Signed Inner Speech
Daniel Gregory
***What follows summarizes some ideas in my recent paper, ‘Inner speech and sign languages’ (Gregory 2026).***
When native speakers of voiced languages (such as English, Mandarin or Hindi) think about inner speech, they naturally tend to think of a mental state with an auditory phenomenology—something like an internal correlate of speech spoken aloud. But what about the inner speech of people whose native language is a sign language (such as American Sign Language, Indo-Pakistani Sign Language or Russian Sign Language)? What do we know about their inner speech?
The answer is that we know surprisingly little. Inner speech can occur in signed languages (see Lœvenbruck et al. 2018, pp. 139–140, reviewing some relevant sources), but it seems that there are no published data about the phenomenology of signed inner speech (or signed-IS). There are, of course, anecdotal reports, such as those here and here, but, as far as I know, no scientific study has yet been published.
Someone with inner speech in a voiced language (or auditory-IS) might assume that signed-IS would basically be the same kind of thing as auditory-IS, except that it would occur in the visual modality rather than the auditory modality. But the matter cannot be so simple. The experience of externally signed speech is perspective-relative. When a signer sees the back of their own hands, for example, the interlocutor sees the signer’s palms. When a signer sees one of their hands moving left-to-right, the interlocutor sees the hand moving right-to-left. What the signer and their interlocutor sees is different. So, is signed-IS akin to seeing externally signed speech from the signer’s perspective, or from an interlocutor’s perspective?
We cannot really ask a question like this about auditory-IS. As Frank Jackson has pointed out to me, auditory-IS seems to be ‘aperspectival’. One does not represent any spatial relationship between oneself and the voice in one’s mind. Auditory-IS occurs in one’s mind, of course, but you do not represent the voice as coming from somewhere in relationship to yourself.
Joanna Atkinson, an important researcher on sign languages, has suggested a third possibility about the phenomenology of signed-IS as well: ‘Future research may reveal that mental representations of sign language are primarily kinaesthetic because muscular feedback predominates how an individual would experience their own productions’ (Atkinson 2006, p. 706).
The phenomenology of inner speech in voiced languages does not typically have a kinaesthetic aspect. Even if some of the cognitive and physiological mechanisms involved in speaking aloud are engaged when one produces inner speech in a voiced language (a separate topic), the phenomenology is usually purely auditory.
The question of what it is like to have signed-IS is, of course, an empirical one. It is entirely plausible that different signers will have different experiences: Some might have a visual experience from one perspective or the other; some might have a kinaesthetic experience. We can see, though, that, whatever the phenomenology of signed-IS, it is a mistake to assume that it is just a visual equivalent of auditory-IS.
This is important. It justifies asking questions about whether philosophical (and other) claims which have been made about inner speech in general are even plausible with respect to signed-IS. In my paper, I show that, whatever exactly the phenomenology of signed-IS, there will be a complication for at least some philosophical theory which is either about inner speech or which attributes a significant role to inner speech. Here, I will review just one such possibility.
Interpretevism (or inferentialism) about self-knowledge is the view that we learn about our own mental states, such as our beliefs and desires, in the same way that we learn about the mental states of others. We learn what others believe and desire by interpreting their behaviour and speech; we learn about our own mental states by interpreting our own behaviour and speech—including our inner speech. Interpretivism is strongly associated with Ryle (1949). Among contemporary philosophers, interpretivism (or some form or part of it) can be attributed to Carruthers, (e.g., (2009, 2011)), Cassam (2014), and Lawlor (2009).
Suppose that signed-IS is kinaesthetic. Could one interpret their signed-IS in the same way that they interpret the externally signed speech of others? They would have to be interpreting a mental state with a phenomenology akin to the movements they make when they sign, rather than a mental state with a phenomenology akin to what they see when others sign to them (or even what they see when they sign externally). This would be like interpreting a mental state with a phenomenology akin to the experience of moving one’s lips when one speaks externally. It is far from obvious that this is even plausible, for inner speech in either a signed language or a voiced language.
The difficulty here only increases when we think about what is involved in interpreting the behaviour and speech of others, and what it would mean to apply the same process to ourselves. The process of interpreting the behaviour and speech of others in order to infer their mental states is what philosophers call ‘mindreading’. If we interpret our own behaviour and speech, then we read our own minds, so to speak. For Carruthers (e.g., 2009, 2011), reading our own minds amounts to applying to ourselves a system which evolved specifically to determine the mental states of others. In general, our access to others’ behaviour and speech is through visual and auditory perception. It is not such a stretch for the interpretivist to say that auditory-IS might also provide usable input for the mindreading system when we turn it on ourselves. But what about kinaesthetic sensations? If we do not rely on kinaesthetic experiences in interpreting others’ behaviour and speech, then would signed-IS—if it is a kinaesthetic phenomenon—even be usable input for the mindreading faculty? If not, then it would not just be far from obvious that signed-IS could provide a source of self-knowledge. It would just be impossible that it could do so.
Of course, I have not done much more than raise the possibility that kinaesthetic experiences might not provide usable data for a process of interpretation. But there is clearly a complication here for the interpretivist. If they want to claim that we gain self-knowledge in part by interpreting our own inner speech, then they will need to say something about signed-IS.
And so it will be for many philosophical theories which are either about inner speech or in which inner speech plays some significant role. It cannot be assumed that what is true of auditory-IS will be true of signed-IS, and vice-versa. Whether it is so in any particular case is a question which will require dedicated attention.
Atkinson, J. R. (2006), ‘The perceptual characteristics of voice-hallucinations in deaf people: Insights into the nature of subvocal thought and sensory feedback loops’, Schizophrenia Bulletin, 42: 701–708.
Carruthers, P. (2009), ‘How we know our own minds: The relationship between mindreading and metacognition’, Behavioral and Brain Sciences, 32: 121–138.
Carruthers, P. (2011), The Opacity of Mind: An Integrative Theory of Self-Knowledge, Oxford University Press.
Cassam, Q. (2014), Self-Knowledge for Humans, Oxford University Press.
Gregory, D. (2026), ‘Inner speech and sign languages’, Synthese, 207: 176.
Lawlor, K. (2009), ‘Knowing what one wants’, Philosophy and Phenomenological Research, 79: 47–75.
Lœvenbruck, H., R. Grandchamp, L. Rapin, L. Nalborczyk, M. Dohen, P. Perrier, M. Baciu & M. Perrone-Bertolotti (2018), ‘A cognitive neuroscience view of inner language: To predict and to hear, see, feel’, in P. Langland-Hassan & A. Vicente (eds.), Inner Speech: New Voices, OUP, pp. 131–167.
Ryle, G. (1949/1963), The Concept of Mind, Middlesex, UK: Peregrine Books.