Many years ago, Bhikkhu Sona penned an essay that outlined what he called the “mystery of the breath nimitta”. You’d better read it, or else none of this will make sense!
The breath nimitta was described in the 5th century meditation manual the Visuddhimagga as a subtle vision that appears to the meditator when their samādhi approaches absorption. This is a major influence on modern meditation teachers. This nimitta is, however, not mentioned in the Suttas. Sona argues that a close comparison between the various Theravadin meditation texts reveals that the idea of the visual breath nimitta essentially arose as a mistake, as the Visuddhimagga took similes from earlier texts and applied them as if they meant actual visions.
Now, this kind of essay is 100% catnip for me. It makes an actual argument, rather than just waffling vague pleasantries. And it does so based on genuine historical analysis, aiming to a conclusion that is genuinely useful for meditators. Awesome!
The great thing about an argument is that you can check if it’s right or wrong. It actually matters if the evidence adduced is correct or incorrect, and if the manner of argument is valid or invalid. Sona’s argument fails, I think, but it does so in interesting ways that make it worth taking a closer look.
Sona opens by saying “there is a significant puzzle to be solved”, and lays out his thesis thus:
a description of the mind of the jhanic meditator found in the Canon itself and quoted in the Paṭisambhidāmagga as a simile involving a comparison of mind with a full clear moon, degenerates to a mistaken literalization of these images as internally produced visual data. Since the contents of mind are not easy to point to, the Buddha frequently used similes comparing visual and other sense objects with mental contents in order for meditators to clearly understand what they should be seeking and experiencing. In religious traditions of all kinds we often find a naive tendency to take literally what is meant as a simile. It seems this process has occurred somewhere along the line and has become enshrined in the Visuddhimagga’s description of the paṭibhāganimitta or “counter-part sign.” It is important that new generations of western meditators not be misled by this probable historical error.
Let me first say that it’s not really persuasive to begin an argument by alluding vaguely to “a naive tendency” in “religious traditions of all kinds”. An argument is established by building up small, knowable things and moving towards generalizations. Coming off the blocks generalizing about how religions are “naive” tells us more about the author’s attitudes than anything else. If we think our sources are naive, we no longer think they have anything to teach us and will be inclined to a swift dismissal of any difficult issues.
But let us look at more specifics. Sona points out a “critical” distinction in the Vimuttimagga. One thing to note here is that Sona is relying on the old, draft translation by Ehara, Soma, and Kheminda. The shortcomings of this are well known, yet he nowhere acknowledges the critical inadequacy of his source. I’m checking his quotes against the now-standard new translation by Nyanatusita, and supply the old=new page numbers.
The quoted passage (68 = 250) says, “the sign of one meditation subject is to be grasped through the touched, namely, mindfulness of breathing.” According to Sona, this “shows that breath meditation is different from other concentration objects in that it is exclusively tactile”. He wants to argue that there is no visual nimitta for breath meditation. But that is not what the passage is talking about. It is describing the sensory basis of the meditation subject, not the nimitta. Some meditations, like fire or earth kasina, are based on looking at an object; others by feeling it. Breath is, of course, purely tactile as a meditation subject, whereas the “air kasina” can be both, since the movement of air can be both observed and felt.
A longer passage (158–9 = 420–3) details the process of breath meditation. Sona points out the rather curious fact that some of the very same descriptions of nimittas that the Vimuttimagga says to avoid are described in the Visuddhimagga as the valid appearance of the subtle sign. It is an interesting contrast, but does it bear the weight Sona places on it?
Now, during the process of meditation, it is common for various images and experiences to manifest, and a meditator should be clearly aware of this and ignore them. Mara appears in many forms. In particular, it is common for various perceptions to arise at the beginning of meditation which, to an inexperienced meditator, appear similar to how the subtle nimitta manifests as one approaches jhāna. Thus the Vimuttimagga, in describing such false visions, specifies “a meditator who from the beginning sees different signs”. A note by Nyanatusita explains that “different” could also mean “deviant”.
The Visuddhimagga, on the other hand, is not talking about a meditator “at the beginning” but, rather, one who is on the verge of jhāna. It too emphasizes that a meditator should remain focused on the breath until that point. The exact relationship between these two presentations is interesting, and would probably reward a closer look. But it is safe to say that the Visuddhimagga is no stranger to the idea that things might appear as a false path at certain stages of meditation which at other stages of meditation would be a sign of success. There is, in fact, an entire section on the ten “imperfections of insight” devoted to this.
These are by no means purely theoretical problems. It is common, if not normal, for meditators to encounter such byways. I regularly have meditators ask me this exact question: “I saw a light/felt a tingle/saw a mist/whatever”, should I focus on that or go back to my meditation?" And I always give the same advice: go back to the breath. The true nimitta arises as a powerful transformation manifesting within the breath, and the very fact that someone is doubting it shows that the hindrances are still present.
Sona further comments on this passage:
The sentence “this does not depend on colour or form” makes it quite clear that the meditator should not expect the sign of respiration mindfulness as a visual image, since it is not possible to conceive of a visual perception lacking both colour and form.
However, compare with the corrected translation in context:
Thus, observing the touch of the wind of the in-breaths and out-breaths at the [tip of the] nose or the [upper] lip gives rise to the sign of the wind, which is not caused by [attending to] shape or color.
The passage has nothing to do with whether the nimitta is visual or not. The point, which really just reiterates the earlier point about the breath being tactile, is that the breath has neither shape nor color, yet it gives rise to the nimitta.
Thus far, the passages don’t really tell us anything about the nimitta. They have spoken of either the sensation of the breath or of distractions. But Sona is quite right to point out that when the section on breath meditation does describe the nimitta, it does so in purely tactile terms:
The sign is like the pleasant touch of a tuft of silk, or a tuft of cotton wool touching the body, or it is like the pleasant touch of cool breeze touching the body.
How are we to read this? Sona has adduced a variety of passages that purport to show the description excludes a visual nimitta. As we have seen, the cited passages don’t prove his case. But this seems to be a strong piece of evidence, as it it directly describes the nimitta.
Now, in context, this passage on the breath nimitta appears well into the discussion of the different kinds of meditations. The basic description of the nimitta is found, as in the Visuddhimagga, in the discussion of the earth kasina (Nyanatusita 270). There the nimitta is defined as meaning “cause” (of samādhi), “perception”, and “reflection, like seeing the reflection of one’s own face.” The so-called “counterpart sign” is said to have the same meaning as the latter, i.e. “reflection”.
Taking this on face value (see what I did there!), the nimitta must be purely visual. But when we come to the section on breath meditation, we learn that the nimitta is tactile. I would suggest that it is possible to read these two passages in a harmonizing way: a nimitta is normally visual, but in breath meditation it may also be tactile.
The Visuddhimagga emphasizes the variability of the breath nimitta, saying that the “single meditation subject appears differently because of difference in perception”. Sona says that “the editors” of the Visuddhimagga “seem rather uncomfortable” with this diversity. I think he’s just reading into it. I’ve always really liked this passage, as it acknowledges the diversity of meditation experiences.
Sona says that “a great mystery is solved” when we realize that terms first used as metaphors were later taken to be literal.
One can only wonder how these metaphorical images, found at the end of the section describing breathing meditation in the Patisambhidamagga, eventually became literal visual events related to meditation practice in later commentarial works.
I would beg to differ. I see no mystery here, except the ordinary mystery of how the mind works. This is the job of perception, trafficking in metaphor and image. Has no-one ever dreamed? Do we not “see” things every night that make concrete images out of vague emotions or memories?
When your mind approaches samādhi, it becomes “bright”. You actually “see” a light, so that you might even open your eyes to check if a light had been turned on. This is not a naive idea from an old tome, it’s a common experience for meditators. But why do things “seem” like a light? Because it is perception, that’s how it works. The mind talks to itself using metaphors, which is why the metaphor of light is perceived by the mind as if it were actual light.
The metaphor of “light” is built into consciousness at a fundamental level: daytime we can see things, nighttime we can’t. Perception manifests as a “light” because that is how the mind communicates to itself that it is becoming clear, that knowing is breaking free of hindrances, that the cloud of ignorance is lifting.
If there are things in ancient meditation texts that describe metaphors that appear to the meditator as a literalized visual image, that’s because that is the job of a meditation book: to describe what happens when we meditate. And these are things that do in fact appear to meditators.
Sona says:
I am sure many a meditator has wondered why the Buddha had failed to mention the critical information of the “sign” and “counter-sign” in breath meditation
Indeed, but he has missed the actual answer to this problem. What is called the nimitta today is referred to under other terms in the Suttas, such as obhāsa, rūpa, and so on. The suttas are constantly talking about meditation in terms of light; indeed the very words jhāna and samādhi carry connotations of light in them.
Of course the commentary is more detailed, because, well, that is the job of a commentary. Not every detail and distinction can be traced back to the suttas. That doesn’t mean they’re wrong; clearly they had built up a body of doctrine over the centuries of meditation experience. Let us not forget that the Visuddhimagga is, by far, the most detailed, practical, and useful meditation manual produced anywhere in ancient times.
I think that some of the differences that Sona points out are interesting and worthy of further study. But if we want to understand the historical evolution of ideas, we have to start by not getting tripped up over terminology.