The Chinese Āgamas: Forgotten Discourses from Early Buddhism

How the Buddha’s early teachings were recited, transmitted across Buddhist communities, translated into Chinese, and rediscovered through modern comparative research.

Many English-speaking students of Buddhism first encounter early Buddhism through the Pāli Canon. They read the Nikāyas and meet a world of wandering monks, forest dwellers, kings, merchants, debates, meditation, illness, death, and liberation. The voice of these texts can feel direct and ancient. It is natural to think: this is early Buddhism.

But the story is wider.

Another great body of early Buddhist discourses was preserved not in Pāli, but in Chinese translation. These are the Āgamas. They were not lost. They were translated, copied, catalogued, and preserved in the Chinese Buddhist canon. Yet for much of East Asian Buddhist history, they were overshadowed by later Mahāyāna scriptures and remained under-read.¹

Today, through comparative research, the Chinese Āgamas have re-emerged as one of the most important witnesses to the early Buddhist teachings.²

After the Buddha: preserving a teaching without books

Map of Ancient Kusinara

The story begins not with a library, but with a death.

When the Buddha passed away, the early community faced a profound question: how could the teaching survive when the Teacher was gone? In the world of ancient India, preservation did not first mean writing books, printing editions, or establishing archives. It meant remembering, reciting, checking, and transmitting together.

Buddhist traditions remember that soon after the Buddha’s passing, senior monks gathered at Rājagṛha to recite and organize the teachings. The historical details of this First Council are debated, but the memory itself is significant. In traditional accounts, the early community gathered in order to preserve orally the core teachings of the Buddha.³

In this memory, Ānanda, the Buddha’s attendant and the disciple renowned for hearing many discourses, became associated with the recitation of the suttas or sūtras. Upāli, known for his mastery of monastic discipline, became associated with

the recitation of the Vinaya.⁴

Whether every detail can be historically verified is less important than the central point: from an early stage, the Buddha’s teaching was treated as something to be preserved carefully, collectively, and publicly.

Oral transmission as collective memory

To modern readers, oral transmission can sound unreliable. We live in a world where writing, recording, and digital storage feel more trustworthy than memory. But in ancient India, oral preservation could be a rigorous discipline.

The early discourses themselves bear the marks of an oral culture. They repeat phrases. They arrange teachings in lists. They use standard openings, recurring formulas, and patterned structures. These are not literary defects. They are memory technology.

Sujato and Brahmali summarize several features of early Buddhist oral preservation: repeated words, phrases, passages, and even whole discourses; standardized expressions; summary-and-exposition structures; framing narratives; verse summaries; numbered lists; similes; and group recitals.⁵ These features helped monks learn, recite, verify, and transmit the teachings across generations.

A helpful modern analogy is to think of this as a distributed memory system. A teaching was not preserved simply because one individual claimed to remember it. It was preserved because a community could recognize it, repeat it, compare it, and confirm it together. In that sense, early Buddhist transmission worked through collective verification rather than private ownership.

This does not mean the texts reached us as word-for-word recordings of the Buddha. They were edited, arranged, and transmitted over time. A careful position is to say that the Āgamas and Nikāyas are not tape recordings of the Buddha, but conservative textual memories preserved by disciplined communities.⁶

From recitation to collections

Over time, recited teachings were organized into collections. In the Theravāda tradition, these collections were preserved in Pāli as the Nikāyas. In other early Buddhist lineages, related collections were preserved in Indic languages and eventually translated into Chinese as the Āgamas.

The relationship between the Nikāyas and Āgamas is best understood through one phrase:

They are sibling traditions, not original and copy.

The Chinese Āgamas often parallel the Pāli Nikāyas, but they are not simply Chinese translations of the Pāli Canon. They come through different school lineages, different transmission histories, and different translation contexts.

A basic comparison looks like this:⁷

Chinese Āgama	Rough Pāli Parallel	Character
Dīrgha Āgama	Dīgha Nikāya	Long discourses
Madhyama Āgama	Majjhima Nikāya	Middle-length discourses
Saṃyukta Āgama	Saṃyutta Nikāya	Connected or grouped discourses
Ekottarika Āgama	Aṅguttara Nikāya	Numbered discourses

These correspondences are approximate, not exact. The collections share many parallel discourses, but they differ in arrangement, wording, school affiliation, and transmission history. The BDK introduction to the Dīrgha Āgama notes many close parallels between the Sanskrit/Chinese Dīrgha Āgama tradition and the Pāli Dīgha Nikāya, while also emphasizing differences in ordering and arrangement.⁸

Those differences are not a problem. They are precisely what makes comparison valuable.

The long road to China

A Plausible Overland Route of Agama Transmission to China

The Āgamas did not arrive in China as a single sealed canon. They came through living networks: Indian monks, Central Asian routes, reciters, translators, patrons, and Chinese translation teams.

By the early centuries of the Common Era, Buddhist communities had spread across North India, Central Asia, and into China. Some transmission lines were connected with early schools such as the Dharmaguptaka, Sarvāstivāda, and Mūlasarvāstivāda. The Dīrgha Āgama, for example, is usually associated with the Dharmaguptaka school.⁹

Translation in China was often collaborative. An Indic or Central Asian monk might recite or explain the text; Chinese collaborators would help render it into Chinese; editors and scribes would prepare the finished version. This was not simply mechanical word conversion. It was a meeting of oral recitation, scholastic memory, Indic Buddhist language, and Chinese literary culture.

The major Chinese Āgamas were translated mainly between the late fourth and early fifth centuries:

• The Madhyama Āgama was first translated in 384 and later retranslated in 398.

• The Ekottarika Āgama was translated around 384–385, though its textual history is especially complex.

• The Dīrgha Āgama was translated in 413 by Buddhayaśas and Zhu Fonian.

• The Saṃyukta Āgama was translated in 435, based on a recitation by Guṇabhadra and rendered into Chinese by Baoyun.¹⁰

The story of the Dīrgha Āgama shows how translation depended on historical circumstances. According to the BDK introduction, Buddhayaśas was invited to Chang’an under Later Qin patronage; in 413 he translated the Dīrgha Āgama with Zhu Fonian, and the work was completed that same year.¹¹

To read the Chinese Āgamas, then, is to hear an early Buddhist voice across several layers of transmission: oral recitation, school preservation, Indic language, Central Asian movement, and Chinese translation.

That does not make them unreliable. It makes them historical.

Preserved, but under-read

The Chinese Āgamas were never lost. They were preserved inside the Chinese Buddhist canon. But they did not become the center of East Asian Buddhist imagination.

In China, Korea, Japan, and Vietnam, later Buddhist traditions came to be shaped more visibly by Mahāyāna scriptures and schools: the Lotus Sūtra, the Prajñāpāramitā literature, the Avataṃsaka Sūtra, Pure Land texts, Chan/Zen literature, and tantric materials. In doctrinal classification systems such as Tiantai, the Āgamas were often placed in an early or foundational stage of the Buddha’s teaching, associated especially with śrāvaka practice.¹²

This did not mean the Āgamas were rejected as non-Buddhist. They remained canonical. But they were often treated as preliminary rather than central.

That is why “forgotten” is useful, if we understand it carefully. The Āgamas were not forgotten in the sense of being lost. They were forgotten in the sense of being preserved but under-read.

Modern rediscovery through comparison

Modern scholarship changed the situation. When scholars and scholar-monks began comparing the Chinese Āgamas with the Pāli Nikāyas, they found not a marginal curiosity, but a major witness to early Buddhist discourse literature.

Bhikkhu Anālayo has been especially important in this field. His studies compare Pāli discourses with Chinese, Sanskrit, Tibetan, and other parallels. Peter Skilling, in his foreword to Anālayo’s Ekottarika-āgama Studies, emphasizes that Anālayo works with Chinese, Sanskrit, Tibetan, and Pāli materials side by side, and that this comparative method helps move beyond older assumptions.¹³

The Saṃyukta Āgama is one of the strongest examples. Anālayo’s Saṃyukta-āgama Studies notes that the Chinese Saṃyukta Āgama translation, preserved as Taishō 99, was begun in 435 based on an original recited by Guṇabhadra and appears to stem from a Mūlasarvāstivāda line of transmission. It also states that the reconstructed Saṃyukta Āgama shares the same basic fivefold division as the Pāli Saṃyutta Nikāya, and that more than two thirds of its discourses have parallels in the Saṃyutta Nikāya.¹⁴

This is why the Āgamas matter. They allow us to see early Buddhist teachings through more than one window.

Sometimes the Chinese and Pāli versions are remarkably close. Sometimes they differ in wording, order, or narrative frame. Sometimes one tradition preserves a simpler version, while another has expanded. Sometimes both preserve a feature that may already have been fixed before the traditions separated.

Modern comparative research does not give us perfect certainty. It gives us something better than sectarian certainty: a disciplined way to compare independent witnesses.

The role of Bhikkhu Bodhi

Bhikkhu Bodhi’s work is also essential, though in a different way. His translations of the Pāli Nikāyas have made the early discourses widely available to English-speaking readers. His editions of the Majjhima Nikāya, Saṃyutta Nikāya, and Aṅguttara Nikāya have become major entry points for modern readers of early Buddhism.

Sujato’s textual history of the Majjhima Nikāya describes Bhikkhu Ñāṇamoli’s translation, edited and updated by Bhikkhu Bodhi in 1995, as reaching “a peak of accuracy, consistency, and readability” that later Nikāya translations have aspired to.¹⁵ Sujato also describes Bodhi’s full translation of the Aṅguttara Nikāya as a major advance in accuracy and readability, introducing the collection to a new generation.¹⁶

For readers of the Āgamas, Bodhi’s work provides a crucial companion: clear access to the Pāli side of the early discourse tradition. Anālayo’s work then helps show how those Pāli discourses relate to their Chinese and other parallels.

Together, these scholar-monks have helped make EBT study both accessible and rigorous.

A necessary caution

The Āgamas should not be romanticized as untouched archives of “original Buddhism.” They are transmitted texts. Each collection has its own history.

The Ekottarika Āgama is especially complex. Anālayo’s Ekottarika-āgama Studies notes debates about its school affiliation, translator identity, translation quality, consistency, possible Mahāyāna influences, and interpolations.¹⁷ Elsewhere in the same study, Anālayo argues that the Ekottarika Āgama shows a particular tendency to incorporate new material, including cases of discourse merger and material likely added in China.¹⁸

This does not make the Ekottarika Āgama worthless. It makes careful study necessary.

The same principle applies more broadly. The Āgamas and Nikāyas are not pure, untouched fossils. They are living textual inheritances. Their value lies not in pretending they have no history, but in studying that history carefully.

Early Buddhism is not simply Theravāda

For many English-speaking Buddhists, “early Buddhism” is often unconsciously equated with Theravāda. This is understandable: the Theravāda tradition preserved a vast and precious Pāli canon, and modern Theravāda teachers and translators have done extraordinary work in making that material available.

But early Buddhism is not identical with Theravāda.

Bhikkhu Sujato defines Early Buddhism as the teachings of the Early Buddhist Texts (EDT) —canonical discourses codified in the Buddha’s lifetime or shortly thereafter and passed down in Pāli, Chinese, Tibetan, and Sanskrit. He distinguishes this from Theravāda, the school established at the Mahāvihāra in Anuradhapura, Sri Lanka, which later spread across Southeast Asia and treats the Pāli Tipiṭaka as authoritative.¹⁹

This distinction matters. Theravāda preserved early materials, but it also developed commentarial, Abhidhamma, ritual, institutional, and meditation traditions. East Asian Buddhism preserved the Chinese Āgamas, but later centered itself largely around Mahāyāna frameworks.

The Āgamas help us see beyond a Pāli-only or Mahāyāna-only picture. They show that early Buddhist discourse literature was broader than any one later school.

Why the Āgamas matter now

The Āgamas matter for history. They matter for textual study. They matter for understanding how Buddhist teachings were preserved across languages and communities.

But they also matter for practice.

The Āgamas repeatedly return to a simple but demanding question: how does suffering arise, and how can it cease?

They teach the four noble truths, the noble eightfold path, dependent arising, the five aggregates, the six sense bases, mindfulness, ethical restraint, meditation, and liberation. Their voice is often spare, direct, and practical. They do not always offer grand cosmic visions. Often they ask us to look carefully at body, feeling, mind, perception, craving, contact, aging, death, and release.²⁰

For readers shaped by Zen or East Asian Buddhism, the Āgamas offer something especially valuable. They are not foreign imports from another Buddhist world.

They are part of the Chinese Buddhist inheritance itself.

To read them is not to abandon Mahāyāna. It is not to replace the Pāli Canon. It is not to create another sectarian identity.

It is to recover a wider memory of the Buddha’s teaching.

Not lost, but waiting

The Chinese Āgamas were not lost. They were waiting in plain sight.

Preserved in Chinese translation, overshadowed by later traditions, and now illuminated by comparative research, they invite us to hear the early Buddhist discourses with a wider ear. The Pāli Nikāyas remain indispensable. But they are not alone. Alongside them stand the Chinese Āgamas: parallel witnesses, transmitted through different communities, speaking in another language, yet often pointing back to the same early questions.

What is suffering?

What is its origin?

What is its cessation?

What is the path?

The Āgamas do not merely tell us what early Buddhists believed. They show us how a community remembered, recited, preserved, translated, and practiced a path of liberation across centuries.

They are forgotten discourses only if we do not read them.

Footnotes

1. The phrase “preserved but under-read” is a synthetic description of the historical situation: the Chinese Āgamas were preserved in the Chinese canon, while later East Asian Buddhist traditions often centered on Mahāyāna scriptures and doctrinal systems. The Chinese Āgama collection’s own introductory materials note that the four Āgamas were preserved in the Chinese Buddhist canon, while also describing the later tendency in Chinese Buddhism to marginalize them under “Hīnayāna” or śrāvaka classifications. ↩

2. Sujato and Brahmali define Early Buddhist Texts as including the bulk of the suttas in the four main Pāli Nikāyas and the parallel Āgama literature preserved in Chinese, Tibetan, Sanskrit, and other Indic languages. ↩

3. The BDK introduction to the Dīrgha Āgama says the First Council was traditionally held at Rājagṛha after Śākyamuni’s passing to ensure the oral preservation of his core teachings. ↩

4. Traditional accounts associate Ānanda with the recitation of the discourses and Upāli with the Vinaya. The uploaded Chinese Āgama introduction also preserves this memory of Ānanda reciting the sūtra collection and Upāli reciting the Vinaya at the early communal compilation. ↩

5. Sujato and Brahmali’s discussion of oral transmission lists repetitions, standardization, summary-and-exposition structures, framing narratives, verse summaries, numbered lists, similes, and group recitals as features of the early Buddhist oral tradition. ↩

6. Sujato and Brahmali explicitly argue that the EBTs are not verbatim records of the Buddha’s utterances; rather, they were edited and arranged over time, while preserving the central doctrinal substance. ↩

7. The BDK Dīrgha Āgama introduction provides a comparative chart between the four Sanskrit Āgamas and the five Pāli Nikāyas, listing the broad correspondences between Dīrgha/Dīgha, Madhyama/Majjhima, Saṃyukta/Saṃyutta, and Ekottarika/Aṅguttara. ↩

8. The same BDK introduction notes that the Sanskrit/Chinese Dīrgha Āgama and the Pāli Dīgha Nikāya have many synoptic parallels, while differences in ordering and arrangement must be recognized. ↩

9. The BDK introduction identifies the Dīrgha Āgama as belonging to the Dharmaguptaka tradition and explains that the Dharmaguptaka school inherited the Dīrgha Āgama as part of its Sūtra-piṭaka. ↩

10. The uploaded Chinese Āgama collection gives the basic translation dates: Madhyama Āgama first in 384 and retranslated in 398; Ekottarika Āgama around 385; Dīrgha Āgama in 413; and Saṃyukta Āgama in 435. ↩

11. The BDK Dīrgha Āgama introduction states that in 413 Buddhayaśas translated the Dīrgha Āgama with Zhu Fonian and that the translation was completed in the same year. ↩

12. This paragraph summarizes the later East Asian reception of the Āgamas. The uploaded Chinese Āgama introduction describes how Chinese Buddhism came to be shaped strongly by Mahāyāna currents and how the Āgamas were often treated as “Hīnayāna” or śrāvaka materials despite being preserved in the canon. ↩

13. Peter Skilling’s foreword to Anālayo’s Ekottarika-āgama Studies notes that Anālayo draws on Chinese, Sanskrit, Tibetan, and Pāli materials side by side and uses current research to present balanced assessments of the Ekottarika Āgama. ↩

14. Anālayo’s Saṃyukta-āgama Studies states that T99 was begun by Baoyun in 435, based on an original recited by Guṇabhadra, probably from a Mūlasarvāstivāda line of transmission; it also says the reconstructed Saṃyukta Āgama follows the same basic fivefold division as the Pāli Saṃyutta Nikāya and that more than two thirds of its discourses have Pāli parallels. ↩

15. Sujato’s textual history of the Majjhima Nikāya says that the Ñāṇamoli translation, edited and updated by Bhikkhu Bodhi in 1995 as The Middle Length Discourses of the Buddha, reached a high standard of accuracy, consistency, and readability. ↩

16. Sujato’s textual history of the Aṅguttara Nikāya says Bhikkhu Bodhi’s 2012 complete translation, The Numerical Discourses of the Buddha, was a major advance in accuracy and readability and introduced the Aṅguttara to a new generation. ↩

17. Skilling’s foreword to Ekottarika-āgama Studies summarizes the main problems surrounding the Ekottarika Āgama: debates over school affiliation, translator identity, translation consistency, and possible Mahāyāna influences or interpolations. ↩

18. Anālayo discusses discourse merger and material added in China in the Ekottarika Āgama, concluding that these problems prevent the collection from being assigned too simply to a single early school lineage. He also identifies recurrent Mahāyāna-related influence in the collection. ↩

19. In How Early Buddhism Differs From Theravāda, Bhikkhu Sujato defines Early Buddhism as the teachings of the EBTs passed down in Pāli, Chinese, Tibetan, and Sanskrit, while defining Theravāda as the school established at the Mahāvihāra in Anuradhapura, Sri Lanka, which later spread across Southeast Asia and regards the Pāli Tipiṭaka as authoritative. ↩

20. Sujato and Brahmali describe the EBTs as a foundational body of literature preserving the teachings of the Buddha and his early disciples; their discussion of the EBTs’ character notes their restrained, realistic style and their distinctiveness from later Buddhist literature. The uploaded Chinese Āgama collection also describes the Āgamas as presenting the Buddha’s teachings on topics such as the four truths, dependent arising, the thirty-seven aids to awakening, monastic life, and liberation. ↩