THE WORD “I”
“I” is one of the easiest words to use. It may also be one of the hardest to explain. I remember. I decide. I suffer. I see. I was there yesterday. I will be there tomorrow.
Grammar makes the speaker look simple: one word, one person, one continuous subject. Philosophy makes the word unstable. What exactly is the “I” in these sentences? A body? A consciousness? A memory, a soul, a point of view? A sequence or a convention? Something that cannot itself become an object? Or nothing over and above changing processes?
Lineage note · TERM
The English word “self” is doing dangerous work here. Ātman, attā, anattā and Hume's idea of a simple and identical self do not name one theoretical object.
Anattā negates attā; it is not another positive name for a thing. Comparison begins by allowing these terms and their problems to remain partially untranslatable.
Earlier Paragraphos Folios have already asked pieces of this question. We return to three of them here. Not because they give the same answer. Because they make the same word difficult in different ways.
The history of a question is not always the history of one answer.
A Lineage can follow documented transmission, reception or conceptual development. This one follows the transformation of a question across materials whose historical connection it does not establish. The order of these texts tells us when we are looking, not who inherited an answer from whom.
This is not a line of inheritance. It is a line of pressure.
Our question is what problem each text is trying to solve when it speaks about what we translate or summarize as “self”. The same human word, “I”, is placed under different philosophical pressures. The transformations below describe changes in our comparison, not a proven historical succession of doctrines.
THE KNOWER
- Time
- c. 7th–6th century BCE
- Place
- South Asia
Approximate placement of early Upaniṣadic material, not a date for this passage or an actual conversation.
You cannot see the seer of seeing.You cannot hear the hearer of hearing.You cannot think the thinker of thinking.You cannot know the knower of knowing.This is your self, the one within all.
Reused from Folio 0002 · THE SEER CANNOT BE SEEN. Its source and review limits are retained.
The pressure begins with the relation between what is known and the knower. Things appear. Thoughts, memories and sensations appear. But what about that to which they appear? Uṣasta has asked Yājñavalkya to explain the self within all, the brahman that is immediate and not indirect. Naming another object does not satisfy his demand.
Yājñavalkya does not point toward personality. The answer does not identify a biography, a preference, a remembered event or a character trait. Its repeated prohibitions make the self difficult to objectify. If I observe a thought about myself, is that thought the knower, or something known?
Every object you identify as the self seems to leave another question behind: what knows that object?
“Subject” and “object” are explanatory terms used here, not a claim that this Sanskrit passage speaks the vocabulary of modern philosophy.
FIRST TRANSFORMATION · WHAT HAS “SELF” BECOME?
An ordinary object is no longer an adequate model. Nor is personality, or one observed content among others. Perhaps the self is the condition or subject of knowing. This is a SUPPORTED INTERPRETATION offered as a Paragraphos reading, not an uncontested metaphysical fact. It draws on the passage's repeated prohibitions.
The early Upaniṣadic text must also remain distinct from its later Vedānta reception. Reading the knower as resistant to objectification does not place the complete system of Advaita inside these few lines. Yājñavalkya is not Śaṅkara. A later framework may interpret this passage, but cannot quietly become its original voice.
NOT MINE
- Time
- mid-first millennium BCE
- Place
- South Asia
Broad intellectual setting. The transmitted discourse is not assigned an exact composition year.
Form is not-self.Feeling is not-self.Perception is not-self.Formations are not-self.Consciousness is not-self.
Reused from Folio 0005 · THIS IS NOT MINE. Its source and review limits are retained.
The direction changes. SN 22.59 examines the familiar constituents of experience and repeatedly calls them not-self. The display condenses five predicates whose intervening arguments remain available in Folio 0005. Even consciousness is included.
In this discourse the search is not redirected toward a hidden knower. The aggregates are tested through affliction, control, impermanence and the appropriateness of identifying with them. Its refrain moves from description toward a way of regarding experience:
This is not mine.
I am not this.
This is not my self.
SN 22.59, the appropriation refrain. Paragraphos working translation, retained from Folio 0005.
Calling consciousness not-self matters for the comparison. We cannot simply remove body, feeling and thought from consideration and assume that consciousness must therefore be the owner left behind. The text includes it within the repeated examination.
Yājñavalkya's passage can make us ask about the knower that cannot be known like another object. SN 22.59 makes us ask why anything in this changing field should be appropriated as mine, I, my self.
From “Where is the subject?” to “What licenses identification?”
This shift is Paragraphos's reconstruction. The grammatical predication “not-self”, the practice of disidentification and a comprehensive philosophical theory of no-self are related, but one cannot stand in for the others without argument.
SECOND TRANSFORMATION · WHAT HAS “SELF” BECOME?
Something claimed, appropriated, identified with, expected to persist or to obey us. The problem is not exhausted by asking which inner object I am. It also tests the grammar of ownership through which changing experience becomes mine.
The self is no longer only something we search for. It is something we keep saying.
Mine. I. My self. The discourse places these claims under pressure in the movement toward disenchantment, dispassion and liberation. That practical purpose will matter when we reach Hume.
NOT A DEBATE THAT HAPPENED
It is tempting to place ĀTMAN opposite ANATTĀ and call this a debate. That is too simple. These texts belong to overlapping but distinct intellectual worlds. Historical relationships among Vedic, Upaniṣadic and early Buddhist traditions are serious subjects of research. This Lineage is not proving that SN 22.59 directly responds to Bṛhadāraṇyaka 3.4.2.
Lineage note · RELATION
Paragraphos distinguishes DOCUMENTED INFLUENCE, DOCUMENTED CONTACT, CONCEPTUAL PARALLEL and PARAGRAPHOS EDITORIAL COMPARISON.
This Lineage uses the last two. They name what the present comparison proposes. No direct historical influence between all three nodes is asserted; an absence of established contact here does not prove contact impossible.
Geographic and chronological proximity is not enough to establish textual influence.
Our comparison chooses two located passages and asks how their demands differ. It does not reconstruct a conversation between their speakers. South Asia is the location of these materials, not the name of one doctrine they must share.
MORE THAN TWO MILLENNIA LATER
Between these South Asian settings and Hume stand centuries, languages, religions, empires, translations and philosophical systems that this Folio is not attempting to compress into one chain. This absence is intentional. It is a limit of our inquiry, not an empty period in the history of thought.
A conceptual Lineage can jump across time. A historical lineage must account for the connection.
The space marks a long interval without pretending to measure it precisely. The next node locates a Scottish philosopher and a book published in 1739–1740. It does not assign his view to Europe as a whole.
WHAT DO I ACTUALLY FIND?
- Time
- 1739–1740 CE
- Place
- Scotland / Europe
Treatise Book I appeared in 1739; the Appendix followed in 1740. Scotland identifies Hume's intellectual setting, not the place where every page was written.
I never can catch myself at any time without a perception
Reused from Folio 0008 · THE SELF YOU CANNOT FIND. Its source and review limits are retained.
Hume reaches the difficulty by asking what impression could supply the idea of a simple, identical self continuing through a life. Introspection turns up particular perceptions: heat, pain, pleasure, thought. The additional, invariant impression required by that particular conception of self does not appear.
Here Hume begins with evidence, rather than the Buddhist discourse's liberative task.
What experiential content grounds this idea? That demand differs from asking what knows every object, or what should cease to be appropriated. Hume's project is a science of human nature, and this argument concerns the intelligibility and origin of an idea.
HUME MUST REMAIN DIFFICULT
The Treatise's bundle or succession of perceptions cannot be a slogan that finishes the problem. Relations among perceptions, association and imagination help Hume explain why identity is attributed to a changing mind. His Appendix acknowledges that his account of their connection remains unsatisfactory.
CONTESTED: whether this amounts to denying every subject of experience. Galen Strawson explicitly resists that interpretation; Don Garrett develops a positive account of Hume's idea of self. Their accessible abstracts establish that these alternatives exist, without giving us grounds to certify their full arguments. The exact source of Hume's Appendix difficulty also remains disputed. Folio 0008 records that unresolved scholarship.
The missing simple impression leaves a further question: what explains unity?
THIRD TRANSFORMATION · WHAT HAS “SELF” BECOME?
An idea requiring experiential grounding; a supposed simplicity; an identity attributed through time. The wider consequences of this account remain open to interpretation. “Hume proved that no self exists” would erase the difficulty we need to keep.
THREE DIFFERENT ABSENCES
These are Paragraphos formulations of the three pressures, not quotations or three versions of one argument.
- YĀJÑAVALKYA · WHAT IS THE KNOWER?
- The self is difficult to find because the knower cannot simply become another known object.
- EARLY BUDDHIST THOUGHT · WHAT SHOULD NOT BE APPROPRIATED?
- The constituents we identify with are impermanent, conditioned and not appropriately regarded as self.
- HUME · WHAT EVIDENCE SUPPORTS A PERSISTENT SELF?
- Introspection provides particular perceptions, not an additional simple, invariant impression corresponding to the enduring self under examination.
The similarity is the disappearance. The reasons for the disappearance are different.
Even “disappearance” needs care. Yājñavalkya does not withdraw the word ātman; he changes the demand to show it. The Buddhist discourse tests identification. Hume questions the evidential basis of a particular idea. What disappears is the ease with which we expected to locate the self, not necessarily the same entity in each case.
DO NOT LET THE WORD “SELF” DO THE COMPARISON FOR US
Yājñavalkya's context includes ātman, brahman, knowledge and immediacy, within early Upaniṣadic inquiry and its concern with liberative knowledge. The Buddhist setting includes anattā, impermanence, dukkha, appropriation and liberation from suffering and rebirth. Hume's includes impressions and ideas, association, imagination and the investigation of human nature.
Lineage note · LIMIT
A comparison becomes misleading when similarity erases purpose. Yājñavalkya, early Buddhist thought and Hume do not pursue the same philosophical project.
The Lineage stops wherever comparison requires us to pretend that they do. A shared absence is not enough to establish a shared argument.
A shared English noun does not create a shared metaphysics.
The comparison fails if it turns an unobjectifiable knower into a Buddhist remainder, makes SN 22.59 an empiricist test of ideas, or treats Hume's introspection as a practice with the discourse's liberative end. It also fails if one selected passage is made to speak for every later member of its tradition.
Buddhist thinkers developed divergent accounts of persons, causal continuity, conventional reality, mental streams, agency and karma. Later personalist and reductionist debates make that diversity explicit. Those arguments are not all contained in SN 22.59. Nor does this early Upaniṣadic passage contain every later account of ātman. Similarities become useful only while these differences remain in view.
IF THERE IS NO SIMPLE OWNER, WHY DOES A LIFE HOLD TOGETHER?
A person remembers, promises, acts, suffers consequences, makes plans and recognizes a history as theirs. Whatever happens to a proposed simple owner, these continuities still call for explanation. Failure to find one kind of entity does not explain why a promise matters tomorrow.
Lineage note · CONTINUITY
Rejecting a simple or permanent self does not make continuity disappear. Why does experience appear connected? Why do actions have consequences? Why does memory matter? Why can one moment belong intelligibly after another?
The answers differ. The problem persists. These questions do not make every selected passage a theory of identity through time.
The question in this heading is an editorial experiment, not a premise shared by all three texts. Yājñavalkya's passage is not primarily a theory of identity through time. SN 22.59 is not a complete account of Buddhist causal continuity. Hume does directly address attributed identity over time, but his account retains the difficulty he acknowledges.
Perhaps “What is the self?” contains several questions: What is the subject of experience? What constitutes a person? What persists? What owns experience? What makes responsibility possible? What does “I” refer to? They can bear on one another without having one answer. Contemporary discussions of personal identity likewise distinguish persistence, personhood and evidence rather than treating them as interchangeable.
ONE WORD
MANY PROBLEMS
SELF
- SUBJECT To whom does experience appear?
- PERSON What constitutes someone?
- IDENTITY What makes this the same person?
- OWNERSHIP What licenses “mine”?
- CONTINUITY What connects moments?
- AGENCY Who acts and answers for an action?
- SELF-KNOWLEDGE What could knowing oneself require?
This is a Paragraphos arrangement of problems, not a taxonomy attributed to the primary texts. Philosophical disagreement may persist partly because different arguments answer different members of this family.
A thinker can reject a permanent metaphysical substance without rejecting persons, agency, first-person experience or practical identity. Another can defend a subject of experience without claiming that personality never changes. Neither possibility settles the historical texts by itself. Each forces us to specify the claim before deciding whether another thinker contradicts it.
Before asking whether the self exists, ask which problem the word is being asked to solve.
PARAGRAPHOS PARALLEL
What makes experience mine?
- YĀJÑAVALKYA
- The knower cannot simply become one more object.
- EARLY BUDDHIST THOUGHT
- Ownership itself becomes philosophically suspect.
- HUME
- The supposed simple owner does not appear as a distinct, invariant impression.
No direct historical influence across all three is asserted. This question is ours. It lets the different pressures become visible without requiring that the texts share its exact formulation or give compatible answers.
DISTANCE AS WELL AS CONNECTION
A history might place these materials in separate chapters: Upaniṣadic inquiry, Buddhist philosophy, British empiricism. We can preserve those histories while asking what happens when we follow the problem across their divisions. The divisions contain knowledge too: languages, institutions, inherited arguments and purposes that a comparison can easily lose.
This is why following a problem requires a stated relationship type, historical caution and conceptual precision. It also requires the possibility of saying that the comparison stops here. The value of returning to an archive is not that everything suddenly agrees. A second reading can reveal that a familiar question had been doing several different kinds of work.
A map of philosophy should show distance as clearly as connection.
THE ONE WHO SAYS I
I woke up. I remember. I changed my mind. I was wrong. I will be there tomorrow. None of these inquiries makes such sentences disappear. The question is what we think stands behind them: a knower, a process, a conventional person, a sequence of related perceptions, something else?
Different philosophical traditions make different demands on the answer. Perhaps “I” is philosophically powerful precisely because ordinary language lets us use it long before we know what it means.
The word remains.
The theories change.
The question survives.
When you say “I”, what do you think the word is pointing to?
Sources / Apparatus
The source records below belong to the earlier Folios and are reused unchanged. This comparison does not certify their outstanding checks. New research is listed separately.
YĀJÑAVALKYA
Yājñavalkya, as represented in the Upaniṣad. Bṛhadāraṇyaka Upaniṣad, 3.4.2.
The supplied excerpt is retained against V. P. Limaye and R. D. Vadekar, Eighteen Principal Upaniṣads, vol. I (Poona, 1958), p. 218, Kāṇva 3.4.2, visually checked. Otto Böhtlingk's independent Mādhyaṃdina edition (St. Petersburg, 1889), p. 27, corresponding 3.5.1, supports the disputed verbal forms. This is a comparison of identified printed editions, not a fresh manuscript collation. Punctuation, unaccented spelling and five reading lines remain editorial.
Original Paragraphos working translation supplied for this publication. It is not attributed to a published translator.
This excerpt ends at eṣa ta ātmā sarvāntaraḥ; it omits ato 'nyad ārtam and Uṣasta's subsequent silence, so it does not present all of 3.4.2. Limaye–Vadekar and Böhtlingk print शृणुयाः and मन्वीथाः. Madhavananda's 1950 edition, p. 469, prints शृणुयात् in the second clause: a real printed discrepancy, probably a printing error, though its cause is not established. No manuscript support or intentional normalization for that reading was identified in the consulted material. The display follows the attested second-person शृणुयाः; शृणुयात् would be third person, not a sandhi equivalent. GRETIL has connected manvīthā, while the displayed मन्वीथाः has direct print support and follows a paused-clause convention. Grammar explains that distinction without proving GRETIL's editorial intention. GRETIL separates the root text from a later commentary attributed to Śaṅkara and warns that the commentary is not proofread. The broad early-material date is not a date for the narrated conversation.
The commissioning editor supplied and authorized publication of this Sanskrit excerpt and original working translation. No modern edition's apparatus or third-party translation is reproduced. This authorization covers the supplied material only.
EARLY BUDDHIST THOUGHT
Attributed to the Buddha. Saṁyutta Nikāya, Khandha Saṁyutta, SN 22.59, Anattalakkhaṇa Sutta.
SuttaCentral, Mahāsaṅgīti root text (pli/ms), Bilara published branch. Opening selections: segments 2.1, 3.1, 4.1, 4.2 and 5.1; refrain within 7.1. Checked against the source file on 1 October 2026.
Original Paragraphos working translation supplied for this publication. The five predications and three-part refrain are translated; the arguments in the commentary are identified paraphrases. This is not a reproduction of a named modern translation.
The Discourse on the Characteristic of Not-Self is an English descriptive title. The mid-first-millennium BCE frame concerns early Buddhist thought; the historical Buddha is commonly situated around the fifth century BCE. Canonical transmission and redaction continued later. No exact composition date for SN 22.59 or stenographic record of a historical speech is claimed.
SuttaCentral identifies original Pāli texts as public-domain material and dedicates its own original material to CC0. The selected text and full context below are attributed to its pli/ms source. Its spelling, including ṁ and …pe…, is retained. Line breaks in the display and refrain are editorial; the opening quotation mark is omitted from the isolated first sentence. The English wording is the supplied Paragraphos working translation.
DAVID HUME
David Hume. A Treatise of Human Nature, Book I, Part IV, Section VI · Of Personal Identity · T 1.4.6.3.
Hume Texts Online, edited by Amyas Merivale and Peter Millican. Its Treatise copytexts are the original 1739 volumes I and II and 1740 volume III; the Appendix appeared with the latter. Short quotations checked against the online text on 4 October 2026.
Only Hume's original English wording from the 1739 public-domain text is reproduced. These two short excerpts do not reproduce modern editorial commentary or a modern translation. The site's stated copytext and editorial policy were checked; no blanket licence over the site's apparatus is claimed.
Comparative relationships
These relationships are Paragraphos readings, commissioned for this conceptual Lineage. Dates and location do not establish transmission.
YĀJÑAVALKYA / EARLY BUDDHIST THOUGHT
Bṛhadāraṇyaka 3.4.2 makes the knower resistant to objectification; SN 22.59 tests identifying the five aggregates as self. Paragraphos asks how the task changes from finding the knower to examining appropriation.
These located passages support the comparison, not a documented debate. Shared geography and broad chronology do not prove that SN 22.59 responds directly to 3.4.2.
EARLY BUDDHIST THOUGHT / DAVID HUME
SN 22.59's aggregate predicates and appropriation refrain can be compared with Treatise 1.4.6.2–4 on the failure to find a simple, constant self among changing experiences. This bounded parallel retains the separate methods developed in Folio 0008.
Disidentification directed toward liberation differs from Hume's empiricist examination of the idea of self. Neither shared vocabulary nor similarity establishes transmission. Hume's wider commitments are contested.
YĀJÑAVALKYA / DAVID HUME
Bṛhadāraṇyaka 3.4.2 and Treatise 1.4.6.2–3 let Paragraphos compare what counts as finding a self. The first obstructs treating the knower as another known object; the second challenges a particular idea of self through the demand for an impression.
Failure to locate an object does different philosophical work in these texts. No equivalence of theories, actual conversation, contact or influence is asserted.
Additional research and verification scope
The primary passages, SN 22.59 refrain, Hume Appendix and accessible comparative references from Folio 0008 were rechecked for the bounded claims made here. The references below add support or identify further review; an abstract is not a checked chapter.
- Galen Strawson · Hume on Personal Identity · The Oxford Handbook of Hume
Oxford's indexed author abstract checked. It explicitly contests readings on which Hume denies subjects of experience or endorses a simple no-ownership view. This supports the limited report of interpretive disagreement. Full chapter access was unavailable; its detailed case has not been verified or adopted.
- Don Garrett · The Idea of Self in Hume's Treatise · The Self (2021)
Oxford's indexed author abstract and bibliographic record checked. Garrett proposes a positive account of the idea of self. This is an identified alternative for review, not a verified reconstruction of his argument. Full chapter access was unavailable.
- Jay L. Garfield · Losing Ourselves (Princeton University Press, 2022)
Publisher-supplied JSTOR book record and description checked for an explicitly argued distinction between selves and persons. Listed as further comparative reading. The full book has not been checked; no page-specific claim is based on it, and its position is not presented as the consensus of Buddhist traditions.
- Eric T. Olson · Personal Identity · Stanford Encyclopedia of Philosophy
Section 1 rechecked for the separation of characterization, personhood, persistence, evidence and personal ontology. The seven-part question family in this Lineage is Paragraphos's own arrangement, not Olson's taxonomy or an attribution to the three primary texts.
- Christian Coseru · Mind in Indian Buddhist Philosophy · Stanford Encyclopedia of Philosophy
Sections 5.6 (Persons: Reductionism and Supervenience) and 6.3 (Nominalism) rechecked for personalist/reductionist disagreement and the distinction between denying a substantial self and retaining an account of persons. These are later debates. They are not silently imported into SN 22.59.