WORKSHOP ON INNER SPEECH: PHILOSOPHICAL AND PSYCHOLOGICAL QUESTIONS

The workshop will feature presentations in philosophy, linguistics and psychology. Topics may include: the nature and function of inner speech, the relation between inner speech, spoken language and other kinds of cognition, the neuroscience, psychology and psychopathology of inner speech.

This event will take place exclusively in person. If you would like to attend the workshop, please notify Ms Jonida Kodra at the following email address: [email protected]. Please be sure to include your name and, if you have one, academic affiliation.

Full list of speakers and titles of talks:

Sharon Geva: What do we really measure when we measure inner speech? Inner speech is often described using a variety of terms such as auditory imagery, self-talk, or verbal thinking, with some terms more frequently used in specific disciplines (e.g., self-talk in clinical psychology or covert speech in neuroimaging). And while these terms can be used interchangeably, they might also reflect different definitions of inner speech. Similarly, there are many tools that are used to measure and characterise inner speech, including both self-report questionnaires and experimental task. Each tool measures different aspects of inner speech and can therefore be associated with a specific definition of inner speech. In this talk, I will first explore the various terms used. I will then present a pre-registered systematic literature review that maps the tools used to measure inner speech to the cognitive constructs they are set to measure. The constructs reflected in studies of cognition include condensation and vividness, monologue/dialogue, spontaneous and task-dependent use, use of self or other’s voice, auditory features, embodiment, frequency and control. I will then present behavioural data to explore whether the mapping of measurement tools to cognitive constructs has empirical support. Lastly, I will examine whether individual variability in non-language functions (e.g., attention and imagery ability) can explain variability in inner speech task performance.

Daniel Gregory: Some properties that inner speech doesn’t have. The primary objective of this talk is to characterize inner speech from a new perspective, focusing on ways that the sensory manifestation of inner speech differs from that of external speech. Inner speech does not exhibit the same variation in pitch as external speech. It does not consist in a continuous stream of sound but in segmented representations. It does not allow the same degree of adjustment in the course of its own production, on account of the rapid rate at which it is generated. I show that each of these differences has some non-obvious philosophical implications.

Marta Jorba: Inner speech and emotion regulation. While the role of inner speech in cognitive functions such working memory, planning, cognitive flexibility, reasoning etc., has been studied quite extensively both in psychology and philosophy, its role and relation to affective experiences seems to deserve more attention. One useful strategy for emotion regulation seems to be emotion naming or labelling (Vygotsky 1934; Kopp 1989). In this talk I examine some recent work on emotion regulation (Nook et al., 2021, among others) and qualify the claim that labeling can be useful for emotion regulation, pointing out the importance of parameters such as time and intensity. Once it is clarified when labeling can have emotional positive impact, the question of why it might be so still remains. I propose that examining the profile of inner speech can be helpful in order to answer this question.

Peter Langland-Hassan: There are no words: the simulative view of inner speech. When we generate inner speech, do we produce words?  Many theoretical treatments of inner speech are surprisingly ambiguous in their answers to this apparently fundamental question.  We must stop the madness.  In this talk I will distinguish two views: the word production view and the simulation view.  The word production view holds that when we generate inner speech, words of a natural language are tokened within our minds; the simulation view denies this.  I will defend the simulation view against the best arguments I can think of for the word production view, while explaining the healthy lifestyle afforded by the simulation view.  It turns out that many who lean toward or openly endorse the word production view have nothing to fear in the view that there are no words in the head.

Helene Loevenbruck: Investigating the orthogonality of intention, agency, and control in inner speech
Gary Lupayn: The consequencres of individual differences in inner speech. People report large differences in their experiences of inner speech. Are these self-reported differences just artifacts of how we probe inner experience? Or do they signal genuine phenomenological differences? One way to find out is by measuring whether the self-reports are associated with differences in objective behavior in consistent and theoretically-meaningful ways. I will report a series of experiments showing links between inner speech and measures tapping into phonological and semantic processing. Despite these positive findings, there is no shortage of cases where differences in inner speech should matter, but do not. I will discuss how these negative findings can serve as stress tests for cognitive theories and what they tell us about the robustness of human cognition.

Jutta Mueller: Working memory and prediction as a potential window into the neural mechanisms supporting inner speech. Inner speech is a commonly experienced phenomenon that is involved in numerous language-related and non-language related tasks and contexts. It shapes how we process information by its linguistic representational format and the cognitive mechanisms operating on it. I will suggest that the investigation of inner speech in working memory and prediction tasks allows to isolate important involved core ingredients and processes even though the tasks conflate inner speech with other task-related processes. The argument will be supported by a set of fMRI and EEG studies on working memory and prediction that were not performed to assess inner speech, yet, shed light on covert language production. I will show that working memory benefits not only from phonological rehearsal, but even more from linguistically coded semantic representations. Further, I will demonstrate that the process of phonological rehearsal may include core processes of language production, namely prediction. Thus, verbal working memory, language production and prediction are tighly intervowen and may jointly contribute to many cognitive tasks which, oftentimes, may include experiences of inner speech. I will argue that, in order to gain insights on how different linguistic computations operate within our brains, future neuro-cognitive research should aim for integrating subjective measures that capture the diversity of inner speech experience with sophisticated experimentation that allows isolating core ingredients of inner speech in service of human cognition.

Wade Munroe: Inner speech and natural language as a language of thought.

Agustin Vicente: Speech productions models: consequences for inner speech. Since Levelt (1989), speech production models have it that speech production begins with a message to be conveyed. Most implementations of this model, including Levelt’s own, assume that such a message has to adjust to the expressive demands of the target language (English, German, Spanish, etc.). In particular, messages have to be composed of “lexical concepts”, that is, concepts that can be expressed by monomorphemic words in the target language. There are two ways to comply with such a demand: (i) our thoughts are built out of lexical concepts; (ii) in order to become messages, our thoughts are translated into a format that meets the expressive demands of the target language. While option (i) is simpler, most theorists (e.g., Levelt, 1996) assume (ii). In my talk, I will argue that the function and nature of inner speech in part depends on how messages are construed. I will try to show that the assumption that messages can be the output of a translation process is problematic. If the assumption is indeed problematic, then option (ii) probably entails that whenever we use inner speech we are having thoughts that we cannot otherwise have. Meanwhile, if option (i) is pursued, inner speech can be just an externalization of already existing thoughts for e.g., attentional purposes. I will argue in favor of option (i).