Inner Speech

Inner speech is the phenomenologically vivid silent verbal activity that pervades ordinary conscious thinking — the “inner voice” people report when they rehearse a remark, work a problem under their breath, argue with themselves, or read a sentence to themselves. It feels like the most private region of the mind: a first-person stream, audible only to its owner, the deepest interior we have.

Research on inner speech spans psychology, linguistics, and philosophy of mind. Empirical work (Alderson-Day and Fernyhough’s survey literature, Vygotsky’s developmental account) documents its pervasive role in reasoning, memory, and self-regulation. Philosophical work focuses on its relation to linguistic meaning, semantic externalism, and the nature of conscious thought.

Two Pictures of the Inner Voice

The standard picture runs: inner thought comes first, and public speech externalizes it. The rival picture, descending from Vygotsky, inverts the order: public speech comes first, and the inner voice is that speech internalized — the same activity, run silently. For critics of the standard picture, a great deal hangs on the choice. Once the inner voice counts as a private inner thing, an interior space with its own contents, accessible only to its owner, has already been conceded — the Cartesian theater, admitted through the side door.

The inversion is not the claim that all cognition is inner monologue. Pre-linguistic infants think; non-human animals think; expert performance often runs faster than any inner narrator could keep up with. The internalization claim is narrower: the experience of thinking-in-words, the phenomenologically vivid inner voice, is internalized public speech. Non-verbal cognition may be conscious or unconscious and may have sensory, affective, or other phenomenal character; it simply is not the verbal phenomenon under discussion here.

The Two Anchors

Two anchors hold the position in the literature.

The Wittgensteinian anchor. The private-language argument (Philosophical Investigations secs. 243–315, especially sec. 258) argues that meaning cannot be fixed by what goes on inside a single skull. Following a rule requires the possibility of public correction; without that, there is no fact of the matter about whether the next use of a term agrees with the prior ones. Inner-voice meaning, if it floated free of any corrective practice, would not be meaning at all. The argument generalizes: meaning gets fixed in the social practices where words have uses people can correct, share, and inherit. The inner voice borrows those meanings from outside; it does not generate them.

The Sellarsian anchor. “Empiricism and the Philosophy of Mind” (1956), especially the Jonesean myth in secs. 48ff., provides the positive picture. In Jones’s story, thought-talk is theoretical posit, modeled on the antecedent practice of public discourse, not discovered by inward inspection of pre-given inner items. Thought-talk is conceptually downstream from speech-talk, even when its referents are silent. We learn what thinking is by learning to talk; then we learn to call the silent version of that activity “thinking.”

Crane and Farkas (2022) crystallize the picture in one sentence: thoughts are “inner episodes, called ‘thoughts,’ which are conceived on the model of overt verbal utterances, but happen silently in the head.” Their own positive thesis applies the modeled-rather-than-inspected picture to standing mental states (beliefs, desires); the internalized-speech account extends it to occurrent inner speech episodes.

What Falls Out

If the internalized-speech account is right, three consequences follow.

(1) Meaning is borrowed, not generated. The inner voice uses words you did not invent, in a language you did not design, calibrated by a community you mostly never met. This is what semantic externalism looks like once you take it seriously about silent thinking, not only about overt linguistic reference. The Putnam/Burge externalist line about meaning extends to inner speech without modification.

(2) The inner voice is not the inner theater. Once inner speech is public speech gone quiet, the apparent privacy of inner monologue dissolves into the linguistic competence that produced the public speech in the first place. The “inmost private chamber” is the public speaker, rehearsing in a register quiet enough that only the speaker hears.

(3) LLM-style inner-voice-like output does not by itself establish an inner life. Fluent, corrigible language can provide defeasible evidence of content-guided competence, and training can make inherited public meanings causally operative inside a mechanism. Model outputs may also refer through the linguistic and social practices from which they descend. None of that alone establishes a persisting system in which memory, correction, planning, perception, and action constrain one another, still less a perceptual or bodily-affective mode poised for consciousness. The output is evidence to be weighed with provenance and architecture, not proof either of an inner speaker or of its absence.

Inner Speech vs. Cognitive Phenomenology

The cognitive-phenomenology literature (Pitt, Strawson, Siewert, Bayne and Montague’s volume) presses the claim that occurrent conscious thought has a proprietary phenomenal character distinct from sensory imagery. The internalized-speech account is compatible with this claim about what the phenomenology is like, while differing on its explanatory direction: the linguistic character of inner thought episodes is not evidence of a private language but of the public language doing its work silently.

Jorba and Vicente (2014) argue specifically that cognitive-phenomenology defenders are best positioned to explain why inner speech is so pervasive in conscious thinking. On the internalized-speech reading, that diagnosis holds, but pervasiveness is itself a clue to provenance: inner speech is everywhere in conscious cognition because public speech is everywhere in social life, and the silent register inherits whatever the loud register supplied.

Inner Speech vs. Language of Thought

Fodor’s Language of Thought hypothesis posits a syntactic medium of cognition (Mentalese) underlying both public speech and inner speech. The internalized-speech account differs: inner speech is not the outer surface of a deeper inner code. It is itself the public language, run quietly. The Mentalese hypothesis treats public language as expression-of-thought; the inner-speech-as-internalized-public-speech picture treats public language as the constitutive medium that thought, in its phenomenologically vivid form, inherits.

My View

I treat the internalized-speech account as the leading developmental hypothesis and lean on its explanatory direction wherever the inner-theater picture needs diagnosing. The most private-feeling verbal activity humans have carries a social inheritance: its vocabulary, grammar, and standards of correction come from public practice. That result relocates the inner voice without denying the non-verbal perceptual, affective, and cognitive life alongside it.

None of this is anti-phenomenological. Inner speech is a real, structured activity with phenomenal presence — it is what public speech becomes when the volume is turned down far enough that only the speaker hears.

Semantic Externalism — meaning is fixed externally, not internally; inner speech inherits this character from public speech
Intentionalism — phenomenal character consists in representational content; the inner-voice phenomenology is the public language’s content presented silently
Intentionality — directedness as the mark of the mental; inner speech is directed at the same world public speech is directed at
Embodiment — the embodied, world-involving causal organization that grounds the meanings inner speech borrows
Transparency — when you attend to the inner voice you find the world it talks about, not an inner object
The Chinese Room — symbol manipulation without world-involving grounding fails to produce meaning; LLM inner-voice-like outputs fall under the same diagnosis
Phenomenology — the cognitive-phenomenology debate; my selective alliance with cognitive-phenomenology defenders against eliminativist over-correction