The Buddha Before Buddhism: Review of Fronsdal

In his new book The Buddha Before Buddhism, Gil Fronsdal undertakes a translation of the Aṭṭhakavagga or Book of Eights, one of the Buddha’s most profound and enigmatic teachings, also widely considered one of his earliest. As is to be expected from his previous translation of the Dhammapada (2005), Fronsdal’s book is lucid and readable, one that should be of great interest to an audience that goes well beyond that of Buddhism. It is the most approachable translation in English of which I am aware.

The text has a couple of key advantages. First of all, it is concise. Fronsdal’s treatment is under 150 pages, including lengthy commentaries. Eminent Pāli scholar KR Norman’s translation that includes only the text itself is just twenty three pages long. (2001: 103-126). The other advantage of the Aṭṭhakavagga is that it appears to present a teaching and practice that is simpler and more down-to-earth than the traditional account found in most canonical texts. To that extent it may reveal in Fronsdal’s words “The Buddha before Buddhism”, before many of its speculative doctrinal and cosmological elaborations became standardized. This is a claim we will be returning to.

Four Themes Plus Four

Fronsdal highlights four main themes he sees in the Aṭṭhakavagga. First, he sees the text as recommending the letting go of all views. Note that this is distinct from a claim of “no-views” (p. 71). It isn’t saying that we should have no views, but rather that we should strive towards an affect of non-clinging to all views, of non-identification with all views, of not becoming emotionally bound up with views as “mine”. Indeed, there is one stanza that explicitly makes the point that the wise don’t speak of “purity” in terms of the having of views nor in terms of the lack of views. “It is by letting go, not grasping, not being dependent …” (839)  In this Fronsdal follows Paul Fuller’s (2005) take on “right view” in early Buddhism.

Second, he sees the text as arguing against sensual craving. Sensual craving is what must be overcome to escape bondage to this world and become free. In several places (e.g., The Discourse on Quarrels and Disputes, Ch. 11) the text gives partial causal explanations of the arising of worldly troubles in terms reminiscent of dependent origination (a matter I discussed in a previous essay). That is to say, we come into contact with things in the world through our senses. Those things we find agreeable or disagreeable. We naturally crave those things we find agreeable and avoid those things we find disagreeable. One thing the Buddha notes as particularly insidious in this regard is sex, which is dealt with in some detail. (Ch. 1, Ch. 7; 835, 926, 940).

Third, Fronsdal sees the text as arguing for a particular understanding of the sage as expert, wise, peaceful, and also as someone who both knows and sees the world. Fronsdal includes a detailed appendix of all occurrences of knowing and seeing in the text, and clearly these are of utmost importance to the attainment of awakening. A sage knows and sees things as they really are, but without clinging to them.

Fourth, he sees the text as recommending a kind of training that one should undergo to become awakened, although this is significantly underdescribed when compared with our standard canonical examples such as the Four Noble Truths, the Eightfold Path, or the gradual training, none of which appear in the Aṭṭhakavagga as such. What is most important in this text is that one becomes a sage through personal effort, through knowing and seeing for oneself, not through the intercession of supernatural beings or the like.

Fronsdal also makes several claims in regard to the text that seems to imply that its dhamma precedes the teachings found in the body of Pāli discourses. I note four additional such claims. Continuing our numbering scheme, his fifth claim then would be that this text doesn’t put forward a teaching or doctrine that is supposed to be superior to other such teachings or doctrines (pp. 1-2, 75). Sixth, the text doesn’t make any explicit reference to the teaching of non-self, nor to famous numbered teachings such as the Noble Truths, Path, and so on (pp. 3, 141). Seventh, the text doesn’t make any claims about extraordinary states of meditative consciousness (p. 17). And eighth, “common Buddhist concerns of rebirth … are primarily discussed in terms of what non-Buddhists believe.” (p. 3).

Evaluating Saññā

There is clearly a great deal to contend with in such a rich text. Much of what Fronsdal says is borne out by the text itself, but some of what he says is questionable and unfortunately depends on how one translates particular words from the Pāli. To begin with, let us take the word “saññā”. It commonly denotes one of the five aggregates that make up the human person, the one ordinarily translated “perception”. (The other four being form, feeling, volition, and consciousness). Fronsdal however translates the word as “concept”. Now, this translation is not entirely without merit: on the Buddhist (and perhaps pre-Buddhist) understanding, saññā is a conceptually mediated process that doesn’t just see, but also categorizes and names things in the world. Hence the word itself has aspects of both “perception” and “conception”. Fronsdal seems aware of the potential problems here since “saññā” is one of the terms he regularly footnotes in the text. His attention to detail, to citing particularly problematic issues of translation with footnotes, is particularly helpful to the reader and should be commended.

We can get an idea of the difference a word can make when we see how say KR Norman translates a couple of stanzas as versus Fronsdal:


(778) Having dispelled longing for both ends, having understood [and renounced] contact, not greedy, not doing that for which he would reproach himself, a wise man does not cling to what is seen and heard.

(779) Having understood [and renounced] perception, a sage should cross over the flood, not clinging to possessions. With barb pulled out, living vigilant[ly], he does not long for this world or the next.


(778) Subduing desire for both sides,
Fully understanding sense experience,
Free of greed,
And doing nothing one would reproach oneself for,
A sage does not attach to what is seen or heard.

(779) Fully understanding concepts and having crossed the flood
Sages don’t cling to possessions.
With arrow removed, living alertly,
They do not long for this world or the next.

Although Fronsdal’s translation is on the whole more readable and elegant, I submit that his choice of that one word is problematic. In Norman what the sage understands is “perception”, while in Fronsdal it is “concepts”. Does this make a difference?  It obscures the progress from one stanza to the next. For it is in “not clinging to what is seen and heard” that one understands perception. In Fronsdal’s translation stanza 778 is about “sense experience” and “what is seen or heard”, which plainly is not concepts, at least in English. (One does not have sense experience of concepts, nor does one see or hear concepts). So on Fronsdal 779 seems to be discussing an entirely different matter from 778. We might then think that 779 is an entirely new claim, which is that by understanding concepts for what they are, sages don’t cling to possessions. But why would understanding concepts help us not cling to possessions? On Norman it is clear that the non-clinging to possessions stems from the non-clinging to what is seen and heard, i.e., perceptions.

One may say that on the Buddha’s picture we do not so much cling to things in the world, or even to our perceptions of things in the world, as to our misconceptions about things in the world. On this understanding, “conception” is the issue rather than “perception”. But I think this move assumes a kind of sophisticated analysis that goes beyond the picture found in the text, which is more focused upon the role our senses and sense-organs play in binding us to the world.

Translation of “saññā” as “concept” also obscures its role in key forms of absorptive meditation, namely the state usually translated “neither perception nor nonperception” (nevasaññānāsañña) that was apparently taught to the Buddha by Uddaka Rāmaputta before his awakening. There are at least two stanzas in the Aṭṭhakavagga in which this kind of meditative absorption might be indicated, one of which (874) Fronsdal notes as “enigmatic” (p. 17; also see p. 92), as indeed it is. It would take too long to discuss fully here, but suffice to say that Ṭhānissaro (2003: 276) believes the stanza refers to the state of neither perception nor nonperception, and Norman (2001: 361) believes it refers to the fourth jhāna.

Another such stanza may be 847.


(847) There are no ties for one who is devoid of perceptions. There are no illusions for one who is released through wisdom. But those who have grasped perception and view wander in the world, causing offence.


(847) Someone freed from concepts has no ties,
Someone freed by wisdom has no delusions.
Those who grasp at concepts and views
Clash as they walk through the world.

This stanza comes at the end of a discussion with Māgaṇḍiya that began with the Buddha rejecting sexual companionship, and then denying that his rejection of sex was due to his clinging to views. Hence the sutta highlights a tight relationship between craving, even sexual craving, and clinging to views. Now, the difference between concepts and views is a matter of complexity: a view is made up of concepts. So on Fronsdal’s translation it seems as though the Buddha is simply saying that we should cease clinging to concepts, and to views which are themselves made up of concepts. Concepts therefore become the whole of the problem.

Norman’s however reveals further nuances: it is not only views that are the problem, but sense perceptions as well. One key problem lies with our sensory apparatus and the way it attaches us to things in the world. Although I would prefer to translate the sentence that there are no ties for one who is “detached from” or “dispassioned by” perceptions rather than to say one must be “devoid of” them, it may indeed be that Norman is on to something. His translation would suggest that the Buddha is recommending a deep state of calm (samādhi) that gets beyond perceptions of form entirely so as to remove any residue of sensual craving. If so, this stanza might amount to an early distinction between samādhi and wisdom (paññā), which with ethics (sīla) became one standard concise formula for the Buddha’s dhamma.

Evaluating the Minor Themes

This brings us to questions about the four other, minor themes that Fronsdal finds in the text. To begin with, does it or does it not include reference to extraordinary states of meditative consciousness? This may seem a rather recondite question. Why should we think it would? If indeed the Buddha learned such states as these from his teachers Āḷāra Kalama and Uddaka Ramāputta before becoming awakened (as Alex Wynne (2007) believes is historically credible), and if the Aṭṭhakavagga is a particularly early text in the tradition, it should be unsurprising to find that the Aṭṭhakavagga reflects meditative techniques learned from some of his most recent meditation teachers. As we have seen, there are a couple of stanzas that may indeed refer to such extraordinary states of consciousness. If nothing else, the “enigmatic” character of certain stanzas suggests they refer to states that are unusual. One can find a few other potential mentions of such states in the final chapter (962, 972, 975), however that chapter may postdate the rest of the text. Indeed it is not entirely clear what the original state of this text might have been. It is possible that it was called the Aṭṭhakavagga (“Book of the Eights“) because originally it only consisted of those four chapters (2-5) that had eight stanzas. If so fully three quarters of the chapters postdate its original core (p. 144).

As for non-self, which Fronsdal says is not found explicitly within the text, there are many passages that, depending on how one translates them, deal with this theme. (E.g., 790, 800, 805-11, 858, 913-14, 919, 922, 950). Fronsdal glosses many of these as expressing “not-mine” rather than “non-self” (e.g., p. 62), but it is not clear whether “non-self” in early Buddhism is really that different from “not-mine”. The standard early Buddhist pericope of “not I, not mine, not myself” includes “not mine” of course. “Not I” is a refusal to identify oneself with the objects of experience, and “not myself” is a refusal to identify a view of self with such objects. The former as such may not be reflected in the Aṭṭhakavagga, but the latter is, insofar as one is to cease clinging to (that is, identifying with) all views. So what we end up with is an understanding of the Aṭṭhakavagga as expressing most of what we find in early (Nikāya) Buddhism regarding the self. It may not have been stated explicitly, that is in the sense of a fully developed doctrine, but nevertheless it can be found.

As for the issue of rebirth, to be fair Fronsdal does not claim it is entirely absent in the Aṭṭhakavagga. It can be found in a few stanzas (e.g., 779, 865), as well as understood in certain doctrinal terms cited in the stanzas. For example the term “āsava” or “defilement” in 913 stems from Jainism (Norman 1997, 34) where it would have had connotations of rebirth.

The broader question about rebirth and other so-called ‘metaphysical’ (or perhaps better, ‘speculative’) issues in the Aṭṭhakavagga is answered by Norman (2003) who points out that the Pārāyanavagga (the text that comes just after the Aṭṭhakavagga in the Sutta Nipāta, 976-1149) has every indication of being just as old, and yet is replete with discussion of issues such as the escape from birth, old age, and death. (E.g., 1043-48, 1082 which also includes the term “āsava”, 1094, 1123).

There remains Fronsdal’s theme that the Buddha in the Aṭṭhakavagga doesn’t put forward a doctrine that he claims superior to other doctrines. To be fair, one does find some stanzas that seem to head that route. E.g., the Buddha says:

(837) Considering the doctrines [people] cling to,
It does not occur to me to say
‘I proclaim this.’
Seeing — but not grasping — these views,
I knew and saw inner peace.

But even so, one wonders what to make of such a claim, when the text is replete with apparent “proclamations”, from the problems of sensual craving, to the descriptions of the sage, to claims about the nature of the training. (All of which by the way are aspects of the Noble Truths). Indeed we find another stanza that reads,

(946) Not deviating from truth (sacca)
Sages, brahmins stand on dry ground.
Giving up everything,
They are said to be at peace.

One may perhaps wonder how one can both give up everything and yet not deviate from truth. Perhaps this is the secret of right view: it is precisely by not deviating from truth that one gives up everything. But of course “truth” implies a truth-claim, which in this context seems to imply a doctrine of such claims. Another stanza:

(783) Of noble character — say the skillful [ones] —
Is a monastic at peace and fully stilled
Who does not speak of one’s own virtue,
Or have any arrogance anywhere in the world.

Here we have the odd juxtaposition of a monastic described as “noble” by “the skillful” (Norman: “experts”), who nevertheless does not speak of his own virtue or presume to arrogance. This is odd because it is a text supposedly spoken by the Buddha himself. He is, he must be, “speaking of his own virtue” by claiming nobility. He claims to be a sage throughout the text, whether directly or by implication.

Either we are to take this as a contradiction, a kind of hypocrisy — not a charitable reading, or we must consider that something more nuanced is being said. As with the case of views and doctrines, the way through this thicket is to realize that it is the mind-state of the speaker that makes all the difference. One of ignoble character speaks of his own virtues with a mind-state of arrogance, that is with an ego-driven clinging to his own good qualities. One of noble character may indeed mention his own virtues but does so without arrogance. He simply states facts. We must take “does not speak of one’s own virtue” with a grain of salt in this stanza: he is in fact speaking of his own virtue, but doing so without an accompanying state of arrogance.

It is the same issue with extolling a doctrine as with extolling one’s virtues. The Buddha may well claim to do neither, but this is transparently untenable; it is an exercise in special pleading. Of course the Buddha is extolling a doctrine, it is all over the text. Of course he is extolling his virtues. We may say that he is doing so without arrogance, without clinging to them as “mine”, without negative emotion, without any intent to debate or dispute, but much more than that we cannot say. For the Buddha is making any number of substantive and even controversial claims in the text. One only has to look at the sutta to Māgaṇḍiya (835-836), where the Buddha rejects the hand of a beautiful woman because she is “full of urine and excrement”, to see this. And plainly the Buddha’s claims cohere into a rough doctrine or teaching.

“Before Buddhism”?

Norman’s (2003) paper throws cold water on a lot of claims made about the Aṭṭhakavagga’s apparent differences from other canonical texts of early Buddhism. It is true that one does not find typical doctrinal formulae or lists in it (cf. Fronsdal pp. 3, 141), but neither does one find anything that disagrees with doctrine. All one is left with when trying to formulate a notion of the dhamma “before Buddhism” is an argument from silence. Of course, insofar as one simply wants to locate a text without very much outside of a this-life path, the precise doctrine of the Aṭṭhakavagga hardly matters. But insofar as the text is supposed to be a window to something historically prior to the Buddhism of the (rest of the) Nikāyas, it may.

As Norman makes clear, although the Aṭṭhakavagga is likely early, it is hardly the only such early text in the Nikāyas. Other texts such as the Khaggavisāṇa Sutta (Snp. 35-75), the Pārāyanavagga, the Ariyapariyesanā Sutta (MN 26), the Dhammacakkappavattana Sutta (SN 56.11), and the Alagaddūpama Sutta (MN 22), among others, contain material likewise considered early. Insofar as we are looking to determine what the Buddha’s teachings looked like “before Buddhism” it seems odd to confine our search to one early text rather than looking at all the early ones we seem to have.

At any rate, in a lengthy afterward Fronsdal goes into a number of arguments that cast some doubt on the historical claims some have made of the text; clearly the issue as to its relative antiquity, and precisely what we should take that to mean in the history of Buddhist ideas, is far from settled. Hence I think we should be cautious about making too much of the Aṭṭhakavagga. If indeed we find it reveals the dhamma in a somewhat sketchier and more preliminary state than we find elsewhere, that is perhaps enough. It is also enough to note that the teaching found there is direct and oriented towards the present life, even if this does occur before a background that likely was more cosmologically speculative than might appear from the text.

Arguably — and it is hard to make any claims about this text that are immune from argument — the Aṭṭhakavagga was composed and preserved as a compact, relatively introductory teaching. Short, easily memorized, relatively jargon-free, and extolling the virtues of non-clinging to views, it might have been useful when teaching early arrivals with different philosophical or religious backgrounds. It might have been intended as a kind of “dhamma 101”, not merely to instruct in the preliminaries to (what would become) settled doctrine, but to induce youthful and energetic monastics to cease squabbling, and not incidentally to cease looking for sex. That is, perhaps we ought to look at the text as much for what it was intended to do as for what it was supposed to show about the views of its author.


Gil Fronsdal is one of the most lucid and elegant translators of early Buddhist texts. His Dhammapada is perhaps the best version in English, and his new version of the Aṭṭhakavagga, fruit of a very long acquaintance with the text, is similarly elegant. It is by far the most readable of any such translation of which I am aware, the kind of work that is both small enough to keep in a coat pocket and profound enough to reward continual re-reading.

Since the subject of the text is almost entirely confined to issues of lived reality it is also a text that can be read and digested even by those with no interest in Buddhist cosmology. With a few caveats such as its strong argument for celibacy, the text could almost be produced in a non-Buddhist, lay context today.

All that said, Fronsdal makes a handful of questionable moves in the translation that (at least in my estimation) make the text somewhat less lucid than it otherwise might have been. They also suggest concepts and arguments from much later in the Buddhist tradition. Fronsdal is far from the first to have made such suggestions; Luis Gómez (1976) claimed to find “mystical” elements of “proto-Mādhyamika” in the text, a claim which I have disputed (2015). While Fronsdal does not go so far as to attribute the position of “no-views” to the Aṭṭhakavagga, I think some of the mystical, “no-view” flavor remains in his translation, perhaps stemming from his having been introduced to it in the context of the much later Buddhist school of Zen. That said, the care which Fronsdal takes to highlight problematic Pāli terms in footnotes is of great help to the careful reader. All too often such decisions are left obscure.

There is at last the question to what extent Fronsdal’s translation, indeed the Aṭṭhakavagga itself, really approaches “the Buddha before Buddhism”. To an extent perhaps it does, but if so, only if seen in the context of other texts from the canon that appear of similar antiquity, insofar as we can gauge such things. I do not believe the dhamma found in the Aṭṭhakavagga deviates to any important degree from that of the other canonical texts, although it appears in a sketchier, perhaps more preliminary form. It is the manner of teaching that makes this text so compelling: direct, human, down-to-earth. But this is the hallmark of all the Buddha’s most profound teachings throughout his long career.


Gil Fronsdal (2005). The Dhammapada. (Boston: Shambhala).

Gil Fronsdal (2016). The Buddha Before Buddhism. (Boulder: Shambhala).

Paul Fuller (2005). The Notion of Diṭṭhi in Theravāda Buddhism. (London: Routledge).

Luis Gómez (1976). “Proto-Mādhyamika in the Pāli Canon.” In Philosophy East and West 26:2.

KR Norman (1997). A Philologial Approach to Buddhism. (London: University of London School of Oriental and African Studies).

KR Norman (2001). The Group of Discourses (Sutta-Nipāta) 2nd Ed. (Oxford: Pali Text Society).

KR Norman (2003). “The Aṭṭhakavagga and Early Buddhism.” In Jainism and Early Buddhism: Essays in Honour of Padmanabh S. Jaini, Part 2, Ed. Olle Qvarnström (Fremont: Asian Humanities Press).

Doug Smith (2015). “Was the Buddha an Anti-Realist?” In the Journal of the Oxford Centre for Buddhist Studies, 11.

Alex Wynne (2007). The Origin of Buddhist Meditation. (London: Routledge).