Abstract
The idea that some people completely lack inner speech is of both scientific and popular interest. In a recent study, Nedergaard and Lupyan compared self-reporting high and low inner-speech-prevalence groups and found that participants in the low-prevalence group performed worse on a verbal working memory test and responded more slowly and less accurately during rhyme judgments. These results represent an original contribution to the study of inner speech. However, the authors go on to draw the unfounded conclusion that their findings, together with previous empirical and anecdotal data, show that some people have no inner speech at all. They have coined the term anendophasia for this trait. This commentary examines Nedergaard and Lupyan’s claim of demonstrated anendophasia; I conclude they present no compelling evidence that some individuals lack inner speech.
Nedergaard and Lupyan (2024) present an inventive study showing that participants who self-reported a low incidence of inner speech (a) performed worse compared to high-prevalence participants on a verbal working memory test and (b) were slower and less accurate in making rhyme judgments. These two tasks are believed to recruit inner speech. The research contributes both empirically and methodologically to the challenging study of inner speech. At several points, however, the authors made the claim that some people have no inner speech at all, and they coined the term anendophasia for this trait (mirroring the recent coining of aphantasia for a lack of visual mental imagery). Both the claim and the term form part of the title of the article, and in some passages it is strongly implied that the claim is supported by the data of the study. Because there is a crucial difference between self-reported low prevalence and complete lack of inner speech, I instead argue that the claim is noticeably divorced from both Nedergaard and Lupyan’s (2024) data and from other data they present in support of the existence of anendophasia.
Research Transparency Statement
Clear evidence that some people entirely lack inner speech would be of great scientific interest, particularly because of the relevance of such evidence to the debate about possible interrelatedness between language and cognition. There is also public interest in the issue, and several science-news publications have reported that Nedergaard and Lupyan (2024) demonstrated the effects of a complete lack of inner speech on cognitive tests (Arnold, 2024; Makin, 2024).
Nedergaard and Lupyan (2024) asked 1,037 participants to complete the Internal Representations Questionnaire, or IRQ (Roebuck & Lupyan, 2020), allowing them to assemble the low and high inner-speech-prevalence groups on the basis of the responses. Curiously, nowhere is it stated that a single participant indicated a complete lack of inner speech. The authors argue that such individuals could be found if a larger participant sample were investigated. However, if the IRQ can reliably quantify the frequency of inner speech, then with 1,037 participants it seems likely that such individuals would already have been identified.
Nedergaard and Lupyan (2024) highlight (a) that introspective reports are not necessarily reliable and (b) that the IRQ cannot determine whether variations in people’s self-reported inner-speech prevalence are due to differences in the awareness of inner speech or to actual differences in frequency. 1 These well-considered calls for caution with respect to interpreting their own data contrast with the categorical claims made elsewhere in the article about some people lacking inner speech altogether.
Nedergaard and Lupyan (2024) also back up their claim (i.e., that some people lack inner speech) with anecdotal reports from popular science features (Felton, 2020; Soloducha, 2020) and with additional scientific data (Alderson-Day et al., 2018; Brinthaupt, 2019; Hurlburt et al., 2013). Central to Soloducha’s (2020) feature is a video interview with an individual who is presented as having no inner monologue. But the person’s actual testimony suggests otherwise: [M]y thought process is more like sporadic or random jot notes, it’s not my voice talking all the time, and it’s more like I’m just seeing lines of thought and maybe they have voices associated to them or tones associated to them, but not necessarily my voice.
On the IRQ this individual might score low on prevalence, but clearly this person is experiencing inner speech. Felton (2020) references a Reddit thread in which users Vadermaulkylo and GohanShmohan deny having any inner speech at all. However, Vadermaulkylo clearly experiences inner speech when reading, whereas GohanShmohan denies having any metacognitive abilities whatsoever—a self-contradictory claim. More importantly, Felton (2020) also references a blog post by Professor Russell Hurlburt (Hurlburt, 2011). Hurlburt has developed and published widely on descriptive experience sampling, or DES, in which participants going about their everyday lives are cued by a beeper at random intervals for several days. At each beep they quickly jot down notes about their inner experience, and after each day they are interviewed in depth about each beep. The DES method is considered to provide the most high-resolution and reliable data available about spontaneous inner experience (including inner speech). Hurlburt (2011) invoked Heavey and Hurlburt’s (2008) finding that 5 of their 30 participants reported no instances of inner speech during the beeps as confirmation that some people never experience inner speech. This inference is invalid, however; as Heavey and Hurlburt (2008) acknowledged, the absence of inner speech at random beeps does not imply its complete absence. When Dickens et al. (2018) sampled 10 participants in everyday life and while golfing, inner speech was reported at 8% and 27% of beeps, respectively. Two participants reported no everyday inner speech, but high frequencies during golf, suggesting that if Heavey and Hurlburt’s (2008) 5 participants went golfing, they would experience inner speech.
Alderson-Day et al. (2018), Brinthaupt (2019), and Hurlburt et al. (2013) presented different measures showing variations in people’s inner-speech frequency. However, the former two articles mention no respondents completely lacking inner speech, and the only such mention in Hurlburt et al. (2013) refers to the 5 participants in Heavey and Hurlburt (2008) discussed above.
Although aphantasia is currently ascribed to respondents scoring below a cutoff on the Vividness of Visual Imagery Questionnaire (Marks, 1973), aphantasia diagnosis based on self-report has been questioned (Larner et al., 2024) and the need for further validation of the concept is recognised (Keogh & Pearson, 2024). Confirmation of anendophasia would similarly involve several approaches, likely including DES and an interview protocol that includes (a) assessment of respondents’ presuppositions about inner speech and (b) direct attempts to elicit inner speech.
To conclude, there is no convincing evidence, in Nedergaard and Lupyan (2024) or elsewhere, to support the assertion that there are individuals who do not have any clinical speech conditions but who nevertheless lack inner speech completely.
Footnotes
Acknowledgements
The author wishes to thank Johan Blomberg, Petter Johansson, Louise Ström, and Trond Arild Tjøstheim.
Transparency
Action Editor: Zhicheng Lin
Editor: Simine Vazire
Author Contributions
