Philip Jaekl is a cognitive neuroscientist and writer. He studied at York University in Toronto, and completed his postdoctoral research at Pompeu Fabra University in Barcelona and the University of Rochester in New York. His writing appears in The Atlantic and The Guardian, among others. He lives in Tromsø in Norway, and is currently writing his debut nonfiction book.
Photo from fizkes/Getty Images.
‘I think, therefore I am,’ the 17th-century philosopher René Descartes proclaimed as a first truth. That truth was rediscovered in 1887 by Helen Keller, a deaf and blind girl, then seven years of age: ‘I did not know that I am. I lived in a world that was a no world … When I learned the meaning of “I” and “me” and found that I was something,’ she later explained, ‘I began to think. Then consciousness first existed for me.’ As both these pioneers knew, a fundamental part of conscious experience is ‘inner speech’ – the experience of verbal thought, expressed in one’s ‘inner voice’. Your inner voice is you.
That voice isn’t the sound of anything. It’s not even physical – we can’t observe it or measure it in any direct way. If it’s not physical, then we can arguably only attempt to study it by contemplation or introspection; students of the inner voice are ‘thinking about thinking’, an act that feels vague. William James, the 19th-century philosopher who is often touted as the originator of American psychology, compared the act to ‘trying to turn up the gas quickly enough to see how the darkness looks’.
Yet through new methods of experimentation in the last few decades, the nature of inner speech is finally being revealed. In one set of studies, scans are allowing researchers to study the brain regions linked with inner speech. In other studies, researchers are investigating links between internal and external speech – that which we say aloud.
The roots of the new work trace back to the 1920s and the Russian developmental psychologist Lev Vygotsky, who said the human mind was shaped by social activity and culture, beginning in childhood. The self, he hypothesised, was forged in what he called the ‘zone of proximal development’, the cognitive territory just beyond reach and impossible to tackle without some help. Children build learning partnerships with adults to master a skill in the zone, said Vygotsky, then go off on their own, speaking aloud to replace the voice of the adult, now gone from the scene. As mastery increases, this ‘self-talk’ becomes internalised and then increasingly muted until it is mostly silent – still part of the ongoing dialogue with oneself, but more intimate and no longer pronounced to the world. This voice – at first uttered aloud but finally only internal – was, from Vygotsky’s perspective, the engine of development and consciousness itself.
Vygotsky’s theory of childhood development contrasted sharply with those of his Western counterparts. William James had a complete disdain for the study of inner speech, because, to him, it was a ghost: impossible to observe. The French developmental psychologist Jean Piaget insisted that private speech signified simple inability – it was the babble of a child without capacity for social communication with no relation to cognitive functioning at all. Through much of the 20th century, Piaget seized the reigns of child development, insisting that children had to reach a developmental stage before learning could occur. Which came first: the chicken or the egg? Vygotsky said that learning occurred, then the brain developed. Piaget said the brain developed, then learning occurred.
Over years of meticulous experiment behind the Iron Curtain, Vygotsky continued to make his case. One thing he did was study children in the zone of proximal development as they worked with adults to accomplish tasks. In the experiments, the child would be presented with a challenge and a tool for overcoming it. In the zone, Vygotsky observed what he called ‘private speech’ – self-talk that children between the ages of two and eight often engage in. This intermediate stage, he held, was connected on one end to a prior period when we had no thread of memory (and no inner voice) and on the other end to true inner speech so crucial to self-reflection, narrative memory, and development of cognitive skills.
Within the newly forming Soviet Union, Vygotsky’s research was stigmatised, in large part because it used intelligence testing to validate some concepts; IQ testing itself had been banned as a challenge to Marxist principles of equality. Despite the roadblocks, in 1934 (the year of his death) Vygotsky finally published his opus on inner speech and childhood development, Thought and Language. It was a potent challenge to Piaget but, shrouded by the Stalinist censor, his ideas remained under wraps.
Meanwhile, in the West, newer work in developmental psychology began to chip away at the acceptance of Piaget. When Thought and Language was finally rediscovered and published in English by MIT Press in 1962, it was the perfect moment. The book provided a rational, alternative way to conceive the development of the mind. And further translations of Vygotsky’s writings led to a plethora of hypotheses ripe for testing.
By 1970, the push to validate Vygotsky’s ideas had picked up steam. A leader of that era was the American psychologist Laura Berk, professor emeritus at Illinois State University, an expert on childhood play. Berk observed children engage in imaginative, ‘make-believe’ play, and demonstrated that the substitution of objects – say a cup for a hat – requires internal thought (and self-talk) rather than impulse. Her studies show that during imaginative play, children’s self-talk helps them guide their own thoughts and behaviour and exert true self-control. She and many other child psychologists demonstrated the importance of the inner voice, beyond a doubt, elevating Vygotsky and burying Piaget for good.
With inner speech clearly established as a chisel for the young mind, many more questions remained. Do people in adulthood experience inner speech in the same way as children – or even as each other? Do most of us even have an inner voice – an internal commentator narrating our lives and experiences from one moment to the next?
These were deeply controversial and introspective questions in the 1970s, and they captured the imagination of Russell Hurlburt, an aeronautical engineer-turned-clinical-psychology graduate student at the University of South Dakota. Hurlburt had envisioned a way to accurately sample others’ random inner experiences. Today a professor of psychology at the University of Nevada, Las Vegas, he’s been honing the technique ever since.
Hurlburt calls his methodology Descriptive Experience Sampling (DES), and it works by sampling the inner thoughts of a given interviewee during those moments when a beeper randomly goes off. After extracting the contents of inner experience from countless interviews, Hurlburt has defined an array of phenomena typically shared by humans – auditory and visual imagery, emotion, awareness of real stimuli and a category of thoughts that occur without words, images or symbols of any kind. The main contribution here, though, is actually DES itself. Before its inception, introspective methods had been shunned for decades, if not centuries, as being too highly influenced by bias to be taken seriously. Now, with DES, Hurlburt believes in the possibility of obtaining unbiased, accurate snapshots of inner experience that includes inner speech.
Freed from the mundane confines of a laboratory, the data come from ‘the wild’, as Hurlburt puts it. A participant wears the beeper, which can go off at any moment throughout the day. They go about their daily activities and are likely to forget its presence. When the beeper does go off, the participant makes a careful note of exactly what their inner experience was immediately beforehand. Subsequently, they are questioned by Hurlburt about that experience in a thorough but open-ended interview.
The interview process itself requires an exacting, friendly yet trial-like probe of what occurred. In one unedited transcript in Hurlburt’s book Exploring Inner Experience (2006), a participant named Sandy is quoted following a beep: ‘I was reading. I was starting with the word “life”… and I had an image in my head – it was a black and white image, by the way – of… OK, I was staring at the word “life” and I had said to myself “life” in my own tone of voice.’
Sandy was referring to inner speech using the word ‘life’. For the next six minutes Hurlburt probed her about this experience. His questions eventually helped Sandy divulge that as she was inwardly speaking the word ‘life’ she simultaneously ‘saw an image of that word in an old-courier like font – black on a white background’ and a moving image of ‘sand pouring’ from a hand of unknown agency below her face.
In a sample of bulimic participants, Hurlburt found the propensity for multiple inner voices at the same time
DES requires careful skill to capture these kinds of experiences accurately – what Hurlburt terms, ‘high-fidelity, pristine’ inner speech as it naturally occurs. He takes care not to bias the participant in any way. ‘There are a lot of people who believe that you talk to yourself all of the time, so that’s a form of external pressure to say you were inner speaking when maybe you weren’t,’ he notes. For example, noted consciousness researcher Bernard Baars has asserted that ‘overt speech takes up perhaps a tenth of the waking day; but inner speech goes on all the time’. Hurlburt’s research shows this isn’t true; he finds that inner speech consumes about 25 per cent of an average person’s day, and thus, he is careful to not communicate any assumption about what type of inner experience a DES interviewee may have had at the time of the beep.
Thanks to the accuracy of DES, Hurlburt has found thought patterns associated with various clinical populations, including those with schizophrenia, bulimia nervosa, and autism. In a sample of bulimic participants, for instance, he’s found the propensity for multiple inner voices experienced at the same time. Take ‘Jessica’, a patient watching television when the DES beep occurred. In the front of her head, Hurlburt explains, she was inwardly saying ‘blond’, ‘skinny’, ‘guys’, and ‘stare’ in what was her own, unspoken voice. At the same time, in the back part of her head, she was saying, in another, quieter inner voice, still her own: ‘Why is it that movies and TV shows always have ‘girls for’, ‘to’, and ‘at’? Importantly, such experiences are not often perceived by the experiencers themselves, let alone revealed to anyone else.
Hurlburt has found that we typically self-talk in voices we regard as our own and, though silent, we attribute to these voices sonic characteristics such as tone, pitch and pacing. We invest them with emotional qualities similar to external speech. Finally, inner speech mostly occurs in complete sentences and is nearly always actively produced rather than passively experienced.
Recently, Hurlburt teamed up with Charles Fernyhough, a leading researcher of inner speech and auditory hallucination at Durham University in the United Kingdom. To conduct their collaboration, they put DES participants into brain scanners using functional Magnetic Resonance Imaging (fMRI), which detects metabolic changes in the brain. Like other participants, these subjects were asked to record inner speech and experience occurring just prior to a beep – but this time, brains would be scanned. These scans were compared with others that were captured when participants inwardly repeated words that they read on a screen – a method for investigating inner speech that had been used in many previous studies. The researchers found that activity increased in a brain region called Heschl’s gyrus during spontaneous inner speech, but not when the self-talk is prompted – indicating that the neural nature of pristine self-talk is unique, indeed.
Picking up where Vygotsky and researchers like Berk left off, Fernyhough has been investigating the role that inner speech plays in developing, evolving minds. After the private speech of childhood has finally been internalised, suggests Fernyhough, inner speech emerges in a multiplicity of ways – each comparable to speech spoken out loud.
Fernyhough calls the most familiar level of inner speech ‘expanded’ because it is basically the same as external speech – grammatical and fully formed, but not vocal. He believes this kind of inner speech is most likely engaged when we are under stress or cognitive pressure. Imagine, for example, while travelling, that you are making an important phone call regarding a lost passport. While on hold there’s a good chance that you’ll mentally rehearse exactly what you are about to say to the official on the other end – your story about how your passport went missing – in language that is full and complete.
To date, Fernyhough and colleagues have devised clever ways of exploring expanded speech through its tight connection with out-loud speech. For example, they’ve shown that external speech can interfere with inner speech when it is required for memorisation. If participants try to silently memorise a list of items – a task that requires inner speech – while they recite out loud the days of the week, they can’t do it. Speaking aloud effectively annihilates the role that expanded inner speech would otherwise play.
In another study, Fernyhough and his colleague Simon McCarthy-Jones captured fMRI images of brain activity linked with a form of expanded speech labelled ‘dialogic’ because it involves envisioning a dialogue with another. To do the experiment, the researchers had participants imagine themselves in a variety of scenarios, like going to their old school or meeting the Prime Minister. In these imaginary scenarios, they inwardly conversed with their old teacher or interviewed the Prime Minister, while the scanner recorded the active areas in their brains. These scans revealed that, in a manner similar to real communication, inner dialogues can also recruit neural regions such as the posterior temporal cortex, involved in what’s called ‘theory of mind’ – the ability to attribute mental states to others that are different from one’s own. This happens even if the dialogue is solely within the mind of the self.
The second broad category of inner speech defined by Fernyhough is considerably more mysterious and enigmatic. He calls it ‘condensed’ inner speech, borne out of Vygotsky’s belief that as speech becomes internalised it can undergo profound transformations that set it distinctly apart from the expanded version. Condensed inner speech is defined as a highly abbreviated and ungrammatical version of regular speech. Although possibly linguistic – comprised of words – it is not intended to be communicated or even understood by others. For example, as a habit in the winter since my younger days, I often think to myself, ‘passlockmoney’ before heading out the door to go snowboarding. For you to understand what I mean, I’m required to expand this term: Remember your ticket or pass if it is still valid, your snowboard lock, and cash or credit card for getting lunch (and après beer).
Sometimes there are holes in your inner speech; sometimes all words are missing and yet you still experience yourself speaking
The variety of ‘condensed’ experiences people have are remarkably unique because we already know the meanings of the contents of our own thoughts – there’s simply no need for a brain to slog out a rich, grammatical format of inner thought that can be understood by others, when thinking to ourselves. Thus, beyond the abbreviation of ‘passlockmoney’, the condensation of inner speech can result in ‘thinking in pure meanings’ as originally stated by Vygotsky. Condensed inner speech can even have most of its auditory qualities stripped away, such that ‘there isn’t much speechy about it’, say Fernyhough.
Hurlburt says inner speech can indeed involve elimination of words entirely, while the linguistic experience remains intact. ‘Sometimes there are words that are missing – “holes” in your inner speech. Sometimes the whole thing [all words] are missing and yet you still experience yourself speaking,’ he states. In this case, the person reports the experience of speaking, including its production, sense of loudness, pace etc, and senses what is being said but does not experience any words in their usual sense.
He has also posited a passive form of inner speech that he calls ‘inner hearing’. ‘It’s possible to inner “hear” your own voice rather than speak your own voice,’ he tells me. Here, people listen to their own voice in their heads, perceiving the same sonic characteristics as expanded speech, but without the agency. Such experiences have been recalled by participants as their voice ‘just happening’, as ‘coming out of its own accord’, as ‘taking place’ rather than ‘being uttered’.
Some people passively experience inner speech in voices not their own – essentially as auditory hallucinations that they cannot control. Founding member of the Beach Boys Brian Wilson described the experience to Larry King in an interview on CNN in 2004: ‘I’m going to kill you. I’m going to hurt you’, an inner voice had continually repeated to him since his initial experiences with LSD in the 1960s. The value of understanding such hallucinations is self-evident: they are a hallmark of schizophrenia, a condition that affects almost 24 million people worldwide.
Of great fascination, Fernyhough has concluded that a small but significant part of the general population also experience auditory hallucinations – a phenomenon the researchers call ‘voice hearing’ to distinguish it from schizophrenia. Such voices have been reported by noted individuals throughout history, says Fernyhough. The Greek philosopher Socrates described what he called a ‘daemonic sign’, an inner voice warning him that he was about to make a mistake. Joan of Arc described hearing divine voices since childhood – the same ones that influenced her motivation to help in the siege of Orleans. The 15th-century mystic and autobiographer Margery Kempe wrote about inner conversations with God. Sigmund Freud was not immune: ‘During the days when I was living alone in a foreign city … I quite often heard my name suddenly called by an unmistakable and beloved voice.’
All this leads to another, confounding question: are verbal thoughts reaching awareness just the tip of a mental iceberg, offering only a glimpse of the unconscious mind? The possibility was posed by Vygotsky, but Fernyhough doesn’t like going there: ‘When we are talking about thinking, we are talking about conscious processes.’
Whatever we ultimately find, the work pioneered by Vygotsky in human development promises to illuminate more secrets of selfhood and consciousness over time. Today, the new science of inner speech is turning up the gas so fast it is lighting the dark.