Understanding Knowledge through Comparison: Upamana in Indian Thought

How do we come to know something we have never directly seen or experienced? This is one of the oldest puzzles in philosophy, and Indian thinkers developed remarkably precise answers to it centuries ago. One of those answers is upamana – the pramana of comparison. As a recognized means of valid knowledge (pramana) in Indian epistemology, upamana explains how we acquire genuine knowledge of the relationship between a word and the object it names – purely through the recognition of similarity. It is not guesswork, and it is not inference in the ordinary sense. It is a distinct, irreducible way of knowing.

What does upamana mean? 🔗

The word upamana is a compound of two Sanskrit roots: upa, meaning similarity or resemblance (sadrsya), and mana, meaning cognition. Together, they give us “knowledge through similarity.” Upamana is formally defined as the knowledge of the relationship between a word and the object that word denotes – and this knowledge is produced through the recognition of resemblance, given a prior description of the unfamiliar object. In simpler terms: someone tells you what an unknown thing is like; later, when you encounter it, the click of recognition that connects the description to the object in front of you – that moment is upamana at work.

This is not the same as inference (anumana), which moves from a general rule to a conclusion. Upamana does not reason from a universal law. It operates through perceived similarity between a known and an unknown entity, and it yields knowledge that neither direct perception nor inference alone could produce.

The classic example: the gavaya 🔗

Indian philosophers have long used a single, vivid example to illustrate how upamana functions. Imagine you have never seen a gavaya – a wild ox. A forester, who knows the forest well, tells you: “A gavaya is an animal like a cow, but larger, wilder, and with a bigger horn on its forehead.” You have no direct experience of the gavaya, but you know what a cow looks like. You store that description in memory.

Some time later, you are walking through the forest. You spot an unfamiliar animal. It resembles a cow – yet it is bigger and fiercer, with the distinctive horn you were told about. In that moment, something decisive happens: you recognize the animal as a gavaya. This recognition – that the animal before you is indeed a gavaya – is the result of upamana. You have gained knowledge of the connection between the word “gavaya” and the creature it refers to, through comparison with an already-known object.

Notice what makes this special. You did not perceive the gavaya directly before. You did not arrive at it through a logical syllogism. The knowledge came specifically from noticing resemblance in a real encounter, informed by a prior description. That is the epistemological structure of upamana.

The three stages of upamana as a knowledge process 🔗

Upamana is not a single mental act – it unfolds in a sequence. Indian philosophers typically identify three stages in the process:

The first stage is receiving a description of similarity. Someone with first-hand knowledge tells you that an unfamiliar object resembles something you already know. This description comes through verbal testimony (shabda), but it is only a preparatory step – it does not itself constitute upamana.

The second stage is encountering the unfamiliar object. At some later point, you come face to face with the thing that was described. This encounter activates the stored description in your memory.

The third stage is the moment of recognition: you perceive the similarity between the object before you and the known object from your prior experience, and you thereby grasp the connection between the word (“gavaya”) and what it denotes. This is the actual epistemic event – the knowledge produced by upamana.

What makes upamana philosophically significant is the claim that this three-part process generates a type of knowledge that is not reducible to any of the other pramanas. It is not mere perception, not inference, and not simply the verbal testimony that initiated the process.

Upamana and the word-object relationship 🔗

One of the most important contributions of upamana to Indian epistemology is its role in explaining how we understand the relationship between words and their denotations. This is not a trivial matter. Language depends on words correctly pointing to the things they name. But how does a person first grasp that a word applies to a specific class of objects?

According to the Nyaya school, upamana is precisely the means by which this word-object relationship is first cognized. When you recognize that “gavaya” applies to the animal in front of you – based on similarity to a cow – you have come to know the denotative relationship of that word. This is knowledge that could not have been established by perception alone (since you had never seen a gavaya) nor by inference alone (since there is no universal rule linking the word “gavaya” to its referent in the way inference requires).

The formal terminology in Sanskrit is precise here: the upameyam is the subject of comparison (the unfamiliar object), the upamanam is the object it is being compared to (the familiar one), and the samanya refers to the shared attribute or quality that makes the comparison valid. As Monier Williams explains, if one says “her face is like the moon in charmingness,” the face is the upameyam, the moon is the upamanam, and charmingness is the samanya. The structure is always: something unknown is made intelligible through something known, via an identifiable shared quality.

Upamana across Indian philosophical schools 🔗

Not every school of Indian philosophy accepts upamana as an independent and valid pramana. The debate over its status is itself an important chapter in Indian epistemological history.

The Nyaya school 🔗

The Nyaya school explicitly accepts four pramanas: perception (pratyaksha), inference (anumana), comparison (upamana), and testimony (shabda). For Nyaya philosophers, upamana is irreplaceable because it explains something that neither perception nor inference can account for – namely, how a cognitive subject first grasps the referential relationship between a word and its denotatum. Nyaya restricts the scope of upamana specifically to this word-meaning function, treating it as a precise and limited epistemic tool rather than a broad analogical reasoning faculty.

The Mimamsa school 🔗

The Mimamsa school also accepts upamana, though with a somewhat different understanding of its scope. The Prabhakara sub-school of Mimamsa accepted upamana alongside perception, inference, postulation (arthapatti), and testimony as valid means of knowledge. Mimamsa thinkers, like Nyaya thinkers, recognized upamana as the knowledge of similarity – generated when an unfamiliar object is brought into a recognized relationship with a familiar one. Kumarila Bhatta’s sub-school went further and accepted six pramanas in total, including upamana within that expanded framework.

Advaita Vedanta 🔗

Advaita Vedanta accepts upamana as one of its recognized pramanas – describing it as the knowledge gained by means of analogy and comparison. It is understood as the cognitive process by which a similarity between two objects generates valid knowledge of what the unknown object is. The Advaita acceptance of upamana reflects the school’s broadly inclusive approach to the means of valid cognition.

Schools that reject or absorb upamana 🔗

Not all traditions grant upamana independent status. Schools such as the Vaisheshika, Samkhya, Yoga, Jain, and Buddhist traditions generally subsume upamana under either inference or perception, arguing that comparison is not epistemically distinct enough to merit its own category. Buddhist epistemologists like Dignaga and Dharmakirti, working within a strict two-pramana framework of perception and inference, reduce upamana to a form of perceptual or inferential cognition rather than accepting it as a separate source of knowledge.

This debate – whether upamana is genuinely irreducible or is merely a variant of inference – remains one of the productive tensions in classical Indian epistemology. Those who defend its independence argue that the specific cognitive event of recognizing word-meaning through similarity has a unique structure that inference cannot capture. Those who reject it argue that any “recognition through resemblance” can always be reformulated as an inferential process.

Why upamana matters epistemologically 🔗

Upamana addresses a real gap in how we acquire knowledge. Much of our understanding of the world is built not through direct encounter, but through structured comparison with what we already know. When we encounter an unfamiliar species, a new concept, or an abstract term, we instinctively reach for something familiar – and the recognition of similarity generates genuine cognitive content, not mere metaphor.

Upamana helps generalize concepts across individuals of varied learning and background – making it an instrument not just for philosophical reasoning, but for communication, teaching, and scientific classification. In Ayurvedic medicine, for instance, upamana was used extensively to explain the features of unmanifested conditions through comparison with known, observable phenomena – a testimony to how practically the concept was applied beyond pure philosophy.

The 7th-century text Bhaṭṭikāvya dedicated substantial discussion to identifying when comparison-based knowledge is most reliable and when it is not – acknowledging that upamana, like all pramanas, is a conditional and fallible means of knowing. Ancient and medieval Indian texts debated as many as 32 types of upamana, reflecting how seriously this method of cognition was taken across traditions.

At its core, upamana insists that similarity is not merely rhetorical – it is epistemic. When two things are genuinely alike in a relevant way, and that likeness allows you to identify and name what was previously unknown, you have not just spoken elegantly. You have come to know something real.

What do you think? If the recognition of similarity is itself a genuine source of knowledge, where do we draw the line between an upamana that yields real understanding and one that merely creates a misleading impression? And do you think modern disciplines – linguistics, cognitive science, or medicine – implicitly rely on upamana-like reasoning without naming it as such?

References

1. https://www.britannica.com/topic/pramana
2. https://www.newworldencyclopedia.org/entry/Pramana
3. https://raajacademy.com/the-six-pramanas-comprehensive-ways-of-knowing-in-indian-philosophy/
4. https://en.wikipedia.org/wiki/Nyaya
5. https://en.wikipedia.org/wiki/Hindu_philosophy
6. https://www.researchgate.net/publication/361182186_Upamana_pramana
7. https://www.jetir.org/papers/JETIR2507672.pdf
8. https://www.carakasamhitaonline.com/index.php?title=Upamana_pramana
9. https://en.wikipedia.org/wiki/Pramana