We tend to think of language as a transparent tool—a neutral medium for expressing thoughts and describing reality. While philosophers have (more or less) come to agree that “the map is not the territory” more recent thinkers have argued that language does play a fundamental role in shaping our perception of the world and our notion of self. By drawing parallels between language loss from brain injuries and the experiences reported during deep meditation psychedelic states, UCL neuroscientist Jeremy I Skipper argues language creates an altered state of consciousness. Only by losing language can we start to dismantle the scaffolding that supports our notion of the world and the self.
Altered articulation
Contemplate this self-report, recounting the experience of a “non-ordinary” or “altered state of consciousness”:
... my brain chatter began to disintegrate ..., my consciousness soared into an all-knowingness, a “being at one” with the universe ... I no longer perceived myself as a whole object separate from everything ... I retained only a vague idea of who I was ... I had stepped beyond my perception of myself as an individual ... My entire self-concept shifted as I no longer perceived myself as ... an entity ... separated ... from the entities around me.
The following is a similar description from another person:
I had a nothing mind, a flotsam mind. I was incredibly focused on the present, with very little awareness or interest in my past or future. My entire environment felt interconnected, like cells in a large, breathing organism. I felt less like myself and more like everything around me.
What caused these altered states? If you have tried an ‘heroic’ amount of magic mushrooms or other psychedelics, the accounts of interconnectedness and loss of “self” might feel familiar. If you have not, you nonetheless might have experienced “oceanic boundlessness” and “ego dissolution” by other means. These include advanced meditation, breath work, dreams, flow and peak states, mystical awakening, and near-death and out-of-body experiences.
You might be surprised to discover that the above reports were given by individuals who had severe brain injuries that resulted in loss of language. Both recovered speech and wrote beautiful books describing their experiences of living in a language-less world. At the time of their injuries, one was a renowned neuroscientist and the other an actor. Why would these “strokes of insight” feel at home in the Erowid Psychoactive Vaults, or Subreddit communities sharing psychedelic “trip reports” or discussing the “jhanas” (advanced meditative states that can occasion “nondualism”)?
___
"I did not know that I am. I lived in a world that was no world. I cannot hope to describe adequately that unconscious, yet conscious time of nothingness."
___
An intimation comes from Helen Keller, the deaf-blind author, educator, humanitarian, and disability-rights pioneer. Keller did not learn a language until middle childhood. She writes of her experience preceding this education in “Before the Soul Dawn”:
Before my teacher came to me, I did not know that I am. I lived in a world that was no world. I cannot hope to describe adequately that unconscious, yet conscious time of nothingness. I did not know that I knew aught or that I lived or acted or desired. Since I had no power of thought, I did not compare one mental state with another. When I learned the meaning of “I” and “me” and found that I was something, I began to think. Then consciousness first existed for me.SUGGESTED VIEWINGYour true self With Joanna Kavenna, Betty Sue Flowers, James Tartaglia, Bahar Gholipour
Keller describes an undifferentiated and “me”-less reality before she acquired language. This sounds similar to the descriptions we began with. Taken together, these three anecdotes suggest that the phenomenology of psychedelics and other altered states of consciousness is related to the human capacity for language. Beyond the mere difficulty of describing “oneness” or other novel states, such experiences might be ineffable because having them involves language being literally unavailable to organise and even dictate how our worlds appear to us. Without the guiding hand of language, “reality” is free to express itself in manners we cannot easily comprehend.
Perennial prattle
Words are not ... facts or things: we need therefore to prise them off the world ... so that we can realize their inadequacies and arbitrariness, and can relook at the world without blinkers.
J L Austin
This view corresponds to a sort of “perennial philosophy of language.” There is at least a 2,500-year history of mystics, philosophers, and psychologists who maintain that language interferes with and obscures our understanding. They often describe a sort of utopia resulting from a language-free mind, one in which we jettison a deluded form of consciousness that we have confused for reality.
Around 400 BCE, the Tao Te Ching begins with “The Tao [or Way] that can be spoken is not the eternal Tao.” A bit later, the Chinese philosopher Zhuang Zhou provides one possible explanation: the “Way has never known boundaries ... But because of ‘this,’ there came to be boundaries.” That is, words categorise, creating artificial borders and divisions that conceal reality. 600 or so years later, the Indian monk and Buddhist philosopher Nāgārjuna gave a different perspective. He observed that “a name ... due to the absence of substance in things, is insubstantial and therefore empty,” pointing to how language deceives us by implying solid, independent realities where none exist.
Around the same time as Nāgārjuna (about 250 CE), the Greek Platonist, Plotinus, maintained that the “spoken word is an imitation” of reality, and an imprecise and inadequate one at that. Fast forward 1,000 years and the German Christian mystic Meister Eckhart preached that “whatever we say of God is not true, and what we do not say of Him is true,” meaning language fails to capture or distorts (ultimate) reality. In “That Nothing is Known” (1581), the philosophical sceptic Francisco Sanches wrote that “there lies in words no power to explain the nature of things.” And More literarily, Nietzsche wrote in 1873: “We believe that when we speak of trees, colours, snows, and flowers, we have knowledge of the things themselves, and yet we possess only metaphors of things which in no way correspond to the original entities.”
At the start of the twentieth century, the Nobel prize winning (and Einstein debating) French philosophical celebrity, Henri Bergson, had much to say about how language reifies a static reality. He noted that “the word... overwhelms or at least covers over the delicate and fugitive impressions of our individual consciousness.” In 1931, Alfred Korzybski, a Polish-American scholar, memorably quipped a similar sentiment, namely that “The map is not the territory” in that words are “filters” that “are not the things they represent,” such that to “use words to sense reality is like going with a lamp to search for darkness.” Around the same time, the British mathematician and philosopher, Alfred North Whitehead, expressed this more formally as the “fallacy of misplaced concreteness,” in which abstractions like those engendered by words are confused for reality, and was himself “impressed by the inadequacy of language to express our conscious thought.”
SUGGESTED VIEWINGMind, Maps and RealityWith Maria Balaska, Quassim Cassam, Hilary Lawson, Lisa Feldman Barrett
Around 1940 the Austrian philosopher Ludwig Wittgenstein’s concept of “language games” illustrated that language shapes our perception of reality by demonstrating that meaning arises from use within specific contexts—for instance, the medical language of “mental illness” frames conditions as biological disorders, influencing both diagnosis and treatment. His critique of the notion of a “private language” underscores this by showing that language, and thus our grasp of reality, is inherently social; our understanding is necessarily constrained and moulded by shared, public forms of expression, rather than by individual, isolated experience.
These views resonate with the linguistic relativity hypothesis of Sapir and Whorf that emerged around the same time in the field of psychology in the United States. This proposes that linguistic structures influence thought and worldview, with language determining cognition:
We cut nature up, organize it into concepts, and ascribe significances as we do, largely because we are parties to an agreement to organize it in this way—an agreement that holds throughout our speech community and is codified in the patterns of our language.
The neo-Whorfian hypothesis is a modern refinement of the Sapir-Whorf relativity hypothesis and is supported by much empirical evidence. It posits that the language we speak influences our cognitive processes and perception of reality, though not always deterministically.
Brain banter
I also see the words fall, they come from up above, as if they were little luminous objects falling from the sky, the language falls on the sacred table, falls on my body, then with me hands, I catch word after word
Maria Sabina as quoted by Allyson Grey
Biologically speaking, from whence does this perennial philosophy come? Language is a tool that is so ubiquitous and useful, that it has made humans into “language cyborgs,” with new cognitive abilities compared to other animals. It reduces the dimensionality of our environments via labels, enhancing our abilities to represent, learn, store, and manipulate information, allowing us to allocate attention and coordinate activity with ease. However, as the perennial philosophy of language reveals, these enhancements likely come at a cost, creating a “reality tunnel” in which the true spatial heterogeneity and perpetual temporal flux of the world is obscured and fixed in time. We “mistake the map for the territory,” confusing the “one great blooming, buzzing confusion” we experienced as infants with our low-dimensional linguistic representations as adults.
Thus, humans live in a linguistically altered state of consciousness that we do not recognise as such because, like the proverbial fish, we are swimming in language. What we call “altered states of consciousness” might naively be thought of as a return to a reality unmediated by linguistic representations, to a more “animalistic” form of cognition. However, this is overly simplistic since we have brains that have been sculpted by and are always in touch with language. Thus, even if we manage to temporarily shut language down, it leaves in its wake unique human artifices resident in the brain, including information otherwise unexpressed in the animal kingdom, like “free will,” “infinity,” “money,” etc., or more overt works of fiction, like aliens, dragons, and machine elves.
To better understand this, it is useful to review modern conceptions of the neurobiology of language. We are amidst a paradigm shift, one in which many cognitive neuroscientists view behaviours as being implemented in complex networks, distributed throughout the whole brain. This is true of language, where, through a process of learning, regions of the brain associated with the relatively arbitrary sounds that comprise words, form networks with the regions in which the information those sounds are associated with is processed, giving them semantic meaning. For instance, the acoustic word-form for “happy” is processed in auditory cortices, and forms contextually determined functional connections with brain regions involved in interoception (our sense of our insides, like our hearts pounding) and emotional processing. These connections are reciprocal, with interoceptive processing of a certain variety resulting in the “happy” label becoming activated.
Let’s now extrapolate this to a toy model of how changes in the brain basis of language might lead to what we call altered states of consciousness. Through psychedelic consumption or years of meditation, auditory cortical functioning is inhibited or momentarily reduced. This severs word forms from the distributed semantic representations that they are bound to. Concomitant with these changes there is a corresponding excitability in regions of the brain typically involved in semantic control processes, either through pharmacological means or increased effort. This gives our brain a new-found freedom—free from the categorical hold of words—in which we can do new things with our semantic representations. If language is a sort of information compression mechanism, zipping rich multidimensional experiences into simpler low-dimensional verbal descriptions, release from language might allow access to an “uncompressed” experience.
___
Language structures our narrative self linearly in time, demarcating past (“I went”), present (“I am here”), and future (“I will go”). Without this structure, we might be more present in the moment or process time less linearly or perhaps in parallel.
___
Specifically, this set of processes might “expand consciousness” and lead to increases in creativity and insight. First, without being micromanaged by words, each of our sensory systems is free to have more granular, pre-categorical perceptions. For example, the word “blue” shapes our conscious experience of a specific range of wavelengths of light, making sensations within that range more difficult to distinguish. Different blues might be more readily and vividly experienced without accompanying labels. Our words help partition the sensory world into distinct modalities. Without them, more cross-modal integration might occur, perhaps resulting in synesthetic experiences like hearing shapes. Altered language functioning might reduce the reflexive categorisation of sensory information, allowing for more flexibility and exploration in perception. Weakened categorical boundaries might facilitate blending of and seeing more connections between what were thought to be distinct entities. So trees might be attributed agency and everything might be experienced as “one.”
This logic as applied to sensory systems can be extended to linguistic narratives. We have a “narrative self” in which the words “I” and “me” carry a great deal of connected semantic structure. Without these labels, self-referential processing and the filtering of everything through self-related concerns would diminish. Boundaries between the self and others might blur. Language structures our narrative self linearly in time, demarcating past (“I went”), present (“I am here”), and future (“I will go”). Without this structure, we might be more present in the moment or process time less linearly or perhaps in parallel. In the extreme, time might cease to exist. Our narratives embed many assumptions about scale (microscopic/human/cosmic), hierarchy, metaphysics, etc. Freedom from these constraints might allow insights that cross scales, levels, and presumed realities. Given that language normally fragments experience into labelled parts, we might better grasp things as a gestalt, i.e., more holistically, transcending part-whole divisions, perhaps leading to nonduality (states characterised by the belief that there is no division between the self and anything else).
At more cognitive levels, language is a filter that directs our attention to learned verbal categories and reduces attention to experiences that do not fit into those categories. When the filter is removed, we might attend to previously unattended information. During everyday communication, attention to the objects in the world and previously spoken words might lead to expectations and regular (if not ubiquitous) prediction of what words come next. Reducing these predictions might open a window for more novel information-processing.
Machine murmur
The world is made of language.
Boiling this all down, we can say that the neurobiology of language is a whole brain process that flexibly organises our experience of the world. This involves categorically organising external sensory information (e.g., into colours, objects, and smells). Language also categorically organises internal information, like labelling interoceptive processes as emotional words, but also higher-level constructs like the “narrative self,” largely involving stories constructed and manipulated with words. These categorical processes are established with learning and become both positively and negatively entrenched in the brain. From this perspective, increased “unity” during altered states can be seen as the result of these entrenched linguistic categories becoming less differentiated. Similarly, ego-loss results from the diminished ability to maintain higher-level language categories like the self.
Supporting this view, studies of psychedelic-induced altered states suggest that these correlate with alterations in language. Psychedelic phenomenology is almost definitionally described as ineffable. When people in studies are asked to produce language when on psychedelics, speech output is slower, reduced in quantity, and less complex, with more errors. Semantics tend to be bizarre and unpredictable and show more spread (“a giraffe banana”) and altered topics (e.g., a decreased focus on the self). Pharmacologically, psychedelics bind to serotonin receptors more robustly expressed in brain regions traditionally associated with language, with activity in these regions reported to be changed and correlated with degree of “ego dissolution.” These same language-related brain regions are impacted by meditation, and this likely applies to other altered states of consciousness.
___
It should be possible to induce artificial intelligence to have psychedelic-like “trips” or exhibit behavioural reports resembling other altered states of “consciousness” by manipulating language
___
Given findings like these and the described framework for understanding the “perennial philosophy of language,” it should be possible to induce artificial intelligence to have psychedelic-like “trips” or exhibit behavioural reports resembling other altered states of “consciousness” by manipulating language. Similar to the neurobiology of human language, multimodal AIs are trained on massive amounts of text, and learn “embodied” semantic representations tied to visual information accompanying words during training (e.g., associating specific pixel values from images with “blue”). In analogy with humans on psychedelics or in advanced meditative states, researchers have inhibited word forms in these AIs, possibly leaving the models semantics representations to be less categorically constrained. They then had the AIs generate text and examined how much that text would be identified as aligning with questions from standardised questionnaires for measuring altered states of consciousness. AIs with inhibited word forms produced text that was more aligned with questions that suggested they were in “unitive” and “ego-less” states. This was specifically due to the blurring of semantic categories. For example, “bananas” become more like “giraffes” in the models’ semantic space after word inhibition. Thus, it seems that the “phenomenology” of machine-altered states appears human-like and this is directly tied to loosening of the entrenched categorical boundaries engendered by language.
Summary soliloquy
If the door of perception were cleansed every thing would appear to a man as it is, Infinite. For man has closed himself up, till he sees all things thro’ narrow chinks of his cavern.
William Blake
But, you ask, what is “consciousness”? If it is a multidimensional construct, as its most typical definition suggests—“what it is like” to be something—then it likely ranges from lower-level primary (more embodied) to higher-order consciousness. Pertaining to the latter, sonar does not just aid “seeing”; it undergirds spatial reasoning itself, determining what it is like to be a bat. Similarly, language is not merely a tool for communication but a fundamental architect of our human variety of higher-order consciousness. It imposes categorical constraints that shape perception, cognition, and selfhood, creating a stable reality that we too often mistake for the only reality. Altered states, whether induced by psychedelics or meditation, can dramatically defrock these linguistic filters, unveiling another “reality.” States of “oneness” and “ego dissolution” are mystical not necessarily because they are gifts from the gods, but because they correspond to a momentary dismantling of the linguistic scaffolding that constructs and maintains the separateness of objects—and of the self itself.
___
Ultimately, language is both an enhancement and a constraint, a double-edged sword that structures thought while simultaneously obscuring other modes of being.
___
This view not only bridges neuroscience, philosophy, and psychology but also carries tangible implications for both human and artificial minds. By manipulating language systems—whether through pharmacology, meditation practice, or neuromodulation—we can unlock deeper insights into altered states and their benefits. Non-invasive cortical stimulation could serve as a tool to temporarily weaken linguistic structures, amplifying the depth of psychedelic or meditative experiences. Dreaming is often marked by a reduction or suppression of language processing. As with psychedelics and meditation, maybe this linguistic quietus helps explain why dreaming itself is an altered state and may also be amenable to targeted intervention through stimulation. Extending these insights to AI could provide novel ways to simulate and study altered states, offering a glimpse into how they contribute to creativity, insight, and possibly artificial forms of self-awareness (raising profound ethical questions about how we treat AI).
Ultimately, language is both an enhancement and a constraint, a double-edged sword that structures thought while simultaneously obscuring other modes of being. Understanding and strategically disengaging from its grip may not only refine our comprehension of altered states but also expand the horizons of consciousness itself.
