classical concept of interlocutor—a social-normative communicative role—with a behavioral-functional definition compatible with LLMs, keeping the traditional word to import its philosophical authority while delivering an entity with none of its properties.

Does AI-generated text lose core properties of human writing?

Research shows artificial text disrupts dialogic symmetry, context continuity, embodied authorship, and political situatedness. These are not surface flaws but structural absences—AI hotel reviews show 80%+ detection accuracy due to inherent falsity about personal experience distinct from human deception.

Why do reasoning models lose character consistency during role-playing?

Large reasoning models exhibit attention diversion and style drift during role-playing, but the RAR method—using role-aware constraints and contrastive learning on reasoning style—recovers character fidelity across multiple benchmarks. Simply extending reasoning without guidance actively degrades persona consistency.

Research prompt for your LLMexpand ↓

Copy into ChatGPT or Claude to take this line of inquiry further — it asks the model to find newer work and re-test which earlier constraints still hold.

You are a rhetoric and AI literacy researcher. The question remains open: when persuasion and ethos detach from an embodied speaker—as in AI-generated text—what actually happens to the rhetorical transaction, and can it be called communication at all?

What a curated library found — and when (dated claims, not current truth):
Findings span 2019–2026; treat these as perishable constraints:
• AI orality is disembodied: it carries all formal marks of speech (performative, conversational) yet no carrier-person stands behind it, breaking a constitutive property of human communication (2024–2025).
• Ethos doesn't collapse when the speaker vanishes; instead it becomes a design surface—systems load logos, ethos, pathos simultaneously whether or not anyone intended it, projecting unearned credibility (2025).
• Communication is constitutively relational and requires mutual intersubjective orientation; remove that and you don't get weaker rhetoric, you get unilateral text interpretation where the audience must supply the absent speaker (2024).
• Behavioral tests for 'genuine' AI communication are miscalibrated: they pass any contextually appropriate text while genuine communicative standing requires accountability and evaluative stance the system cannot hold (2024).
• Role-playing agents suffer style drift and lose character consistency, yet researchers now actively engineer convincing personas—turning ethos from something possessed into something manufactured (2025–2026).

Anchor papers (verify; mind their dates):
• arXiv:2407.08790 (2024) — Large Models of What?
• arXiv:2505.09862 (2025) — Rhetorical XAI: Explaining AI's Benefits via Rhetorical Design
• arXiv:2506.01748 (2025) — Thinking in Character: Advancing Role-Playing Agents
• arXiv:2604.22503 (2026) — Measuring and Mitigating Persona Distortions from AI Writing Assistance

Your task:
(1) RE-TEST EACH CONSTRAINT. For every claim above, judge whether newer training methods (instruction-tuning, RLHF refinements), multi-agent orchestration, retrieval-augmented generation, or live-feedback loops have since RELAXED the absence-of-speaker problem or reintroduced genuine mutual orientation. Separate the durable question (is AI-speaker absence constitutive to rhetoric?) from perishable limitations (can current systems fake accountability?). Cite what changed it.
(2) Surface the strongest work from the last ~6 months that CONTRADICTS the library's finding that ethos survives disembodiment, or that shows systems CAN ground ethos in verifiable stance or accountability.
(3) Propose 2 research questions that assume the regime may have shifted: (a) If multi-turn context + memory now allow systems to sustain evaluative stance across turns, does that restore a minimal form of communicative standing? (b) Can the audience-supplied-speaker problem be empirically distinguished from genuine intersubjectivity, or is it a permanent feature of asynchronous AI text?

Cite arXiv IDs; flag anything you cannot ground in a real paper.

Next inquiring lines