SuttaCentral: designing the Dhamma

When the time came to look again at SuttaCentral (this was late last year), one of my interests was to bring a refined and careful aesthetic sensibility to the Dhamma. In the past couple of years I’ve developed an interest in typography, due to publishing my own books. I knew that it is now possible to create quality typography on the web: and if anything should be typeset beautifully, it’s the suttas. As I’ve written previously, there is a long tradition of design and typography in writing the suttas, which springs from the very earliest Indian manuscripts. How, then, do we present the Suttas in the best possible way?

There are many issues to consider here, and I’ll try to cover a few of them.

Websites are dead

It sounds like a paradox, but it’s not really. No-one uses “websites” any more. They google, or they tweet or they Facebook. Maybe they read blogs, or check out some articles in a newspaper, or watch videos on YouTube. In none of these cases is someone actually using a “website”, in the basic sense of a structured group of HTML pages. The website is irrelevant. What matters is the content, what is on your screen at this moment, not what the theoretical “website” consists of.

This is the downfall of almost all sites that present the suttas. They are about the site; and often enough, they commit the cardinal, deadliest of deadly sins: using the website as a glorified ad for the organization. As soon as you think like that, you’ve failed. The content is no longer there for its own sake, but as a trick to drive the reader to something else: a product, or raising funds, or just plain old PR.

How do you escape from this? My favorite un-website is gov.uk. Seriously, it’s awesome. It won a major design award last year, and there’s nothing to it. There’s no smiling faces of happy people, no sloganeering, no clever flash, or any of the other meaningless crap that bogs down most websites. It is based on the basic premise that no-one wants to be in a government office, so let them do their business and get out as quickly as possible. It’s the same principle as Google: not, “How do we keep people?”, but “How do we get people where they want to be?”.

I wanted to bring this attitude to the suttas. No-one wants to trawl around a site trying to find a sutta hidden away in a list of links for other, lesser, things. No-one wants to download huge images or be told about the great work our organization is doing or see an ad for a meditation technique, and so on. Our task should be to bring the suttas to the surface in the fastest, best possible way, so that readers can get where they are going, and not have to deal with a website.

So we developed a navigation structure that is extremely shallow: 2 or 3 clicks and you’re at any sutta. The menu is available from any page, and the search is instant (and is getting more awesome every week: just wait until our new search is launched). It is the fastest and most powerful way of navigating the suttas that we can think of. If you’ve got a better way, let us know!

Buddhist Typography

I approached this from two angles. First, what are the fundamental Buddhist design principles? And second, how do we express them using the best practices of web typography?

Consider Buddhist design, the classic shapes of Buddhist art: the Buddha image, the mandala, the stupa. They all share an aesthetic sensibility. They are strongly and simply symmetrical. They are oriented towards the center and the top. This is no accident: the centering is an expression of psychological and spiritual oneness, and the upwards orientation is an expression of the aspirational nature of the path. They use the basic shapes: circle, square, triangle, to create a sense of order amid the complex and chaotic details of samsara.

I also looked at traditional manuscript design. Buddhist manuscripts are always symmetrical, using the universal principles of good typography: evenness, spacing, clarity. Due to the nature of the materials, they are usually horizontally oriented, unlike the vertical structure of a webpage or a book. But many of the design elements are close to those of Western typography. As just one example, the number of characters per line varies from manuscript to manuscript, but on average it approaches the Western ideal of around 80 characters/line.

What is equally important, especially from a Buddhist point of view, is not just what is on the manuscript, but what is not on it. With the exception of highly ceremonial designs, most Buddhist manuscripts are plain and workmanlike. The important thing is the words, and there’s pretty much nothing else on the page. This is contrary to most web presentations of the texts, which love to clutter the page with multiple reference numbers that interrupt the flow of reading.

These ideas are all reflected in the design for Suttacentral. The page is clear and uncluttered. All the reference material and other stuff is hidden away: you can call it out when you want it. But our default presentation is for people who want to read the text. The design is centered and symmetrical. The headings appear in small-caps in the center of the text, drawing the eye inwards and upwards. The small-caps also evoke a sense of seriousness and formality. Notice, also, that these textual elements are grayed out: this is because they are not part of the text itself, but have been added by redactors. So we create a visual reminder of the editorial layers of the text.

Western Typography

In the basic layout I was guided by the application of Bringhurst’s classic principles to web design. The page is a grid design, with a regular vertical movement and constant proportions. The basic set of proportions is selected from the Dhammacakkappavattanasutta. Notice how the Buddha structures his teachings in terms of interlocking numerical sets: 2 extremes, 3 turns, 4 noble truths, 8-fold path, 12 aspects. This is no accident: it provides a sense of order and interrelationship, a sense of the wholeness of the Dhamma. This same sense of proportion is what guides visual artists in creating harmony and beauty.

The most basic element in a page of text is the line height. This is the first and major impression that a page presents: a series of lines. So I took this as the basic unit in the text grid. Since the most basic number in the Dhammacakka is 2, the line height is assigned to two. And as this is a vertical height, the horizontal movement should express the second most basic number, three. So all the elements on the page are organized according to this 2:3 vertical:horizontal ratio, which is then expanded according to the proportions of the Dhammacakka: 4, 8, 12. Of course, a web page is dynamic, and is interpreted by browsers, so we can’t get the precision that we can on a printed page; but anyway, that’s where the basic proportions come from.

You don’t see these things. They are invisible guidelines. But so many people have commented, “Wow, SuttaCentral looks so great!”, it shows how strongly we respond to proportion and order.

One of the ideals of typography is that it recedes. When you read a text, the important thing is the words you are looking at. Anything that distracts the eye is an impediment to readings. And that means pretty much anything at all. A study of reading, for example, showed that use of bold text actually slows down readers: the eye is distracted. This is why the traditional ideal is to aim for a uniform “greyness” of text. Any excessive gaps or unevenness is avoided. The sense of beauty that emerges from a well-designed page is no accident; it is the cumulative effect of countless tiny decisions in the use of space, proportion, and color.

Typography is not just the overall proportions, but the meticulous attention to tiny details. We have managed to implement most of the elements of quality typography, some of which are commonly ignored in even expensive print editions: hanging quotes, proper (and letterspaced) small-caps, correct use of punctuation (including dashes and quotes), verses centered on the longest line, and so on.

One element we have not achieved is proper justification. Our text really should be justified, as this reinforces the symmetry and balance, and for this reason Buddhist manuscripts always use justified text. It is a complex topic, but the essence is that well-justified text is currently impossible to achieve on the web. I have been looking into possible solutions, but so far nothing that is workable.

Fonts

It is now possible to display pretty much any language well on the web, due to the widespread adoption of Unicode, together with the acceptance of @font-face for delivering fonts. The problem is no longer just getting the glyphs you need, but getting a font that actually works well.

There are plenty of free fonts available, but few of them have the glyphs we need. Moreover, rendering fonts on the screen is still a complex and dicey matter. I’m lucky, I use Ubuntu on a good screen, so pretty much any font looks okay. But on low-res screens, and a variety of inferior operating systems, it’s almost impossible to find a font that actually works well across all devices. In addition, I wanted the flexibility of the full range of typographic possibilities, including true small-caps, which are very rare in a free font. To add to the complexity, as the design evolved it became clear that we needed both a serif and sans-serif, in order to clearly structure the information on the page. These both need to both have proper typographical features, be available in small-caps, render well on all screens, and include the various exotic language glyphs.

Ultimately, I simply couldn’t find a free font that met all these criteria. The closest is my beloved Gentium; but this is designed for compact print, and is cramped on a screen. Eventually I came across the wonderful Skolar, which you are reading this blog in right now. The basic version is available online on WordPress, but the full font, which includes masses of specialized and advanced features, is only available commercially. Despite my love of open source, in this case this was the best option.

Skolar is an astonishingly good font, which has been widely adopted in the design community, but which has so far largely passed by its intended market, scholarly publications. It is designed from the ground up to be used with a wide variety of international glyphs, rather than most fonts, where the diacritical marks are bolted on after the fact. It has the classic attributes of a great screen font, with strong lines, low contrast, and moderately large x-height. In addition, it benefits from detailed hand-hinting by the creator, which is the single most important factor in ensuring good appearance on inferior screens.

I have been able to work directly with the font designer, David Březina, to adapt and extend Skolar for SuttaCentral. The initial version includes support for the Pali and Sanskrit diacriticals. David kindly created the Dhamma-wheel you see in the SuttaCentral masthead. This is the Unicode character for the Dhamma wheel, not an image. I chose this as a homage to the Unicode project in enabling communication and learning across all peoples. The next version of Skolar, which will be live on our site soon, has support for Vietnamese characters, which were developed by David specially for SuttaCentral. I’m delighted that we can make this little contribution towards quality typography for everyone.

The companion sans is Source Sans Pro. This was created by Adobe and released as their first open-source font. Originally designed as a UI font, it has outstanding readability at small sizes, which makes it ideal for its use on SuttaCentral. It covers all the glyphs we need and comes in a wide range of weights and styles. In addition, it harmonizes well with Skolar. (I peeked at the innards of a pdf that David Březina sent me, and he uses it as companion for Skolar, too!). I’ve adapted it slightly for use on SC.

Conclusion

I could go on for quite some time, but will have to call it quits for now. Hopefully I have conveyed some sense of the care and consideration that has gone into crafting SuttaCentral. As always, design is in service of a higher purpose: to convey the Dhamma.

The Suttas speak of how the Buddha would speak in a voice that is clear, pleasant to listen to, articulate, and distinct. These principles are fundamental to communication. The design of a Sutta should be attended to with the same care that the Buddha used when speaking.

Unfortunately, most sutta websites so far seem to have used the design principle of throwing a bunch of text at the screen and, I don’t know, praying that it will work out okay. This is neither good scholarship nor good communication.

The presentation of a text conveys, unconsciously but unmistakably, what the designer thinks of the text. If a speaker mumbles, grunts, and slurs, it is a sign that they don’t care about what they are saying, or about the person listening. And if a web designer doesn’t bother to pay attention to the manner in which a text is presented, they throw out millenia of experience with the written word, which universally attests that the manner of presentation deeply influences how the matter is received.

The new SuttaCentral is live

I have some wonderful news today: our new SuttaCentral is finally live!

You can see it here at: http://suttacentral.net/

We have been working very hard on this for some time now, and are very happy with the results so far. I will be writing on various aspects of this project in the coming weeks, so for now let me just introduce a few of our main features.

  • A complete new design from top to bottom. The whole site, in fact has been rewritten from the ground up.
  • Much faster more powerful navigation.
  • Instant search!
  • Many added new references and improved structure
  • Many texts are now hosted on site, including the main Suttas of the Pali, the Vietnamese translations, and some English and Korean translations.
  • Brand new translations from the Saṁykuta Āgama (both the main and shorter version), and the Upāyika. Together with the older translations of the Ekottara and Dīrgha Āgamas, SuttaCentral now has the most comprehensive collection of Āgama translations available anywhere.
  • We also have newly available online versions of Bhikkhu Bodhi’s classic translations from the Dīgha Nikāya
  • The Pali and Chinese texts have a very cool instant word lookup function. (We only have one Chinese text online for now, more coming soon.)
  • A very powerful, intuitive pretty URL structure: you can go anywhere on the site directly from the address bar if you know the abbreviated uid for the text.
  • A Feedback forum: tell us what we can do to make it better! (BTW, while feedback anywhere is good, generally it will be better to leave specific suggestions for the site on that forum rather than here.

Regular readers of this blog will recall that some time ago I complained about how there was no ready good source of Early Buddhist Texts online. Today, we begin to change that.

There are many areas where we want to expand and improve the site, but I hope that it is already something of help for anyone who wants to learn what the Buddha taught.

I have been absent from this blog for some time, as I have wanted to commit myself fully to getting SuttaCentral ready. Now that it is up, I hope to spend some more time back here.

Get the stuff out of the way, let the Dhamma shine

The New SuttaCentral

In recent weeks I’ve been working hard on the revamped SuttaCentral. Not quite ready for prime time yet, but it won’t be long. SuttaCentral has for several years been the only place on the web you can get reliable and comprehensive data on Sutta correspondences: how to find out whether there are any other versions of a Sutta, and where and what they are. It’s rapidly becoming much more than that, and in future posts I’ll talk about some of the great innovations we’ve put together.

One of the biggest changes is the inclusion of texts on our site itself. In the past we merely linked to texts hosted elsewhere. That means we have full control over the presentation and markup, and can integrate the texts in ways never before possible.

The Upāyikā

Suddenly, all sorts of new things are coming together. One exciting development is the appearance of brand new translations by Sāmaṇerī Dhammadinnā (ex Giuliana Martini) of the important Tibetan work, the Abhidharmakośopāyikānāmaṭīkā, which she refers to more succinctly as the Upāyikā. This is one of the few Tibetan texts that contains substantial translations of Suttas from the early Āgamas. Previously the only substantial work on this text has been in Japanese. Now Sāmaṇerī Dhammadinnā has undertaken a series of translations with detailed notes and commentary, and has graciously allowed SuttaCentral to host the text. In keeping with our focus on making accessible the actual texts, we will just present the translated Āgama Sutta portions, and will link to the full essays.

The Upāyikā contains the full text of the many (hundreds?) of passages referred to in Vasubandhu’s famous Abhidharmakośa. It has long been known among scholars of early Buddhism, but we have been ignorant of the detailed contents. Now, thanks to the patient, detailed, careful work of Sāmaṇerī Dhammadinnā, we can all read these precious teachings.

Cut by numbers

But when I was reading her essay, I have a little hiccup moment. Nothing to do with the translation or the essay: I stumbled over the reference system for the texts.

Each translation is given a number, such as Up 9001. When I saw this, I did a double take. What, there’s over 9000 quotations in this book? I knew it was extensive, but that’s a bit much! I was confused, and didn’t know how to interpret the numbers. I looked through the essay and notes, but couldn’t find any explanation. So I emailed Sāmaṇerī Dhammadinnā, and got the answer. It was explained deep within a nearly page-long footnote:

“… the quotation number as established in Honjō 1984 and successive supplementations in his publications (for example, “Up 9001″, which stands for quotation number 1 in chapter IX of the Abhidharmakośabhāṣya and Upāyikā)…”

Problem solved, no big deal. But somehow this little detail sparked a connection with many other details that I have encountered in my recent work on SuttaCentral. It made it clear for me all the countless ways we, as Buddhist scholars, obscure the work we should be illuminating.

At the time of reading the translations, I was inspired, excited. This is an amazing thing! That such teachings could have survived over thousands of years; transmitted in India in countless generations, then brought to remote, forbidding Tibet and rendered in the local tongue by Śamathadeva, an expert scholar, to be preserved, copied and passed down to us. And now, translated into English so we can all see, once more, the heart of the Dhamma beating in the Suttas.

And then a little reference number changed my emotions from excitement to confusion, from inspiration to doubt, without me even noticing. Suddenly, instead of focussing on the teaching, I was focussing on understanding the reference number. Sure, only a little thing. I got over it. The point is, I shouldn’t have been there in the first place.

What went wrong? Basically, the reference system ignores conventions. Chapter 9, section 1 should be referenced as “9.1”, not “9001”. This referencing method is simple, informative (because it clearly differentiates between chapter and section), and is used in millions of documents. It is so basic that it’s the default method for numbering “chapter/section” in every software for structured documents, like word processors or LaTeX.

Unconventions

So what’s the difference? Who cares? Once you know the method, you can find what you want, so big deal.

The big deal is that we shouldn’t have to relearn the method. There is already a method, time-honoured and worn smooth by custom and usage. It works fine. And when we use it, it vanishes. If I had seen “Up 9.1” I would not have even noticed it. It is there when I need it, and simply disappears when I don’t.

Think about the road signs, the warning notices, the nutrition information on food. All this structured data is presented in consistent, clear formats so that you can immediately recognise it and use it when you need it without having to think about it. If you have to think, “Is that a speed limit or a no-parking sign?”, it’s failed already.

So why use this unconventional method? Because it was established by the pioneer scholar in this text, Honjō Yoshifumi. This is standard in scholarly circles: keep consistency with previous work. Which is of course a good thing on the whole, but it does result in the perpetuation of outdated data (like the absurd practice of referring to Pali texts by reference to the volume and page of the century-old, often poorly edited editions of the Pali Text Society). Why did Honjō use this method? I have no idea. Maybe they do things differently in Japan.

People matter

Do we really need to keep this consistency? It will only ever be noticed by that subset of sentient beings who are a) human, b) scholars of early Buddhism, and c) fluent in Japanese and English. Not, I suspect, a double-digit audience. On the other hand, there are many people who would be interested to read these in Sāmaṇerī Dhammadinnā’s excellent, lucid translation. How many? Thousands, at least, probably much more. So why not think about those thousands instead of the tiny handful of professionals?

This is why the Buddha said he did not transgress the conventions of the world: by using conventions properly, they fade into the background so we can focus on what matters.

I know, I know! Why get worked up about something so small? Because, as I suggested earlier, this is just one of a million similar details. The more I work on this material, the more I realise just how abrasive the Buddhist scholarly method is. It’s death by a thousand papercuts. Arcane terminology; massive footnotes; obscure references; dead languages; paywalled journals; pedantic explanations of the obvious and unstated assumption of the unobvious; and the pervasive sense of a dialogue between scholars, not a dialogue with people actually interested in Dhamma.

All this, and much more, is just grit in the gears. Every Pali term is a cypher creating confusion in every single person who reads it (apart from Pali scholars). Every reference number is a meaningless string of symbols unless you are familiar with the field. As long as you place this stuff in the foreground, you are having a conversation between experts. Which, of course, needs to happen. It’s just that it shouldn’t be the only conversation. An outsider who listens to this conversation feels confused and excluded, and, unless they are unusually persistent, they will just give up.

Smooth like butter

We should get the stuff out of the way. Someone who wants to read the Buddha’s words should not have to learn referencing systems, Pali, and abbreviations. They should just get to what they want seamlessly and intuitively.

I was invited for a meal with a very kind and warm Vietnamese family the other day. The kids were playing with iPads. One was very young, I don’t know, maybe 4. I watched how they interacted with the screen, with each other. How intuitive it is! They just pick it up, doing complex tasks without even trying. Wouldn’t it be amazing if we could bring that kind of experience to something profoundly meaningful, like, say, the words of the greatest spiritual teacher ever?

These things don’t happen by accident. They come because the engineers at Google, Apple, Facebook, the gaming companies, or wherever, do their work, and then get out of the way. They know it’s not about the code, it’s about the experience. So they sweat every tiny detail until the code disappears.

People use these things because they have a positive emotional response: they like it. And when you get that kind of engagement, you don’t have to make someone do it. If anything, you have to make them stop.

We should give at least as much time and care and attention to presenting the Buddha’s words as a coder gives to a game. A reader should just read. That’s hard enough, and it’s simply unreasonable to expect more.

Who reads?

A survey of Americans showed that 19% read the Bible daily. Daily. And this was seen by Christians as a worrisome sign of declining religiosity. They asked, “Why don’t all of us read the Bible every day?”

How many Buddhists read the Suttas every day? 1%? Frankly, I doubt if 1% of Buddhists have read any Suttas ever. Why is that? Because the Bible is inherently better, more compelling as scripture? I don’t think so. I think it’s because we haven’t tried. Because we get lazy and complacent and used to speaking to our little circle. Because we don’t have the vision to imagine things could be different.

There’s no great secret here, nothing mysterious. Just paying the same kind of careful attention to every detail in presenting the texts as the scholars have in preparing them.

A little faith

Dhamma is inherently compelling. We just need to get out of the way. The Buddha didn’t say, “MN39.4, note 17”. He said, “I teach for one who feels!” As his students, we should be letting people hear his words, not ours. SuttaCentral, even the new edition, is very far from realising this ideal. But one step at a time.

So, let’s get to work. You tell me: what can we do to make the experience of reading the Suttas better? What’s your papercut? What’s that moment when you tried, but stumbled over something you didn’t immediately understand, and then you gave up?

Looking back over this essay, I can see how I am not yet practising what I preach. My sentences are full of Pali or Sanskrit names and terms, and assumed ideas. I’m leaving them there as evidence for the prosecution. I’m guilty as charged. But, your honour, I really want to change! I promise I’ll be better next time. But I’m a hopeless addict. I just can’t let go of my Pali, not by myself anyway. I need your support.

The exalted extract of the Vinaya

As part of the Authenticity Project, I’ve been going back over the Asokan edicts. Here at the Buddhist Library in Sydney, they have the original publication of the edicts, with pictures, text, translations, and everything: a wonderful old resource, much better than anything available on the web.

I’ve looked into the question of the texts recommended by Ashoka in the Bhabra edict. As is well known, Ashoka recommends a list of texts for the Sangha and lay followers to study. The texts are all obviously part of the early Buddhist canon, yet it is not possible to identify them all easily. This is because the Suttas have never had universal, unique names.

Several attempts to identify them have been made. Probably the best known in Buddhist circles is that of Ven Thanissaro, who did a nice discussion and translation some years ago. As I reviewed them, however, I have become convinced that he is mistaken in his identification of the first of the texts, and that earlier scholars had it right.

The text is the “Vinayasamukase”, which could be rendered the “Extract of Vinaya” or the “Praise of Vinaya”; the root meaning is something like “drawing up”. Thanissaro says that the term vinaya-samukkamsa occurs only once elsewhere in the canon, in the obscure Parivara. With no explanation in the text itself, he relies on the commentary, which identifies these with the “four great standards”, which were laid down originally as a guide to what may be considered as allowable foods in the afternoon. The commentary makes this identification, no doubt, because the following sets of dhamma deal with allowable types of food. But this is a very weak link, as the text in question, the Ekuttarikanayo, is just a list of often unrelated numerical sets; and the term sāmukkaṃsā has nothing to do with the four great standards, a problem that the commentary and sub-commentary address with a convoluted explantion.

Moreover, the text does not even mention the vinaya-sāmukkaṃsā, merely the cattāro sāmukkaṃsā. At least, that is what the VRI text has, perhaps Thanissaro had a different reading. Regardless, one should not rely on such a doubtful text.

It is curious that the term is not further explained, unlike almost all the other terms in this section. I suspect this is because the author expected the reader to be familiar with it and thought no explanation was necessary.

This brings us back to the frequent use of sāmukkaṃsā in a quite different sense: the sāmukkaṃsikā teaching of the Buddhas: suffering, origin, cessation, path (yā buddhānaṃ sāmukkaṃsikā dhammadesanā, taṃ pakāsesi dukkhaṃ samudayaṃ nirodhaṃ maggaṃ). This is part of a standard passage, where the Buddha gradually leads a person on to higher and higher teachings and reveals the Four Noble Truths when they are ready. In this context, the meaning of sāmukkaṃsā as either “extract” or “exalted” fits well.

Several early scholars (e.g. H. G. A. van Zeyst in Encyclopaedia of Buddhism, Vol. II, Fascicle 2, S. 178 – 187) identified the vinayasamukase with this passage, and further, with the Buddha’s first sermon, the Dhammacakkappavattana Sutta, of which the passage may be regarded as a “summary”. This identification is made in the commentary to the Nettipakarana: Idaṃ dhammacakkanti yāyaṃ bhagavato catusaccavasena sāmukkaṃsikā dhammadesanā, idaṃ dhammacakkaṃ.

This is a far more plausible suggestion than Thanissaro’s idea that the very first of Ashoka’s recommendations for the Sangha is how to work out what they can eat in the afternoon!

The question arises, then, as to why this is called the “Vinaya”-extract. Well, the text in question occurs several times in the Vinaya in the period just after the Buddha’s Awakening, which is a very famous passage. More generally, though, vinaya only gradually came to have its specialized meaning of “monastic discipline”, and in early texts means “education”, “training”, and often just means the same as “dhamma”.

The “four noble truths” fit in well with the “four sāmukkaṃsā” in the Parivara, and there is little wonder that such a well-known term should have needed no explanation.

All in all, then, the early scholars were right: Ashoka recommended that the Sangha and laity “frequently listen to and reflect on” the four noble truths as taught by the Buddha in the Dhammacakkappavatana Sutta.

On authenticity

In the past few weeks, I’ve started a project with Bhante Brahmali, which we call “The Authenticity Project”. We have heard skeptical voices that doubt the authenticity of the early Buddhist texts, while among traditional Buddhists the question is rarely even raised. Yet we have not found any source that collects and analyzes the many and varied reasons for regarding them as authentic. So we decided to do it ourselves. The project is developing, and will possibly end up on Wikipedia, and perhaps as a journal article in some form. I’ll share it with you when it is in better shape; at the moment it’s very rough.

The problem is exemplified by the Wikipedia page on the Pali canon. I noticed that the scholars who affirmed the authenticity of the texts were all experts in the field, while the ones who doubted were scholars of later Indian Buddhism and Tibetan tantra. Yet if you are not familiar with the field, it just seems as if scholars do not agree. So I changed the page to acknowledge the backgrounds of the relevant scholars.

I am interested to hear your ideas on this topic. Clearly authenticity matters, as people in all different traditions and religions get very excited by it. But it is not so obvious why this is so: for many people, if it works, it’s good enough. The Buddha in the Sandaka Sutta even warned against over reliance on the authenticity of the texts, saying that, since the teachings may be ‘well heard or badly heard’, one’s spiritual life should not depend on this.

It’s also interesting to hear what different people regard as persuasive. When speaking to various people, almost always they will come up with some different perspective on why the texts should be seen as authentic, or not. We’re interested to gather as many such perspectives as possible, and present them with appropriate analysis and documentation. So, what do you think?

Concerning baskets

The hero Kartaga grappled with the Swan-woman. Long they wrestled. Moons waxed and waned and still they wrestled; years came and went, and still the struggle went on. But the piebald horse and the black horse knew that the Swan-woman’s soul was not in her. Under the black earth flow nine seas; where the seas meet and form one, the sea comes to the surface of the earth. At the mouth of the nine seas rises a rock of copper; it rises to the surface of the ground, it rises up between heaven and earth, this rock of copper. At the foot of the copper rock is a black chest, in the black chest is a golden casket, and in the golden casket is the soul of the Swan-woman. Seven little birds are the soul of the Swan-woman; if the birds are killed the Swan-woman will die straightway. So the horses ran to the foot of the copper rock, opened the black chest, and brought back the golden casket. Then the piebald horse turned himself into a bald-headed man, opened the golden casket, and cut off the heads of the seven birds. So the Swan-woman died. (Frazer, The Golden Bough, ch. 66.)

The basket question

There is a question that has been keeping many of you awake at night, tossing & turning, endlessly ruminating over the meaning of the paradox at the heart of the matter. In the interests of a good night’s sleep, I hereby propose to solve the paradox and ease your torment. The question that has been driving you mad is: what does pitaka really mean?

Fear not! A solution is at hand, a solution that may surprise, even delight you.

Pitaka means “basket”, of course; but that’s not the problem. The problem is that the early Buddhist texts, known as the “Three Baskets” were oral scriptures. It would make perfect sense to think of a “basket” of manuscripts; but what are we to make of a “basket” of memorized texts?

The word pitaka is not used of the Buddhist texts in the canon itself. They refer to themselves, rather, as angas, or “limbs”. Pitaka is used, however, to refer to a traditional text, obviously referring to the Vedas, where the Buddha, most famously in the Kalama Sutta, says we should not accept something merely because it is “included in the basket” pitakasampada. There is inherent in the use, therefore, the notion that something in the pitaka is, or may be, authoritative. It is not just ordinary words or sayings, but a collection that has been specifically preserved and passed down as an authority. It is something valuable.

And that is, I think, coming closer to the point of the metaphor: a basket is something for preserving something valuable. A basket defines and separates, it keeps things together, keeps them clean, protects from the rain, from rats, from unwanted prying hands.

This is important, and I think often missed: the metaphor must originally have referred to the container, not to the contents. Of course, by the time of the Buddha the metaphor had already lost its grounding and referred to the contents as a whole. But in its original meaning, it referred to the container. But oral texts have no container—or do they?

Who uses baskets?

Containers are such an ordinary part of our life that we give them little thought. We put something in a plastic bag, stick in in a box, or place it in a cupboard, and rarely do we spare a thought for the inner life of the container. How thankless a task! To be nothing in itself, merely useful because of what is in you. I hope, after reading this, that you treat your containers with a little more compassion.

We find containers everywhere in human culture; the simplest hunter-gatherer societies will weave little bags or baskets. But they have few possessions, and little need for a container, beyond keeping together today’s gleanings from the forest. Containers come into their own in the town.

The shift from the village to the town is one of the most important moves in human society, the beginnings of the urbanization that still continues today. And the defining capability that makes towns possible is the ability to store food. In a village, you are never far from the forest, and a day’s hunting, fishing, or gathering is usually enough to supply food. Farming is done, sure, and gradually comes to supplant the gathered food. But it is not as essential as it is for a town. Once you have more than a few families living in a stable location, it is absolutely essential that you have a reliable source of storable food.

This food is primarily grain. This implies a substantial cluster of technologies and infrastructure: roads and carts for transport, granaries for storage; fields with specialist farmers using ploughs and possibly irrigation; and a stable political authority. It also implies a shift in relation to time: no longer is what is gathered in the morning sufficient for the afternoon (as depicted in the Agganna Sutta), but what is harvested each year must be put aside, deferred, sacrificed if you will, for the future. This implies an ability to plan ahead, to conceive of life as a predictable cycle. It also implies a degree of control, or at least understanding, of nature.

In the baskets, then, is held the very life of the community. If thieves were to take it, fire to burn it, water to rot it, or rats to eat it, starvation and the possible annihilation of the society lie close at hand.

If the basket originates as a metaphor for “safekeeping, preservation”, and if this metaphor is of special relevance for the orgins of towns, can we pinpoint it more precisely? I believe we can: at the founding of the Kuru kingdom, roughly 1100 BCE. (The following picture summarizes Michael Witzel’s ideas on the topic in “The Development of the Vedic Canon and Its Schools: The Social and Political Milieu” http://www.people.fas.harvard.edu/~witzel/canon.pdf)

The ancient Aryan peoples, after the fall of the Indus Valley Civilization, spread through north-western India. At the south-easternmost extension, according to the Rg Veda, were the Kuru peoples at the confluence of the Ganges and Yamuna rivers. The centre of gravity of the Vedic peoples continued its south-eastern shift, possibly driven by climate change in the north-west, and so the Kuru region became the centre of the Aryan world. Its remnants are what the archaeologists call the “Black and Red Ware Culture”, and among their technical innovations was the production of iron, unknown in the Rg Veda or the Indus valley.

The Rg Veda as a “collection”

The Vedic literature itself marks the advent of a newly unified and centralized Aryan culture. The literature has a curious structure: the main texts of the Rg Veda, the so called “family books” consist of groups of verses that belong to individual family or clan traditions. They seem very much like the local rites and lore, preserved in jealous isolation, of the sages of the clans. These clan groups were not very large; rather more like villages or clusters of villages than towns. There is agriculture, but more emphasis is on the pastoral and nomadic life, and little respect is there for established buildings and the settled life of towns.

Yet this picture of distinct clan traditions is precisely not how the Rg Veda is handed down, and has not been since the Kuru age. On the contrary, there is a standard, received body of Vedic literature, and all the clans had the works of all the other clans. This is no small thing: the mantras held the secrets to the cosmos, to magic, to success in war, to the favours of the gods. Sharing them would have been comparable to sharing of military intelligence among modern nations. To gather them all in one place would have required persuading the sages of the clans to share their secret lore, and to spend endless hours teaching the memorized verses.

Instead, the individual “Family Books” books are collected within a larger structure, one that is based on a very different way of life. The “Family Books”, 2–8, have been prefaced and addended by new collections of verses. The purpose of these verses is to harmonize and unify the collection. For example, Rg Veda 1.59 speaks of Agni as the lord of all people, the trunk of which all other fires are branches, above the gods, centre of the earth, king of settled lands. And the final verse of the Rg Veda is, even more emphatically, the verse of unification.

HYMN 10.191–4. Agni.

1. THOU, mighty Agni, gatherest up all that is precious for thy friend.
Bring us all treasures as thou art enkindled in libation’s place
2 Assemble, speak together: let your minds be all of one accord,
As ancient Gods unanimous sit down to their appointed share.
3 The place is common, common the assembly, common the mind, so be their thought united.
A common purpose do I lay before you, and worship with your general oblation.
4 One and the same be your resolve, and be your minds of one accord.
United be the thoughts of all that all may happily agree.

RV_10,191.01a saṃ-sam id yuvase vṛṣann agne viśvāny arya ā |
RV_10,191.01c iḷas pade sam idhyase sa no vasūny ā bhara ||
RV_10,191.02a saṃ gacchadhvaṃ saṃ vadadhvaṃ saṃ vo manāṃsi jānatām |
RV_10,191.02c devā bhāgaṃ yathā pūrve saṃjānānā upāsate ||
RV_10,191.03a samāno mantraḥ samitiḥ samānī samānam manaḥ saha cittam eṣām |
RV_10,191.03c samānam mantram abhi mantraye vaḥ samānena vo haviṣā juhomi ||
RV_10,191.04a samānī va ākūtiḥ samānā hṛdayāni vaḥ |
RV_10,191.04c samānam astu vo mano yathā vaḥ susahāsati ||

This is no longer the voice of a tribal sage, an invoker of the gods for the immediate needs of clan. It is the voice of a nation-builder, whose priority is to persuade a disparate and epically fractious association of peoples to act together as one.

Notice the various elements of the hymn: the invocation of the god Agni, unifier of all the people; recognition of the blessings he has brought (my knowledge of the Vedic is insufficient, but it seems to me that what is implied here, as well as material wealth, is the treasures that have just been recited, i.e. the Vedas themselves); the repeated use of terms beginning with sam-, indicating unity; the association of this unity with happiness; and the injunction to action, the actual performance of the fire oblation that is central to Agni’s rite, and which constitutes such a central part of the verses just recited.

Buddhist adoption of the Vedic “container”

This verse has an uncanny resemblance to the final statement in the Buddhist patimokkha, which is recited each fortnight as a statement of unity in the Sangha. A very similar statement is found in all the patimokkhas. The Pali is, characteristically, the shortest though in this instance possibly not the oldest. Here I offer a translation of the Pali, with the additions in brackets being mostly common to the Sarvastivada and Mahasanghika groups of schools.

This much is passed down in the sutta, included in the sutta of the Blessed One (the Tathagata, the Arahant, the Fully Awakened One), which comes up for recitation each fortnight (and also other major and minor teachings that have come down). Therein each and every one should train (together), in harmony, with mutual rejoicing, without dispute, (with unified recital, one like mixed milk and water, living happily and at ease).

Ettakaṃ tassa bhagavato suttāgataṃ suttapariyāpannaṃ anvaddhamāsaṃ uddesaṃ āgacchati. Tattha sabbeheva samaggehi sammodamānehi avivadamānehi sikkhitabbanti.

Here, for comparison, is one of the Sanskrit texts, that of the Mahasanghika:

etakoyaṃ punastasya bhagavato tathāgatasyārhataḥ samyak saṃbuddhasya dharmavinayo prātimokṣasūtrāgato sūtraparyāpanno yo vā anyopi kaściddharmasya anudharmo tatra samagre hi sarvve hi sahite hi saṃmodamāne hi avivadamāne hi ekoddeśe hi kṣīrodakī kṛtehi śāstuḥ śāsanaṃ dīpayamāne hi / sukhañca phāsuñca viharante hi anadhyāvācāya śikṣākaraṇīyā

All five of the aspects we have noted recur here: the invocation of the Buddha; the acknowledgement of the treasures gathered in the text just recited; the repeated use of sam- to indicate unity; the association of this unity with happiness; and the injunction to action by training in accord with the rules just recited.

The conclusion seems inevitable: the Buddhist Sangha copied the literary form of the Veda, specifically when it came to the concept of the framing narrative, the bookends that preserve the fragile words, in other words, the basket.

This inheritance is even more apparent in the Sanskrit texts than the Pali. I have checked 4 Sanskrit patimokkha texts, those of the Mulasarvastivada, Sarvastivada, Mahasanghika, and Mahasanghika-lokuttaravada. As you can see, these represent two of the main groups of early schools, the Sarvastivadins and Mahasanghikas. Generally speaking, agreement between these texts is a sign of a pre-sectarian heritage. Perhaps the agreement of these texts is a sign that the Pali has, in this instance, undergone loss of text.

All the Sanskrit texts add the word sahita, which I have translated in brackets above as “together”. But this is the normal word used for the Vedic collections, as in the “Rg Veda Samhita”, the “Collection of the Rig Veda” (and other Vedas). The word sahita in fact appears in Pali as a poetic term for sacred scriptures, in Dhammapada 19-20, where the commentary confirms that this is a word for the Tipitaka of the Buddha’s words. I find it to be a remarkable coincidence, if coincidence it be, that this passage, with so many evident connections with the Vedic Unity verse, should include the very word for the Vedic collections themselves!

This passage is a stock one, and appears in many places. In fact, the passage we have just quoted is a secondary one, for the Buddha is referred to in third person: this is not direct Buddhavacana. In other places, however, similar words are attributed to the Buddha, notably in the Kinti Sutta (MN 103). There, the Buddha is concerned to downplay the significance of Vinaya as the source of harmony, and emphasizes rather the harmonious practice and recital of the 37 Bodhipakkhiya dhammas, using the same phrase we have seen above. Many scholars, myself included, have seen in this and similar passages a hint that the earliest canonical collection of Buddhist teachings consisted largely of these 37 Dhammas. It seems that the redactors borrowed the phrase to use in the conclusion for the patimokkha.

The Buddhists were aware, almost painfully so, of the Vedas and of the prestige and influence that this ancient body of knowledge bestowed on the Brahmins. It must have been frustrating, since the Buddhists believed that their teachings were far superior to what they regarded as the endless, vapid rituals and invocations of the Vedas. Yet despite the many limitations of the Vedic literature, they got one thing right: it lasted.

The literary innovations of the Kurus included gathering the family hymns, organizing them in minute detail (for the structure of the Rg Veda is subtle, rigorous, and complex), adding the framing narrative, and organizing for all the “families” to recite the whole scripture. The result of this was that the text was preserved with uncanny precision for thousands of years in the oral tradition. It was nearly a thousand years old when the Buddha was alive.

The problem facing the Buddhist Sangha was not dissimilar from that facing the Kurus. Just as the Aryan people were up to that stage semi nomadic, semi settled people, the Sangha was semi-nomadic, semi-settled; just as the Aryans were united in a general cultural sense, so were the Buddhist Sangha united in a sense of belonging to the same community; and just as the Aryans were divided into independent clans each with its own scriptures, so the Buddhist Sangha was divided into various monasteries, districts, and teacher’s lineages, each with its own characteristic texts.

It is little wonder, then, that given such similar problems, and given the highly successful solution to that problem implemented by the Kurus, the Sangha used the Vedic precedent. Note that in the passage above we are hearing the voice of the Sangha, not of the Buddha, as he is referred to in the third person as Bhagava.

Everything is neatly wrapped

If we consider the Buddhist texts in this light, this principle is absolutely pervasive. Each sutta starts with an introduction (nidana) giving the setting and main characters, and asserting that what is to follow is the words of the Buddha (usually). It then finishes with the monks rejoicing (and in several traditions, determining to practice what was taught). Thus each text is defined by its nidana, as if it were wrapped in a leaf for safety. Without such definition, each “sutta” would blur into the next, and they would become irretrievably mixed.

Of course, this is an ideal situation, for in reality not every sutta, especially the shorter ones, actually has such a nidana. But there is a sense in which they all are ideally supposed to have one; and the inadequacy of the current situation is seen in that the suttas do in fact blur into each other from time time. Bhikkhu Bodhi, for example, in his recent translation discusses the cases where it is difficult to tell whether a text should be considered as one sutta or two.

Not only is each sutta “wrapped” as by a leaf, they are then “wrapped” in groups of ten to form a vagga, or “chapter”. Each “chapter” is then wrapped in a larger section, and so on. The largest scale “wrapping” is the life of the Buddha himself, within which all the events and teachings of early Buddhism take place.

This puts in a new light the Vinaya passages that give instructions on redaction technique. With his usual cynicism, Gregory Schopen, titles his essay on this topic, “If You Can’t Remember, How to Make it Up”. But this is precisely what the texts he quotes are not saying.

The passage refers to the situation where a monk has difficulty remembering the setting or other framing details of a sutta. In such a case he should insert the usual tropes (e.g. by saying the sutta was at Savatthi and so on).

In light of what we have seen above, the purpose of this is precisely to preserve and maintain the actual sutta—the words of the Buddha—by ensuring that they are “wrapped” well. That the incidental details on the wrapping are not authentic is clearly of little importance: what matters is the actual substance of the teaching.

Likewise, if there is difficulty remembering the sutta as a whole, it should be written down. Again, this is a method for preserving the text without change, not a method for inventing a new text, as Schopen implies.

This reasoning applies to the majority of the techniques that are often regarded as “artificial” in the Pali and other Early Buddhist texts—repetitions, stock phrases, unified language, and so on. Each of these was employed for the same purpose: to make it possible to preserve the teachings as-is, without alteration, except in incidental details of structure and organization.

Just as the Vedic redactors formed a resilient, lasting “basket” in which to preserve their sacred texts, the Buddhists created a “basket”of astonishing sophistication and flexibility which has, despite all the ravages of time, succeeded in its purpose: to preserve the word of the Buddha.

And what of the soul?

Which brings us at last to the little fable at the start of this essay. It is one of a large genre of stories, of which Frazer gives many more examples, and which are particularly prevalent in India. The King (or the wizard or the ogre) keeps his soul in a safe place, stored far away. He cannot be killed as long as the soul is safe, locked away in apparently impenetrable “baskets”. In just the same way, the people of the town keep their grain in a safe storage, and they will survive as long as the grain is safe.

Similarly, the Buddhist community keeps its scriptures locked away in safe places, stored in the basket, the cabinet, the shrine. If the texts were to be destroyed, we would lose our very life, our soul; Buddhism would be cast adrift, like a garland that is not tied together with string. The fable warns of false pride and complacency. No matter how well we think our “soul” is guarded, sooner or later it will meet its end. Thus it will be for all the Buddhist scriptures, and for the Buddhist religion as a whole. The early Buddhist community, attuned to the contemplation of impermanence, was keenly aware of this, and used every means at their disposal to postpone that grim day as long as possible.

The body as metaphor

While we’re on the topic of misconstrued meditative metaphors, here’s another chestnut that well and truly deserves roasting: the body. The formula for third jhana mentions that one ‘experiences bliss with the body’. Most interpretations of jhanas say that they are purely mental experiences, based on the unification of mind-consciousness, and that it is impossible to experience anything through the five senses while in such a state.

But then, we can’t just have everybody agreeing on everything, can we, because that would be just so so dull. So others take the word body quite literally here, and say that this shows that we can experience the body (and other physical senses) in jhana.

You’re probably guessing that I’m going to side with the non-literalists here, and you’re quite right. I’ve discussed this in more detail elsewhere, but I just noticed this little sutta that brings out the metaphorical nature of the language used in higher Dhammas quite nicely. Here it is, Anguttara 4.189.

Bhikkhus, these four are things to be realized. What four?

There are things to be realized with the body, to be realized with mindfulness, to be realized with the eye and to be realized with wisdom.

What should be realized with the body? The eight liberations.

What should be realized with mindfulness? Previous births.

What should be realized with the eye? The passing away and rebirth of beings.

What should be realized with wisdom? The ending of defilements.

Bhikkhus, these four are things to be realized.

Notice especially here the use of ‘body’ and ‘eye’. Now, it is clearly quite impossible that ‘eye’ means a physical eye here; no-one would argue that one can physically see beings getting reborn. In this context of subtle, abstruse, higher Dhammas, the eye is not a physical eye, but a metaphor for a refined inner vision.

And in just the same way, the body is not a physical body, but a metaphor for the wholeness and directness of experience. As if this were not obvious enough from the context, notice that the things to be realized with the body are the eight liberations, which include the four formless attainments. These are by definition beyond any kind of physical reality. Elsewhere, the Buddha says that even Nibbana is to be realized with the body.

The body is not the body, the eye is not the eye, and thought is not thought. These are all words, inadequate, struggling, messy words, creeping up from the evolutionary slime, groping and grasping towards the light. As long as we keep them weighed down by the mundane, we can never speak of higher things. And since these higher things are things of the mind, if we cannot speak of them, we cannot imagine them. And if we cannot imagine them, we cannot realize them. And that is rather a sad state of affairs.

Why vitakka doesn’t mean ‘thinking’ in jhana

I shall give you a simile; for it is by means of a simile that some wise people here understand the meaning of what is said.

—THE BUDDHA

Here’s one of the most often contested issues in Buddhist meditation: can you be thinking while in jhana? We normally think of jhana as a profound state of higher consciousness; yet the standard formula for first jhana says it is a state with ‘vitakka and vicara’. Normally these words mean ‘thinking’ and ‘exploring’, and that is how Bhikkhu Bodhi translates them in jhana, too. This has lead many meditators to believe that in the first jhana one can still be thinking. This is a mistake, and here’s why.

Actually, right now I’m interested in a somewhat subtle linguistic approach to this question. But I’ve found that if you use a complex analysis of a problem, some people, understandably enough, don’t have time or interest to follow it through; and often we tend to assume that if a complex argument is just a sign of sophistry and lack of real evidence. So first up I’ll present the more straightforward reasons why vitakka/vicara don’t mean thinking in jhana, based on the texts and on experience. Then I’ll get into the more subtle question of why this mistake gets made.

For most of this article I’ll just mention vitakka, and you can assume that the analysis for vicara follows similar lines.

Meaning & etymology

Already in the Pali Text Society dictionary we find the combination vitakka & vicara rendered as ‘initial & sustained application’. This was taken up by Ven Nyanamoli in his translations, but was later removed by Bhikkhu Bodhi as he strove to complete Nyanamoli’s project of effectively finding one English word to translate each significant Pali word.

Etymologically, vitakka harks back to a Sanskritic term (vi-)tarka. This appears in both Pali and Sanskrit literature in the sense of ‘thought’; but more pregnantly also as ‘reflection, reasoning’; in some cases more pejoritively as ‘doubt, speculation’. The Pali Dictionary suggests it is from an Indo-European root, originally meaning ‘twisting, turning’, and related to the English ‘trick’. However, I can’t find any support for this is Indo-European dictionaries; nor can I find it in the Vedas.

In the Suttas

The primary source work is the Dvedhavitakka Sutta (MN 19). This is where the Buddha talks in most detail about vitakka specifically, and describes how he discovered and developed it as part of the ‘right thought’ (sammasankappa) of the eightfold path. Note that the terms sankappa and vitakka are often, as here, synonyms.

The Buddha describes how he noticed that thinking unwholesome thoughts leads to suffering, while thinking wholesome thoughts leads to happiness. And he further realized that he could think wholesome thoughts nonstop all day and night, which would not lead to anything bad; but by so doing he could not make his mind still in samadhi. So by abandoning even wholesome thoughts he was able to enter on the four jhanas.

A similar situation is described in AN 3.101. There, the Buddha speaks of a meditator who abandons successively more refined forms of thought, until all that is left are ‘thoughts on the Dhamma’ (dhammavitakka). Even these most subtle of thoughts prevent one from realizing the true peace of samadhi, so they must be abandoned.

Clearly, then, the right thought of the eightfold path, even thoughts of the Dhamma itself, must be abandoned before one can enter jhana.

In experience

Let’s not even worry about experience of the jhanas; then we’d just end up trying to define what a jhana is. Let me give you a test. Sit quietly, now, for five minutes. Watch your mind, and notice what happens when you think and when you don’t think.


Okay, done now? What happened? Well, let me guess: most of the time you were thinking of this or that, but occasionally there were spaces of silence. And those spaces of silence were more peaceful. Even this much, even just a few minutes of sitting quietly, and you can experience the peace of a quiet mind. And yet in jhana you’re still thinking? Impossible!

Not to mention jhana, anyone who has been on a meditation retreat will have experienced those blessed moments, sometimes several minutes or longer, when the mind is clear, still, and silent. Not all the hindrances are gone, and not all the jhana factors may be present, yet there is a degree of stillness.

How language evolves

If vitakka does not mean thinking, then why did the Buddha use such a misleading word? The answer is simple: it was the best he had. Why this is so, and how such situations can arise, is a fascinating question that takes us into areas of linguistic philosophy, specifically, how we develop words for speaking of refined topics.

My understanding in this area was sparked by Julian Jaynes, who devoted quite some time to this topic in his magnum opus, The Origin of Consciousness in the Breakdown of the Bicameral Mind. I don’t have the book with me, so this comes from my (usually unreliable) recollections.

The basis of his ideas can be expressed in some simple axioms. The first:

  • Axiom 1: All abstract words are derived from more concrete words by way of metaphor.

By metaphor here I don’t mean, of course, the conscious use of metaphor as a poetic device. I mean the embedded use of metaphor that pervades all language; like, say, the use of ‘embedded’ in this very sentence.

The idea is that, whether considering the origins of language in history, or the learning of language by an infant, we must begin with what is concrete. We point to the earth and say, ‘ugh’, then point to the water and say, ‘erg’. I can’t point to ‘solidity’ or ‘liquidity’. We must gradually learn these abstract concepts based on the more concrete ones.

There is a universal pattern we can discern in this process:

  • Axiom 2: Metaphors move from what is better known to what is less known.

We start with knowledge that is shared. But when one person learns something that others have not, they must draw the others on from what is known towards what is unknown. Jaynes called these things the ‘metaphier’—the relatively concrete, well-known thing on which the metaphor is based—and the ‘metaphrand’—the relatively abstract, less-known thing that the metaphor is intended to illustrate.

Which leads us to our third axiom:

  • Axiom 3: A metaphrand brings something over from the metaphier, and leaves something behind.

If the basis on which the metaphor is made (the metaphier) has nothing in common with the object of the metaphor (the metaphrand), then there would be no illumination. On the other hand, if they had nothing different, they would be the same thing.

But what is it that is common, and what is lost? Since we are speaking of the movement of language from the coarse to the subtle, we can say that:

  • Axiom 4: When words are abstracted, subtle aspects of the metaphier carry over to the metaphrand, while coarse aspects are left behind.

This is all very abstract, so how about some ‘concrete’ examples. Let’s look closer at ‘earth’.

In English, we have two different words for ‘earth’ (as in the ground, not the planet) and for ‘solidity’. This is such a natural part of our language that we don’t think that it’s anything special.

In Pali, by contrast, the same word, pathavi, is used for both ‘earth’ and ‘solidity’. (There are other words for these, too, but I will keep it as simple as I can). In, say, a Vinaya text that discusses digging, it is clear that pathavi just means ‘earth’ in the ordinary concrete sense of the dirty stuff in the ground. On the other hand, in a philosophical or meditative text that discusses the contemplation of the ‘earth element’ (pathavidhatu), it is clear that a more abstract notion is meant. Parts of the body such as the skin, bones, and hair, are said to be the ‘earth-element’, so clearly this doesn’t mean ‘dirt’. In fact, pathavi is given an explictly abstract definition in the Suttas as ‘hardness, solidity’.

Both languages have a concrete idea of ‘earth’ and an abstract idea of ‘solidity’. And from the Pali it seems obvious that one rose from the other. However, from the English perspective we can’t see that in this case; the metaphorical roots of ‘solidity’ are lost in the mists of time. We no longer feel it as a metaphor. It is just a word that means what it says. In times long past, however, it must have arisen from its own metaphorical roots, which may or may not be the same ‘earth’; in fact, the etymologists say that ‘solid’ is from an ancient Indo-European root *solo-, originally meaning ‘whole’. Since ‘whole’ is itself an abstract concept it must have come from a still deeper metaphor. Interestingly enough, ‘earth’ is also an Indo-European term, meaning ‘ground’; but neither of these is related to pathavi.

So, while the general process is universal, the historical details are arbitrary. Why a language abstracts a certain word and keeps another close to its roots depends on all kinds of random factors. It is not simply that English is a more evolved language than Pali.

Take, for example, the Sanskrit term trsna. This means ‘thirst’, and the English word is indeed derived from the same Indo-European root (which originally meant ‘dry’) and keeps the same meaning. On the other hand, in Pali trsna is split in two: tasina stayed close to its metaphier, and means primarily ‘thirst’, while tanha has almost totally lost its metaphorical connections and just means ‘craving’.

Notice, also, that these words can themselve be used as the basis for further metaphors. We can speak of a ‘solid’ character, or an ‘earthy’ character; but these are not the same kind of thing. Similarly, we can have a ‘thirst’ for knowledge, or tasina can be used to mean craving, just like tanha.

But in all these cases we still feel the metaphor. The words stay close enough to their concrete roots that we know their meaning is being stretched to new forms.

This topic of how language evolves is a fascinating and profound one, and we could take it in all manner of directions. But for now let’s return to our main topic, the Buddha’s description of jhana.

How did the Buddha speak about jhana?

Following the principles sketched out above, what can we say about how the Buddha spoke of jhana?

One thing that seems clear from the historical record is that the Buddha was the first teacher to describe in straightforward, empirical terms the experiences of higher consciousness. Earlier teachings, such as the Upanishads, seemed so overwhelmed by states of transformed consciousness that they had no choice but recourse to a mystical evocation of a divine encounter.

The Buddha, in what must have been a striking innovation, used only simple, empirical terms to describe jhana and other states of higher consciousness. In common with his typical empiricist approach, this means that he used words that remained as close as possible to their ordinary meanings. He wanted people to understand these states, to refer to their ordinary consciousness, and to see how that can be developed and transformed to become something wonderful.

So there is this twofold tendency. On the one hand, the Buddha emphasized countless times how powerful and radically transformative the jhanas were. They are the ‘higher mind’, the ‘expanded mind’, the ‘unexcelled mind’, the ‘radiant mind’, the ‘liberated mind’, the ‘light’, the ‘bliss of Awakening’, the ‘end of the world’; they are ‘beyond human principles’, and are ‘distinctions of knowledge and vision worthy of the Noble Ones’.

At the same time he emphasized how attainable they were. If one is dedicated to following the full course of training that he outlined in places such as the Samannaphala Sutta, one could realize a gradual evolution of blissful consciousness eventually culminating in the full release of jhana.

Any understanding of jhana must take full account of both these aspects, neither reducing jhana to an mundane state of easily-attained relaxation, nor making them so exalted and abstract that they seem unreachable.

I should notice, incidentally, that the common expression found in Abhidhamma literature of ‘mundane jhana’ is very misleading. This has nothing to do with the experience of jhana itself. It simply means that jhana, when practiced outside the eightfold path, leads to rebirth.

What do the words in the jhana formula mean?

If we look closely at the terms in the jhana formula, then, we find that they are words that have a more coarse physical or psychological meaning in everyday language. They are common words that everyone can understand, and can relate to their own experience. And in every single case, they clearly have a more subtle, abstract, evolved meaning in the context of jhana. We have moved from the ordinary mind to the ‘higher mind’, and everything about the experience is transformed.

So, for example, the first word in the formula is viveka. This normally means physical seclusion; going away from others into the forest or a solitary spot. In jhana, however, it refers to a mental seclusion, where the mind turns away from the senses and withdraws into itself. The Pali texts make this distinction clear, as elsewhere they speak of three kinds of seclusion: physical, mental (i.e. the jhanas), and seclusion from all attachments (Awakening).

The next word in the formula is kama. In ordinary language this means the pleasures of life, especially sex, but also food, drink, luxuries, and other pleasures of the senses. In jhana, however, it has a more subtle nuance, referring to the mind that inclines to taking pleasure in any experience through the five senses.

Then there is the word akusala. Normally this means ‘unskilful’, as, for example, someone who is no good at a certain craft. One who is kusala, on the other hand, is clever and adroit. In the jhana formula, however, kusala includes any tendency of the mind that creates suffering.

Similarly there is the word dhamma, which is what akusala qualifies. Dhamma in ordinary language has a variety of meanings, such as ‘law’, ‘custom’, and so on. In jhana, however, it takes on a far more subtle meaning, that is, any object, quality, or tendency of the mind. The akusala-dhammas, or ‘unskilful qualities’, especially refer to the five hindrances which must be abandoned before entering jhana.

And so on. I could go on through the entire jhana formula and show how each word is related to, but abstracted from, its more concrete everyday basis, its ‘metaphier’. But I think that’s enough examples.

So what do vitakka & vicara mean?

Finally we are ready to return to our original question. Now we can look again at the claim that vitakka must mean thinking in jhana, because that’s what it means in everyday discourse. And I trust that this claim now appears a lot less plausible than it might have earlier.

If this is true, then vitakka (& vicara) are the sole exceptions. Every other term in the jhana formula takes everyday words and transforms them, in what the Buddha emphasizes at every turn is a special, exalted, and refined context. Only vitakka is exempt from this, and means exactly the same thing in higher consciousness as it does in lower consciousness.

This argument is not merely implausible, it is totally impossible. Words just don’t do that. And they specially don’t do that in a context like jhana, where the very point of the state of mind is that it is integrated and whole. How can such a coarse, ragged, disturbing thing as ‘thought’ continue, while everything else has become so refined?

Let us consider again our Axiom 4: When words are abstracted, subtle aspects of the metaphier carry over to the metaphrand, while coarse aspects are left behind.

Sit again for a couple of minutes. This time, don’t be quiet: have a think. Look at what thinking is like. Raise a question: what is the nature of thought? Then stop: be silent: look at the space that reverberates after the words have ended.

When you think, the most obvious aspect, the coarsest aspect, is the verbalizations. But they don’t happen alone. There is a kind of lifting of the mind onto an object. This is normally quite subtle, and we don’t notice it because we are interested in the words. It becomes more obvious sometimes when you try to think about something, but your mind is not really interested. It’s as if you keep moving the mind towards that topic, but nothing much happens. You can also feel it when the words stop. The ‘thought’ in some sense is there, apart from the verbalizations. It’s a subverbal thought, a placing or hovering of the mind in a certain way.

This is what vitakka refers to in jhana. This is the subtle aspect of ‘thought’ that is carried over into jhana, when the coarse aspect, the verbalization, is left behind.

And as with vitakka, so with vicara. Vicara is the ‘exploring’ of something, and in ordinary language refers to wandering about a place on foot. Psychologically, it normally means a more sustained reflection or examination of a thought, a keeping in mind of the topic that vitakka has brought to mind. In jhana, it follows the same process. The coarse verbal reflection is long gone, and in its place is the gentle holding or pressing of the mind with its object.

Early definitions

Unfortunately, there are no further definitions of these terms in the very early strata of texts. However, in the next strata, the late sutta/early Abhidhamma phase, we do have definitions. Our first example comes from a sutta in the Majjhima Nikaya; on text-critical grounds, however, it seems this should be viewed as a proto-Abhidhamma work. Sankappa is defined in the Mahacattarisaka Sutta (MN 117) as takko vitakko saṅkappo appanā byappanā cetaso abhiniropanā vacīsaṅkhāro. Vitakka is included in this definition; and notice the last term, cetaso abhiniropanā, which means ‘application of the heart’.

The earliest Abhidhamma text, the Vibhanga, gives a similar definition of vitakka in the context of jhana: takko vitakko saṅkappo appanā byappanā cetaso abhiniropanā sammāsaṅkappo. This text also adds a similar definition of vicara: cāro vicāro anuvicāro upavicāro cittassa anusandhanatā anupekkhanatā. Notice the last terms here: ‘sustained (anu- application (sandh) of the mind, sustained (equanimous) observation (ikkh)’.

Don’t be fooled by the fact that these definitions include both ordinary and abstract terms. This is merely a feature of the Abhidhamma definitions in general. They are concerned to show the range of meanings that terms have in different contexts, so that one can understand what terms have the same or different meanings in various sutta passages. It is a means of referring to and defining terminology, and it is not meant to imply that they have the same meaning in all these cases. On the contrary;, the overall tendency of these definitions is exactly as we have been describing: they move from the relatively coarse to the relatively subtle.

Those who are proponents of the ‘vitakka always means thinking and nothing else’ school of interpretation will, of course, reject these texts as inauthentic. And they are quite right; I would not try to argue that these definitions came directly from the Buddha. But that does not mean that the definitions are wrong. They come from a time shortly after the Buddha, likely within a couple of hundred years, when the monks were still immersed in the early Suttas and, crucially, spoke Pali (or something very like it) as a native tongue. They had access to a far more diverse and richer linguistic context than we do, and their opinions must be taken seriously. While on a doctrinal level it is true we can see certain (minor) shifts from the Suttas to the early Abhidhamma, linguistically they belong to the same period, and we would need strong and clear grounds before rejecting their linguistic explanations.

The waywardness of language

Consider once more the process of the gradual abstraction of words from a more concrete metaphorical basis (metaphier) towards a more abstract metaphorical object (metaphrand); from the relatively coarse thing that provides the illumination to the relatively subtle thing that is illuminated. As we saw above, this process is largely arbitrary. Accidents of history, anthropology, and usage will influence which words get used in which sense, and this process will occur in different ways in different languages; and even within the same language.

One of the consequences of this arbitrariness is that there is a certain unpredictability, even obfuscation, in how abstract words are formed. The speaker intended certain aspects of the underlying metaphier to be carried across to the metaphrand, while the listener understood something else. This happens all the time, and is the main reason why, in any higher discipline, experts spend a lot of time arguing over terminology. We can’t simply agree on the meaning of a word by pointing to what it stands for and saying it.

One of the most intriguing ideas that Jaynes introduced was the notion of ‘paraphiers’ and ‘paraphrands’. These are unintended implications or connotations that are carried over from the original idea to the subsequent one. Central to Jaynes’ thesis, in fact, is the highly challenging notion that our ability to consciously reflect on ourselves as subjects arose in just such a way.

Leaving aside this intriguingly counter-intuitive idea, Jaynes’ essential point is that the paraphiers direct our attention in unexpected ways. And attention creates realities. This is not merely a matter of a kind of poetic allusion or idea. When our minds are drawn towards something—perhaps a new way of seeing or thinking—this creates a new world in our mind, and as we know from our basic Buddhism, such new mental worlds create the world outside.

In the context of jhana, the notion that vitakka always means thinking and nothing else creates realities in meditation. It encourages certain kinds of expectations and responses. By doing so it shapes the nature of the meditative experience. This in turn effects speech about meditation, and a whole range of more concrete realities: books, retreat centers, teaching careers, relationships, organizations.

This is another fascinating aspect of Jaynes’ theory. The process of abstraction creates powerful mental worlds that then become expressed in material forms, thus returning from the abstract to the concrete. The forms that emerge as expressions of the mind then serve to reinforce and validate the particular mental abstractions that gave rise to them in the first place. Jaynes discusses how this happens in religions through the creation of idols, temples, and the like. When enough people share an idea, they band together to create physical representations of their own mental world; and these physical representations in turn confirm and reinforce the idea.

It is in this way, I believe, that the innocent term vitakka has taken on a whole new life. In Pali it had a certain spectrum or flexibility of meaning, such that the Buddha could prod it out of its everyday meaning of ‘thought’ and tease it into a new meaning, ‘application of the mind on to its object in profound meditation’. The English word ‘thought’, however, lacks such flexibility, and remains stubbornly and exclusively verbal. When used as a metaphier for the less-knowable ancient word vitakka, the unexpected and unintended connotations of thought, its paraphiers, are transferred over.

The process of jhana is, at its heart, nothing more than the deepening stillness of the mind that lets go of all pre-occupations and worries. The Buddha used, as he must, everyday words to point to something that moved beyond the everyday. And it is no small irony that one of the crucial terms in this journey from perplexity to stillness, a word whose less edifying connotations include ‘doubt, speculation, the endless twists and turns of the mind’, has itself provoked such doubts and endless discussions.

The first Jataka?

I’ve been taking the opportunity to read Bhikkhu Bodhi’s wonderful (as always!) translation of the Anguttara Nikaya. If you haven’t got it, what are you waiting for?

I noticed a little sutta in the threes, and it struck me that this humble little text is probably the best candidate for the title of the first Jataka story in Buddhism. The text is AN 3.15 Pacetana. There are no known parallels for this text, which is, however, not unusual for the Anguttara Nikaya.

What are the Jatakas?

Jatakas tell of events in the past lives of the Buddha, and sometimes of other personalities from the Buddha’s life. They are everywhere in traditional Buddhism; told in sermons, read to children at bedtime, recited in great ceremonies, depicted in artwork. Yet one of the striking things about Buddhist texts is how few there are in the early teachings. While the Pali tradition alone preserves over 500 stories canonically, and many more in later collections, only a dozen or so are found in the early Nikaya/Agamas. There is an excellent essay on the topic by TW Rhys Davids. It’s a little dated, but still well worth a read. However, Rhys Davids does not notice our current sutta.

I’m greatly struck, in fact, by how few stories there are in the Anguttara (and other Nikayas). Buddhism is one of the greatest story-telling traditions in the world, and yet the Buddha himself seems not to have told many stories at all. The vast majority of suttas are straightforward statements or dialogues on ethics, meditation, and the like. They are often illustrated with similes, and somewhat less frequently with short parables. But there is very little in the way of extended narrative; and most of the narrative that there is is by way of background, not spoken by the Buddha himself. Of course there are exceptions, like the Agganna and Cakkavattisihanada Suttas of the Digha; but few and far between. Why this is, I do not know.

The Pacetana Sutta is another exception. It is a simple story, at the end of which the Buddha identifies himself as the main character, thus qualifying it as a Jataka. While there are several other Jatakas in the early Agamas, most of them contain features that strongly suggest they are of a later date than most of the early Suttas. Alone among the Agama-Jatakas, so far as I can tell, the Pacetana contains a range of features suggesting that it is early.

The Story

Here’s a summary, from the Dictionary of Pali Proper Names. You can find the full translation here.

There was once a king called Pacetana who asked his wheelwright to make a pair of wheels for a battle which was to take place six months later. When but six days remained of this period, only one wheel had been made, but the other was finished within the stipulated time. Pacetana thought that both wheels were alike, but the wheelwright proved to him that the one he had made hurriedly was faulty in various ways, owing to the crookedness of its parts. The Buddha identified himself with the wheelwright and declared that one must be free from all crookedness in order not to fall away from the Dhamma and the Vinaya.

Why is it special?

This is one of those cases where there is not one thing that is particularly unusual. Rather, there are many little details that taken together suggest, to me, that the text is somehow distinctive, and may have originally been, or been considered as, the first Jataka story. Here’s a list of points I noticed.

  1. The king is unknown. He is not a stereotyped king, such as the Brahmadatta found in so many later Jatakas.
  2. The Buddha does not identify himself as a Bodhisatta in the past. This is a common feature of all the early Jatakas. There is no suggestion here that the Buddha in the past knew that he was destined for Awakening, or that he was engaged on a spiritual quest spanning many lives, or indeed that what he did in that life had anything to do with Awakening. In other early Jatakas, in fact, he specifically denies that the practices of those days lead to Awakening. (Bearing this in mind, I won’t refer to him as the Bodhisatta in this discussion.)
  3. While in most canonical Jatakas the Buddha identifies himself as a great king or sage of the past, here the Buddha identifies himself with a lowly chariotmaker. In many other suttas of the Anguttara, the chariotmaker is listed as a lowly, menial occupation like a scavenger, flower-collector, or rubbish-sweeper. (Ven Bodhi’s translation obscures this point somewhat; the translation uses the more distinguished “chariot-maker” here, and “cart-maker” in the other contexts; but the Pali in both cases is rathakara.)
  4. There is a distinct absence of miracles and wonderous events. Even when pressed by the king, the chariotmaker is unable to do his job well. He is clearly constrained by the usual demands of his craft, unlike the near-superhuman abilities of Bodhisattas in many other stories.
  5. The chariotmaker would seem to be engaged in wrong livelihood, or at best an ethically dubious trade: making weapons of war. There are, it is true, many Jatakas of the later periods where the Bodhisatta is depicted as breaking various precepts, but this is still noteworthy. It is also perhaps significant that the chariot, specifically the two-wheeled chariot, was the distinctive war vessel of the Aryan people and was, it seems, the decisive technological innovation that spurred their great success in spreading their culture across the world; as unstoppable as the Wheel of Dhamma itself…
  6. The moral of the story is the importance of gradual development. While not at all unusual in the Suttas, this differs from certain later trends, which emphasized an instantaneous realization.
  7. There is considerable confusion about the name, both of the king and of the sutta. Variants include Pacetana, Sacetana, Paccetana, etc.; and the sutta is sometimes called “Cakkavatti”. This is interesting, since the sutta refers to the “rolling of a wheel”, but not to the “Wheel-turning Monarch”, which is what the Cakkavatti usually refers to. The previous sutta, in fact, mentions the Wheel-turning Monarch. There seems to be some confusion; perhaps—and this is very speculative—the Pacetana Sutta was the kernel from which the idea of the Wheel-turning Monarch was derived.
  8. On the same topic, and on more solid ground, it is very striking that this sutta, which deals with “rolling forth a wheel”, is said to have been set at Benares, in the Deer Park. This is, of course, where the Buddha taught his first sermon, the Dhammacakkappavattana Sutta, the “Rolling Forth of the Wheel of the Dhamma”; and the same words are used here. While some later Suttas were taught there, it was not a very common setting. This detail is even more striking when we realize that very few of the suttas in the Anguttara have a proper setting. In almost all cases they simply have an abbreviated setting, or none at all. When the settings are included, it is usually because they have some special relevance for the teaching. So it seems certain that the setting is significant here. Probably it is meant to establish some connection with the First Sermon. And perhaps it is meant to suggest that this is the First Jataka…
  9. Not only is this the first Sutta in the Anguttara spoken by the Buddha to be given a proper setting, it is the first story told by the Buddha in the Anguttara.
  10. Unlike the later Jatakas and most of the more substantial suttas, this does not have an ABA structure. Rather, the story is told, then the moral is drawn out. Once again, in itself this is nothing spectacular, but it does suggest, however mildly, a lack of systematic revision.
  11. The king converses in quite a familiar manner with his chariotmaker. This doesn’t sound like a magnificent monarch of the Moriyan era, nor yet a legendary king of old. It sounds like a little lordling of a smallish garrison city. The preparations for war are, to say the least, perfunctory: one chariot! Of course, in some ways things are no different today: military contracts still don’t come in on time…
  12. When the wheel is rolled forth, the text speaks of giving the wheel “impetus” to roll. This is, in Pali, abhisaṅkhāra. This term is more normally found in the Pali texts in a more refined, abstract usage, where it is equivalent to cetanā, or intention; in fact, it tends to be used in somewhat technical discussions of kamma and the like. Here it appears in its simpler, older, physical meaning. I wonder whether there is any connection with the king’s name: Pacetana, the “One Whose Will is Done”, perhaps?
  13. Similarly, when describing the flaws in the wheel, the text uses terms in a physical sense that are more often found in a psychological sense in Pali: savaṅkā sadosā sakasāvā. Also kusala is found in its older sense of “skilled” rather than the more familiar ethical sense.
  14. In yet another unusual feature, when the Buddha draws out the moral of the story, he says that if “any bhikkhu or bhikkhuni” sees flaws in themselves, they should abandon them. The inclusion of bhikkhunis like this is very unusual in the Pali texts; a quick search only turns up four or five similar cases. Call me biassed, but I have a suspicion that the editors of the Pali canon, whether by accident or design, excluded the bhikkhunis by default. The Satipatthana Sutta, for example, mentions the bhikkhunis in the Sarvastivada version, not in the Pali. If this is correct, it is another sign that this sutta may be early, and has escaped significant editorial alteration.
  15. The text conforms to the “waxing syllables principle”. This means that in various lists of terms, the words with more syallables come later in the list, such as the phrase I quoted earlier: savaṅkā sadosā sakasāvā. This has been shown in detail by Mark Allon to be an outstanding stylistic feature of the early Pali texts, and an indication of their origin in oral culture.
  16. Perhaps the most significant feature of all is the text’s use of numbers. Just like the Dhammacakkappavattana Sutta, the text is carefully built around a subtly interlocking set of numbers. Where the Dhammacakkappavattana Sutta features “two extremes”, the eightfold path, and the “four noble truths”, in “three rounds” totalling “twelve aspects”, the Pacetana Sutta has the two wheels, each with three parts, each with three potential flaws. These are to be completed in 6 months, but at 6 days short of 6 months only one is done; and the second, poor quality job, is finished in 6 days. The three parts of the wheel parallel the three trainings, each of which also has three potential flaws. While the text doesn’t draw out the connection, it seems a parallel is implied thus: rim=body; spokes=speech; hub=mind. Now, the use of numbers in this way can be seen from a number of angles. But what is this text telling us at its heart: that careful craftsmanship produces a long-lasting, stable wheel. And the long-lasting and stability of the Wheel of Dhamma was indeed one of the major concerns of early Buddhists. I suspect our Sutta is suggesting to us, and perhaps to the early generations of redactors, that well-constructed, formally symmetrical texts, using such mnemonic devices as interlocking numbers to create memorable structures, are the key to preserving the Dhamma.

Buddhism & LaTeX

That’s ‘LaTeX’ the typesetting system, by the way, not the kinky shiny stuff. Sorry to disappoint.

One of the reasons for my prolonged absence from this blog the last couple of months is that I’ve been re-setting all my books using LaTeX. The only way I can get stuff like that done is by ignoring other stuff.

In this post I’d like to explain why Buddhists should use LaTeX. The basic reasons are simple. It’s free and open source (FOSS); and it’s the best.

Here’s the long version.

The Suttas often mention how beautiful it was to listen to the Buddha speak. He was supposed to have had a sweet voice like a cuckoo, clear, well-articulated, neither too loud nor too soft. It was not just the content of his speech, but the manner of it that was persuasive.

Just as spoken words can have qualities that are either attractive or repulsive, so too do written words. And just as the manner of presentation of spoken words substantially effects how they are received, so too with the written word. The craft of setting beautiful written text is called typography.

There is a long history of evolution of writing, which has gradually accumulated a ‘canon’ of rules that embody the best practices of presenting the written word. These style guides are not merely arbitrary, they arise from long cultural experience of what makes a text easy to read or not.

Similar principles have been followed in typography both East and West. See, for example, this inscription from King Ashoka, one of the earliest examples of Indian writing.

Without knowing anything about it, it is easily recognisable as writing. The lines are regular, with similar sized characters, and spaced so that the characters are clearly distinct, but not so far apart that they lose a sense of organic connection.

But as an early example, it is somewhat rough around the edges. The lines aren’t straight, and the characters not perfectly formed.

Here is a later, much more sophisticated piece of Buddhist typography, from 18th century Burma.

Here the typographic art has been raised to perfection. There is an unmistakable sense of care and precision about the piece; the harmony between the form and the content; the shapes of the letters, the integration of the decoration; the use of space; and so on.

This is what happens when care is taken to produce good typography. It’s not the expensive materials and decorations that matter, it’s the attention to detail.

And that is solely what typography is: attention to detail. Each little facet of the page is, in and of itself, trivial; but taken as a whole a well laid out page creates a harmonious form, which immediately, without fuss or pretension, informs the reader that they are reading something that is worth taking time over.

If you don’t get what good typography is, you won’t get the point of LaTeX. Typography leads a reader gracefully into the text, and invites them to care for the text’s meaning. Software is only a tool, which is there to help you accomplish this. So you have to know what you want. Learn what makes good typography first, then you’ll know what LaTeX is for.

In modern Buddhism, much of the typographical tradition is lost. This is, of course, a widespread outcome of the democratisation of text, so that anyone can now lay out text with a word processor. And it is exacerbated in Buddhism, where many of the books are produced by volunteers with no specialist training or interest in typesetting. And so virtually all the free distribution books in Buddhism (and many of the commercial ones) have bad typography. Mine included! (There are exceptions: the recent hard cover editions of Ajahn Chah’s teachings were produced in LaTeX, and they look lovely.)

To my mind, this is simply disrespecting the Dhamma. Do we think so little of the Dhamma that we are content to bash out any old crud as long as it fits on the page?

One of the basic issues underlying this is the use of word processors. Word processors are intended to do what their name says – process words. They are intended as tools for writing, and they can also be used for simple typesetting. If you want to do an invitation for your daughter’s 10th birthday, fine, use Word or LibreOffice. But if you want to make a book, use something designed for the job.

For a long time I used OpenOffice—later renamed LibreOffice—to lay out my books. Bit by bit, I tamed it and bent it to my will. And it did a pretty good job, once I learned how to do stuff like justify all text by hand using character styles.

What are you talking about? I hear you say. You just hit ‘justify’ and there the text is, all neatly lined up.

Ahh, no, it’s not quite like that. Word processors typically justify text by using a very simple algorithm. They see if the number of characters in a line fits the line (which of course rarely happens). If it doesn’t, they move the last word on to the next line and stretch out the remaining words to fit. The result is loose, sloppy text, with arbitrary gaps. Once you notice this, you can’t mistake it. It looks like crap.

This is how most Buddhist free distribution books are produced, using either (pirated) copies of Word, or LibreOffice, or sometimes a graphics program (since most of the volunteers who work on the books are graphic designers, not typesetters).

A proper typesetting system does much more. It evaluates what is the optimum way of distributing characters over an entire paragraph. Stretching here, compressing there, until the minimum ‘badness’ is achieved. And yes, ‘badness’ is a thing: it is the measure LaTeX uses to quantify how far from the ideal a justified line is.

Traditionally, all this stuff was done by professionals who hand set a page, using years of experience and study of their craft to produce beautiful, optimally laid-out pages. Look at an old book, and you’ll see a well-laid out page.

When computers started to be used for typesetting, the standards fell rapidly. So much so that one of the founding geniuses of computer science, Donald Knuth, took it upon himself to write the perfect typesetting program. It took six years, and he came up with TeX. This has has some modifications over the years, but is essentially perfected and stable.

Others took up the project and created LaTeX, which is essentially a programming language that uses a set of macros to make using TeX easier. Normally you don’t have to worry about TeX, just use LaTeX.

LaTeX is the standard for pretty much every scientific journal—it is particularly good at laying out equations—and is used widely in academia and elsewhere. There is, I have to admit, a geeky thrill in using LaTeX programs that written by NASA.

There is a good reason why such high-level institutions use LaTeX. It’s the best.

And there’s a good reason why most publishers use Abode’s desktop publishing software. LaTeX is, to put it mildly, unintuitive. That doesn’t mean it’s hard. I just means you have a learning curve before you can get it to do what you want.

Desktop publishing software can produce a good book, but it is not without its problems. The main FOSS program is Scribus, and while this is excellent for newsletters, magazines and the like, it lacks some basics for books, such as footnote support. Adobe software is good, but it’s expensive. And despite the premium price, in several areas it still produces output that is inferior to LaTeX. Of course the final result depends how you use the tool, but it is still true to say that LaTeX can produce output that is at least as good as Adobe.

A couple of months ago, I reached a crisis point with LibreOffice. There was so much formatting cruft in the file, it just wouldn’t work the way I wanted it to. Seriously, open up a years-old and oft-edited file one day and see what’s inside. It’s disgusting! (For a .odt file, rename the extension .zip, extract it, and open the .xml file with a text editor.)

Fortunately, I’m living with a serious geek, Ven Nandiya (he used to design physics engines for gaming). He learned LaTeX at Uni, and walked me through the basics.

Soon enough we were producing reasonably functional LaTeX documents, and I set out on redoing all my old books. I’ve done most of them, and am working on finishing touches now.

For those who want to try it out…

Okay, so hopefully some of you are interested to help create some high quality Buddhist books with LaTeX. But you have no particular computer expertise, and you don’t live with a geek god. LaTeX is very well documented, but most of the documentation is intended for serious users, and taking the first steps can be difficult. So here’s a few tips to get started.

First is obvious—do some googling and read the basics.

Second, install TexLive. This might already be present on your system. Essentially, TexLive is a distribution of software packages, documentation, and the like, that lets you produce LaTeX documents. It will be a large download, but it means that pretty much everything that you want is right there on your computer.

Third, play with creating some simple LaTeX documents. Here’s how. These instructions work for Linux, I’ve no idea if they apply on other systems.

You don’t need any special tools. I tried using LyX, which is supposed to make using LaTeX easier. But I found it frustrating and never got anywhere. I suspect it’s just that LaTeX is simply not suited to a WYSIWYG approach. There are various LaTeX editors available, but I find just using Gedit, a simple text editor (with code highlighting) and a terminal works fine.

The basic process is this. You edit the LaTeX file in a text editor. Then you use the terminal to process the file. That produces a pdf of the final product. Not so hard, right? Here’s the details.

Create a file on your desktop. Name it test.tex. Open the file with a text editor (notepad or the like) and paste in the following. Or better still, write it by hand so you get used to LaTeX.

\documentclass{book}

\usepackage{lipsum}

\begin{document}

\tableofcontents

\chapter{This is a chapter}

\lipsum

\lipsum

\section{This is a section}

\lipsum

\chapter{This is another chapter}

\lipsum

\lipsum

\end{document}

Save the document on your Desktop. Open a terminal. Write cd Desktop and press enter (capitals sensitive!). This opens the terminal in the Desktop directory. Write pdflatex test.tex and press enter.

Several files will automagically appear on your Desktop. Most of these are just logs and can be ignored. One of them will be a pdf file. But hold on: LaTeX counts page numbers on the first run through and creates a Table of Contents on the second run. So retype pdflatex test.tex in the terminal and press enter once more. (Hint: to cycle through previous commands in the terminal, use the up and down arrows on your keyboard.) Now open the pdf file. And Bingo! Your very own, LaTeX-produced pdf file.

Notice how many of the aspects of a book have been done automatically. Page numbers are there, a Table of Contents is there, and clearly formatted. The text is a reasonable size, well laid out, with plenty of white space.

So what’s going on here? Let’s have a look at what this code means.

\documentclass{book} % Commands in LaTeX start with a backslash. Every LaTeX document begins by specifying a document class (book, report, article, and so on). This tells LaTeX to use a set of defaults that are suitable. ‘Book’ is, obviously, the basic class for producing books, but the ‘memoir’ class is much more powerful and flexible and probably a better choice for serious booking. However, I only found this out after achieving most of what I wanted using the ‘book’ class, so I stuck with it.

\usepackage{lipsum} % ‘Package’ is the LaTeX (and more broadly Linux) word for a piece of software. You can specify any number of packages to use in a LaTeX document. These provide extra functionality and control. Think of them like plugins for your browser or apps for your phone. This one tells LaTeX to use dummy Lipsum text when the command \lipsum appears in the document. Tip: Packages and other LaTeX stuff (like document classes) come with excellent documentation, which you already have in your TexLive installation. To get the relevant documentation, just type in your terminal texdoc nameofpackage. No really, it’s that simple. Try texdoc lipsum and see for yourself. A nicely formatted pdf pops up explaining what the package is and how to use it.

\begin{document} % Should be obvious. Everything before this point is called the ‘preamble’, something like the <head> in an html page.

\tableofcontents % Inserts an automagically produced and formatted table of contents at this point.

\chapter{This is a chapter} %Inserts a numbered chapter heading.

\lipsum %Inserts dummy text.

\lipsum

\section{This is a section} % Inserts a numbered section heading.

\lipsum

\chapter{This is another chapter}

\lipsum

\lipsum

\end{document} %What it says. Note that everything on a line that follows a % on a line in LaTeX does not appear in the document. This is used for making comments in the file.

That’s a good start, but there’s much more to be done. What if we want to change something from the defaults? This is where things start to get tricky.

The defaults on LaTeX are brilliantly thought through (in the main), but they are obviously not suitable for everything. We will usually need to do things like change the page size, the margins of text on the page, and so on. And the default fonts are fine, but kinda stodgy (academia, you know), so we’ll want to liven things up a little.

So let’s just look at making a couple of basic modifications to the text we’ve produced so far.

You may have noticed that the pdf you produced has headers at the top of each page (except, intelligently, not the chapter title page). This is good, but the headers also appear on blank pages that have been inserted before the chapter title pages. This is because LaTeX by default follows the traditional practice of starting new chapters on a right hand page. You can change this behaviour in a number of ways. It is, of course, possible to keep the blank pages and eliminate the headers from them. But the simplest solution is to insert an ‘option’ in the specification of your document class that will eliminate the blank pages altogether.

Your original command was:

\documentclass{book}

Notice the two parts of this. The ‘command’, which is written with a backslash, and the ‘argument’, which is enclosed in curly brackets. Some commands also allow ‘options, which are inserted with square brackets between the command and the argument. In this case, we insert the option openany to instruct LaTeX to start a new chapter on any page.

\documentclass[openany]{book}

Run LaTeX again, and the annoying blank pages are gone.

For Buddhist texts, it is usually necessary to display unusual diacritical marks in Pali and Sanskrit, etc. Until recently, this was a hassle in LaTeX, as it was developed in the pre-Unicode days. As such it was limited to the fonts specially developed for LaTeX by its inventor, and using other fonts required special hacks.

There are a couple of recent extensions to LaTeX that can use any font installed on your computer very easily. One is XeTeX, but I had problems getting it to do microtypography. The other is LuaTeX, which is supposed to be unstable, but I have used it for lots of things with no problems. So, try this.

To your preamble (that’s the stuff between \documentclass and \begin{document}) add the following lines:

\usepackage{fontspec}
\setmainfont{Gentium Basic}

That tells LaTeX to use the package fontspec, allowing it to specify fonts, and chooses Gentium Basic as your main font. (You do have Gentium Basic on your computer, right? Don’t use plain old Gentium, as it doesn’t have proper bold face.) Then add some Pali or Sanskrit text after the \chapter, such as:

Upayo, bhikkhave, avimutto, anupayo vimutto. Rūpupayaṃ vā, bhikkhave, viññāṇaṃ tiṭṭhamānaṃ tiṭṭheyya, rūpārammaṇaṃ rūpappatiṭṭhaṃ nandūpasecanaṃ vuddhiṃ virūḷhiṃ vepullaṃ āpajjeyya. Vedanupayaṃ vā…pe… saññupayaṃ vā…pe… saṅkhārupayaṃ vā, bhikkhave, viññāṇaṃ tiṭṭhamānaṃ tiṭṭheyya, saṅkhārārammaṇaṃ saṅkhārappatiṭṭhaṃ nandūpasecanaṃ vuddhiṃ virūḷhiṃ vepullaṃ āpajjeyya.

Now run LaTeX. But rather than using the previous command pdflatex, which tells LaTeX to make a pdf file, we use lualatex, which tells LaTeX to use LuaTex to create a pdf file. Don’t worry, it just works (assuming, of course you have TexLive installed). Enter: lualatex test.tex.

And there you have it. Gentium Basic used throughout. And nicely formatted Pali text.

But wait! Oops, the Pali text doesn’t justify properly. One of the words hangs over the line. OMG LATEX IS SUCH CRAP!!! Okay, now’s the time to breathe. Calm down. It’s only software.

What’s happening is that, whereas word processors and the like use a simple system to do line breaks, LaTeX is much more finicky. Normally this means that it handles things like justification and hyphenation really well. But in unusual cases—like an unknown language—it doesn’t know what to do. So rather than bodging up a quick fix, it leaves the problem so that you’ll notice it and can fix it up properly.

The simplest solution is to add \-, which tells LaTeX to hyphenate at that point. So saṅkhārappatiṭṭhaṃ becomes saṅ\-khārappatiṭṭhaṃ. Run lualatex again. The problem is gone, and the text justifies nicely. *Sigh of relief*

That’s enough to get started.

O, and one other thing. Most of the time you won’t be writing directly as LaTeX, but will be adapting something from somewhere else. Remember, LaTeX is for typesetting, word processors are for writing. So you want to take your .odt or .docx or .html file and turn it into LaTeX. Heh, heh! Good luck with that.

The problem is that word processors end up with a complex mishmash of style and structure information, and there’s simply no way to anticipate how that should be changed into LaTeX. Not least because much of the information—like page format, colors, and so on—need to be changed anyway, and the program can’t guess what information you want to keep. Html files should work better, as they are a structured format not dissimilar to LaTeX.

There are various ways of converting word processing files to LaTeX. Here’s some. But after experimenting we found the best way was to use Abiword, a light weight word processor that natively exports to LaTeX. It captures most of what needs to be kept—footnotes, text formatting like italics, chapters—with a minimal amount of cruft (unlike LibreOffice’s LaTeX export…).

Open your .odt or .doc file with Abiword and export to LaTeX. Open up the .tex file, clean any cruft (I said it produces minimal cruft, not none at all…), and you should have the start of a basic LaTeX file.

One thing Abiword doesn’t do very well is export non-standard characters. It recognizes a few of the common characters with diacriticals in Pali/Sanskrit, but in most cases it just omits them. Indeed, it doesn’t just problematize obscure characters, but even some common punctuation. This is because Abiword exports to vanilla LaTeX, which is not Unicode. But we’re using LuaLaTeX so we want the Unicode characters, not the code.

When Abiword gives you the Unicode for the missing character it’s easy enough to insert the relevant character using the methods described here. However in the cases where the code is simply missing, not only is the resulting file inaccurate, but there’s no simple way to search for the missing characters.

Whether this is a hassle or not depends on what you are doing. If you’re quoting lengthy Pali passages, as for example a chanting book, then never mind, you can just paste the text again from source. But if you have Pali/Sanskrit terms or quotes embedded in your text (like me), then it’s a nightmare to find and replace them.

Luckily enough, Ven Nandiya came to the rescue. He’s hacked Abiword to export the Unicode codes of all non-standard Unicode characters. Awesome! So now the LaTeX file still doesn’t have the proper characters, but it does have the unique Unicode code for each character. In addition, each code is preceded by xx, so you can find them easily in your document.

Which is great, except you don’t have it. Ven Nandiya doesn’t want to submit his hack to Abiword for inclusion, as he says it’s real quick & dirty. But if you want it, let me know.


One of the best aspects of LaTeX is that it has an enthusiastic and very helpful online community. That’s the beauty of free and open source—it’s all about helping each other. That’s the real reason why the Buddhist community—made up largely of enthusiastic volunteers who just want to help doing something good—should feel a sense of kinship with LaTeX and other FOSS projects.

We shouldn’t spend lots of money on programs designed for massive corporations. Nor should we pirate software, which is unfortunately endemic in Buddhism.

Lurking at the core of these bad practices is the subtle lack of confidence that secretly thinks that free and voluntary stuff is not as good.

But Buddhism itself puts the lie to that. Everything is, or was until very recently, Free and Open Source.

And just as Buddhism has produced and maintained some of the world’s most amazing things without the need for commercialisation, so too the FOSS community show that the same spirit of community co-operation is not just alive, but produces the best results.

Don’t believe me? Which encyclopaedia have you used the most recently—Wikipedia, or Britannica? Why? Because Wikipedia is better, obviously. But that’s too subjective. What about something really hard and technical and objective, like, say, supercomputers. No room for Mickey Mouse thinking there. Well, the designers of the world’s fastest computers would appear to be fans of FOSS: almost all of them use a derivative of the Linux FOSS operating system.

This is not a trivial thing, it is an amazing one. The voluntary work of people coming together to help out is still, despite all our corporatization, one of the major forces in human culture. It’s on the cutting edge of technology, just as it has been for 2500 years of Buddhism.