Śruti and Smṛti

Two words do most of the sorting in Hindu textual tradition, and both of them are about transmission rather than about quality.

Śruti, "what is heard", names the Veda. Smṛti, "what is remembered", names a great deal of what came after. Almost every short account then adds a ranking — śruti above smṛti — and stops there. The ranking is real. But taken alone it produces two misunderstandings that are worth dismantling before anything else, because nearly everything interesting about this pair lives underneath them.

Two words, and what they actually name

Start with what each category holds.

Śruti is the Vedic corpus, and it is larger than the word "Veda" often suggests in casual use. It includes the hymn collections, the ritual manuals attached to them, the forest-texts and the Upaniṣads. All of that is śruti.

Smṛti, in its strict and original sense, is narrower than the textbooks imply — we will come back to that. In ordinary modern use it covers the law-books, the two epics, the Purāṇas, and much else.

Now the first misunderstanding. It is tempting to hear śruti and smṛti as "sacred" and "less sacred", or "God's word" and "human books". Neither pair is what the terms assert. Both words describe a route: one thing was heard, the other is held in memory. What the tradition builds on top of that difference is not a scale of holiness. It is an argument about evidence.

Where the authority of the remembered books comes from

Here is the question the short accounts skip. If the Veda is the authority, why does a remembered text have any authority at all?

The classical answer is one of the most interesting pieces of reasoning in the tradition, and it is set out in the Mīmāṃsā school, which exists to work out how the Veda is to be understood and applied.

The objection is stated first, and stated brutally: since dharma is based on the Veda, whatever is not Veda should be disregarded. The remembered texts are the work of human authors, and human beings can be wrong.

The answer runs like this. When a respected text says that a certain rite is to be performed, and the people who preserve and follow that instruction are the same people who preserve the Veda, the best explanation is that there was a Vedic passage behind it. Perhaps we still have that passage. Perhaps it was known to the author of the remembered text and has since been lost — and after the last lesson, with its thirteen surviving branches out of a thousand, "since been lost" is not a convenient excuse but an ordinary description of what happened.

Follow what that argument does. It does not say smṛti is a second-rate authority. It says smṛti has no authority of its own at all — and that it carries Vedic authority, at one remove, by inference.

That is a stronger claim and a more fragile one at the same time, and it is the key to everything that follows.

Which is why śruti wins a conflict

Now the famous rule stops being arbitrary.

Where a remembered rule and a heard passage conflict, the remembered rule is set aside. The reasoning given is exact: in a conflict, the source of the Vedic instruction is there in front of us and fully recognised, while the source of the remembered rule "has got to be found" — it is an assumption. And the assumption was only ever made because nothing contradicted it. Once a Vedic passage does contradict it, there is no longer any reason to assume the lost text existed.

So śruti does not win because it outranks. It wins because the case for the other side has just collapsed.

The same reasoning produces something a modern reader rarely expects: a test for setting a remembered rule aside. If a visible worldly motive explains why a rule exists, no lost Vedic text needs to be assumed behind it. The classical example is unsentimental — priests who stood to gain a large piece of cloth were the ones who found a rule requiring the sacrificial post to be covered with cloth.

A tradition that builds that test into its own theory of scripture is not a tradition that accepts everything it is handed.

What "smṛti" meant before the textbooks widened it

Now the terminological correction, which matters more than it sounds.

The Manusmṛti, setting out the roots of dharma, gives a definition: śruti is to be known as the Veda, and smṛti as the dharmaśāstra. In that formulation smṛti does not mean "all non-Vedic Hindu literature". It means the legal and normative literature specifically — the class of texts to which the Manusmṛti itself belongs.

The wider modern usage, where smṛti becomes a bin containing the epics, the Purāṇas and everything else, is a later broadening, and it is the usage this lesson has been using so far because it is the usage you will meet. It is serviceable. It is just not what the classical definition says, and when you see a sentence like "the Purāṇas are smṛti", it is worth knowing that the sentence is using the wide sense.

The oldest texts do not sort themselves this way

If the two-category scheme were simply built into the tradition from the beginning, we would expect the oldest texts to observe it. They do not, and the evidence is easy to check.

In the Chāndogya Upaniṣad, a student named Nārada is asked what he has already studied. He lists the Ṛgveda, the Yajurveda, the Sāmaveda, the Atharvaṇa as the fourth — and then itihāsa-purāṇa, story and old-tale, "as the fifth Veda". Not as an appendix. As the fifth item in the row.

The Bṛhadāraṇyaka Upaniṣad does something similar in a famous image: what has been breathed out from the great being is the Ṛgveda, the Yajurveda, the Sāmaveda, the Atharvāṅgirasa — and then history, old tales, the sciences, the Upaniṣads, verses, aphorisms, explanations and commentaries. Commentaries are in the list.

Both of those passages are śruti. Śruti, in other words, does not draw the line where the later scheme draws it.

That does not make the scheme wrong. It makes it what it is: a classification the tradition worked out about itself, over time, for good reasons — not a boundary that was always there and that the earliest texts were observing.

Rank is not importance

Which brings us to the second misunderstanding: that being in the lower category means mattering less.

The clearest counter-example is the one text almost everyone reaches for first. The Bhagavadgītā sits inside the Mahābhārata and is, formally, smṛti. It is also one of the three foundations on which the Vedānta schools build, and Śaṅkara, Rāmānuja and Madhva each wrote a full commentary on it — the same three thinkers who wrote on the Upaniṣads. A text can be formally subordinate and central at once.

And beyond the Vedāntic mainstream the picture is wider still. Śaiva and Vaiṣṇava communities have their own revealed corpora, the Āgamas, which their traditions treat in scriptural terms; a flat statement that "scripture means the Veda and nothing else" describes one position rather than the whole field.

Then there is ordinary life, where the ranking barely appears. Most Hindus meet the tradition almost entirely through the wide smṛti layer and through practice — a story, a festival, a family observance. Śruti is where authority is formally located. Smṛti is where nearly all of the religion is actually lived, and calling that a lesser encounter would be a description no one inside it would recognise.

What to carry out of this

Three things.

The pair is about routes, not about worth. One category is defined by having been heard, the other by being remembered. Anything you are told about their relative holiness is an addition to what the words say.

Smṛti's authority is Vedic authority, inferred. That is why śruti wins a conflict, and it is also why a remembered rule can be set aside when a worldly motive for it is visible. Both halves come from the same argument, and the second half is the one usually left out.

The scheme is a later achievement, and it does not cover everything. The oldest texts list story and commentary beside the Vedas; whole communities give scriptural weight to texts outside the Vedic corpus; and the text people actually read most is technically in the lower category. Use the classification. Do not mistake it for a map of what matters.