Does silent reading of direct speech activate voice-selective auditory cortex more than indirect speech, and does attention to the text moderate this?

Why quotation marks can make the brain 'hear' a voice

Reading is silent, but it is not necessarily soundless in the mind. When you read a line like *She said, "I'll be there at six,"* your brain may treat the quoted words as an actual utterance, briefly simulating the sound of a voice. Brain imaging work comparing direct quotations with indirect reports found that direct speech produced higher activity in the temporal voice areas—regions of the auditory cortex that normally respond to human voices—than indirect speech did [2]. In other words, the punctuation and phrasing that mark someone as speaking seem to cue the brain's voice-processing machinery, even though no sound reaches the ears.

The same pattern appears in work on the emotional side of inner speech: silent reading of direct quotations has been associated with elevated responses in voice-selective auditory cortex compared with indirect speech [1]. This is consistent with the idea that direct quotations invite readers to generate an inner voice, while indirect speech stays more like reported information.

Attention acts like a volume knob on the inner voice

The voice-selective response is not automatic and fixed. Direct brain recordings from people reading silently found that activity in auditory cortex, including voice-selective areas, was strongly modulated by attention—the response grew or shrank depending on how much the reader attended [3]. That suggests the 'inner voice' is partly something the reader does, not just something the text does to them.

Put together, the two threads fit a simple picture: direct quotations provide the cue to imagine a voice, and attention determines how loudly that imagined voice registers in the brain [2][3]. A distracted reader may show a weaker voice-selective response even to a vivid quotation, while an engaged reader may show a stronger one.

The attention question is plausible but not yet directly tested

It is worth being precise about what has and has not been shown. The studies that compared direct and indirect speech reported the voice-selective difference but did not test how attention to the text changes it [1][2]. The study that demonstrated strong attention effects on auditory cortex activity during silent reading did not compare direct with indirect speech [3]. So the combined explanation—that attention moderates the direct-versus-indirect difference—is a reasonable inference from separate findings rather than a result any one study directly established.

That leaves a clear open question: does paying closer attention to a quotation widen the gap between direct and indirect speech in voice-selective cortex? The existing evidence makes that prediction plausible, but it has not yet been confirmed in a single experiment that manipulates both quotation type and attention together.

What this means for how you read dialogue

For ordinary readers, the practical takeaway is that quotation marks are not just typographic decoration. They appear to signal the brain to treat the words as a voice, recruiting the same auditory regions that process real speech [2]. This may help explain why dialogue in a novel can feel more vivid and immediate than a summary of what characters said.

It also suggests that how absorbed you are in a text shapes how strongly that inner voice comes through [3]. Reading a quoted passage closely and attentively may make the imagined voice more pronounced, while skimming may leave it faint. The evidence supports this as a likely pattern, though the direct test combining quotation type and attention remains to be done.

Sources used in this answer

Related research questions