Daniel Gregory: Introduction to the Symposium on Inner Speech

Introduction to the Symposium

Daniel Gregory

Introduction

Produce this sentence in inner speech (i.e., in your ‘inner voice’ or your ‘silent monologue’, as it is also sometimes known): ‘I think, therefore I am’. At the moment you produced that sentence in your mind, was it true?

              Setting aside the complications which philosophers have raised about every word in the sentence, the default answer is ‘yes’. Notice, though, that you could only have uttered a true sentence in inner speech if producing inner speech really does amount to speaking in your mind. And it turns out that inner speech might not really be a kind of speech at all.

              On the face of it, inner speech consists of auditory mental imagery. It involves conscious auditory experiences which are not caused by sounds in the external world. But auditory images are usually representations of sounds—as when you imagine the sound of cymbals crashing or a lion roaring. So, inner speech might just be a representation of the sound of speech, rather than a kind of speech. If this is so, then producing the sentence, ‘I think, therefore I am’, in inner speech cannot amount to saying something true at all. All you are doing is imagining what the sentence sounds like when spoken aloud. An auditory imagining can represent a sound more or less accurately, but it cannot be true or false. Imaginings are simply not truth-evaluable.

              Perhaps there are other ways that you can judge or affirm Descartes’ famous sentence, but producing it in inner speech is surely what most of us did when we first encountered it as undergraduates—and every time since then. But, if inner speech is not a kind of speech, then we might never have spoken in our minds a sentence which was supposed to serve as a self-verifying rebuttal of external world skepticism. In fact, we might never have spoken in our minds a sentence which was apt for assessment as true or false at all!

*                         *                         *

Inner speech did not receive a lot of attention from analytic philosophers historically, but a small industry has developed around the topic in the last 15 or so years. This symposium will give a sense of the issues about inner speech which philosophers have started discussing.

              Peter Langland-Hassan takes up a version of the issue raised above. He reflects on whether there really are words in inner speech, or just auditory images of words. He discusses what he calls the ‘double-duty view’ (discussed in Gregory (2026, pp. 175–176), referencing Daniel Stoljar, in conversation). A photograph of the famous Hollywood sign is an (external) image of a particular token of the word, Hollywood. Specifically, it is an image of the token consisting of those steel structures on Mount Lee. At the same time, there is a token of the word, Hollywood, on the photograph paper. The patches of ink which represent the sign on Mount Lee themselves form an instance of the word, Hollywood. So, there is a token of a word which also represents a distinct token of that word. Could inner speech consist of auditory mental images of words which also have all of the properties necessary to be tokens of words? If so, then inner speech would involve imagistic representations of speech, but it would also simultaneously be a kind of speech itself. (Producing the sentence, ‘I think, therefore I am’, in inner speech would amount to imagining the sound of the sentence and actually saying the sentence.) Langland-Hassan reviews various disanalogies between inner speech and the case of the photograph of the Hollywood sign, drawing the double-duty strategy into question, but leaves open whether it might survive on a functional analysis of words.

              In contrast to Langland-Hassan, Wade Munroe argues that inner speech and external speech are fundamentally continuous phenomena—both just different kinds of speech. He then urges that theorists of speech and language have gone badly astray in focusing almost entirely on interpersonal speech and interpersonal uses of language. Munroe quotes Chomsky: ‘inner speech is most of speech. Almost all the use of language is to oneself’ (Chomsky 2002, p. 148). So, speech and language cannot be understood without taking inner speech into account—and perhaps giving it a central role.

               One point Munroe makes is that traditional speech act theories might not be able to accommodate inner speech. Speech act theories typically attribute major significance to the intentions with which speakers produce utterances. Appealing to work in cognitive science, Munroe argues that we can neither plan our inner speech in advance of producing it, nor guide it while we are producing it, in the way that we can plan and guide our external speech. If it is an essential feature of actions that we have the ability to plan and guide them, then instances of inner speech cannot be actions—and therefore cannot be speech acts.

              However, insofar as we do seem to perform speech acts in inner speech—we seem to give (ourselves) instructions and ask (ourselves) questions etc.—speech act theorists may have made a major error in emphasizing intentions in their analysis of language use.

              Marta Jorba takes up the topic of inner speech and self-knowledge, i.e., knowledge of our own mental states. It is a tempting thought that inner speech is a source of knowledge about what we believe and desire and so on, but it is hard to see how. If one has knowledge of the intention to express, e.g., a belief in inner speech, then it seems they cannot learn that they hold this belief by expressing it. On the other hand, if one does not know that they hold a belief, then it seems they cannot formulate the intention necessary to produce the inner speech expressing it. Jorba proposes that, while we can often have knowledge of our mental states without expressing them in inner speech, we can gain a further, distinctive kind of knowledge of those mental states by formulating linguistic expressions of them. Formulating a linguistic expression of a thought in inner speech also enables it to play different functional roles.

              In my contribution, I summarize some ideas from a recent paper on inner speech in sign languages (Gregory 2026). For speakers of voiced languages (e.g., English or Hindi), the experience of inner speech is typically sound-like. They have what we can call ‘auditory inner speech’ (or ‘auditory-IS’). It is natural for people with auditory-IS to assume that speech in a signed language (or ‘signed-IS’) is effectively the same kind of mental state as auditory-IS, except that it occurs in the visual modality. But this is not a safe assumption. There is a surprising dearth of empirical research on the phenomenology of signed-IS, but we can infer from the nature of sign languages that signed-IS may have some complexity which auditory-IS lacks. For example, the experience of signed speech is perspectival to a far greater extent than the experience of voiced speech. What a signer sees (their own hands moving from behind etc.) is different from what their interlocutors see (the signer’s hands moving from in front etc). So is signed-IS like seeing signing from behind (the signer’s perspective) or from in front (an interlocutor’s perspective)? There is a third possibility as well. Joanna Atkinson has suggested that signed-IS might have a largely kinaesthetic phenomenology, given that the experience of producing signs is one of making bodily movements (Atkinson 2006, p. 706). In the paper, I argue that, whatever the phenomenology of signed-IS, it will complicate at least some claim which has been made about inner speech. In my contribution to this symposium, I give one example, in the context of inferentialism about self-knowledge.

*                         *                         *

It is not obvious why analytic philosophers did not seem to have much interest in inner speech until relatively recently. It is a huge part of the mental lives, and the language use (if that is what it is), of most people. It also connects with some of the biggest topics in the field, such as speech act theory and self-knowledge. Whatever the reason, though, I hope that this symposium will convince readers that inner speech is something very much worth investigating now.

Atkinson, J. (2006), ‘The perceptual characteristics of voice-hallucinations in deaf people: Insights into the nature of subvocal thought and sensory feedback loops’, Schizophrenia Bulletin, 43: 701–708.

Chomsky, N. (2002), On Nature and Language, Cambridge University Press.

Gregory, D. (2026), ‘Inner speech and sign languages’, Synthese, 207: 176.