Daniel Gregory

Inner speech is known as the “little voice in the head” or “thinking in words.” It attracts philosophical attention in part because it is a phenomenon where several topics of perennial interest intersect: language, consciousness, thought, imagery, communication, imagination, and self-knowledge all appear to connect in some way or other to the little voice in the head. Specific questions about inner speech that have exercised philosophers include its similarities to, and differences from, outer speech; its relationship to reasoning and conceptual thought; its broader cognitive roles—especially within metacognition and self-knowledge; and the role it can play in explanations of auditory verbal hallucinations and “thought insertion”.

The problem of dream skepticism – i.e., the problem of what can justify one’s belief that they are not dreaming – is one of the most famous problems in philosophy. I propose a way of responding to the problem which is available if one subscribes to the theory that the sensory experiences that we have in dreams consist of images (as opposed to false percepts). The response exploits a particular feature of imagination, viz., that it is not possible to simultaneously have two separate imagistic experiences in the same modality.

In November 2022, OpenAI released ChatGPT, an incredibly sophisticated chatbot. Its capability is astonishing: as well as conversing with human interlocutors, it can answer questions about history, explain almost anything you might think to ask it, and write poetry. This level of achievement has provoked interest in questions about whether a chatbot might have something similar to human intelligence or even consciousness. Given that the function of a chatbot is to process linguistic input and produce linguistic output, we consider the question whether a sophisticated chatbot might have inner speech. That is: Might it talk to itself, internally? We explored this via a conversation with ‘Playground’, a chatbot which is very similar to ChatGPT but more flexible in certain respects. We asked it questions which, plausibly, can only be answered if one first produces some inner speech. Here, we present our findings and discuss their philosophical significance.

There is a growing literature in philosophy dealing with the phenomenon of inner speech, that is, the activity of speaking to oneself in one’s mind. This paper highlights a feature of inner speech which has not yet been noticed in this literature: that there is something distinctive that it is like to make a sincere assertion in inner speech. The paper then traces out two implications of this observation. The first relates to the question of how we should characterise inner speech; the second relates to the question of how inner speech may play a role in self-attributions of belief.

A theory which has had significant influence seeks to explain auditory verbal hallucinations as utterances in inner speech which are not properly monitored and are consequently misattributed to some external source. This paper argues for a distinction between inner speech and imagined speech, on the basis that inner speech is a type of actual speech. The paper argues that AVHs are more likely instances of imagined speech, rather that inner speech, which are not properly monitored : 86–107, 2012), Cho and Wu and Cho and Wu, although they prefer a quite different explanation of AVHs).

This dissertation explores the phenomenon of inner speech. It takes the form of an introduction, which introduces the phenomenon; three long, largely independent chapters; a conclusion; and an appendix. The first chapter deliberates between two possible theories as to the nature of inner speech. One of these theories is that inner speech is a kind of actual speech, just as much as external speech is a kind of actual speech. When we engage in inner speech, we are actually speaking, but we are doing so silently. The other theory holds that inner speech is a kind of imagined speech. When we produce inner speech, we are imagining performing the action of speaking. The chapter argues for the theory that inner speech is a kind of actual speech. The second chapter argues against a theory which holds that inner speech is dialogic. On this theory, a subject represents different perspectives in inner speech and a dialogue can take place in the same sense in which a dialogue can take place between different individuals in external speech. The chapter borrows some important material from the philosophy of language to show that this position, though it might have some intuitive appeal, is ultimately implausible. The third chapter is concerned with the question whether inner speech can be a source of knowledge of our own beliefs. It shows that the view that inner speech can be such a source is subject to an adapted version of a problem from the epistemology of testimony: roughly, what justification do we have for believing that we believe what we say in inner speech? It makes use of some material from the recent debate about cognitive phenomenology to develop a version of the view which is not subject to this problem. It then provides some initial discussion of the merits of this view. The appendix takes up a more practical issue regarding inner speech. There is a theory that auditory verbal hallucinations – i.e. experiences of voice-hearing – take place when someone produces an utterance in inner speech but loses track of the fact that they have produced the utterance. Accordingly, they have an experience as of something being said and, not realising that they are the source of the experience, postulate some external cause, i.e. someone else speaking. The appendix develops an alternative account which has been suggested in the literature, at times drawing upon earlier work in the dissertation.