14 December 2018

Dependent Arising: Presence And Time

We can think of the following essay as a coda to my critique of the Perennial Philosophy since dependent arising is often presented as a singular universal metaphysical truth. In this essay I will begin by stipulating that paṭicca-samuppāda is a metaphysical doctrine and then proceed to draw out the implications of this premise.

The first task is to establish exactly what paṭicca-samuppāda says, using the standard methods of philology: analysing the grammar, syntax, and lexemes of the sentences. With a clear understanding of what the traditional formula says, we can try to understand what it means. I will show how the effect and condition relate under paṭicca-samuppāda. In addition, Buddhists were forced to accept a particular account of time and I will show why it had to be that account and no other. By the end of Part I, we will have a pretty good idea of how paṭicca-samuppāda performs as metaphysics.

If anyone thinks this is an elementary exercise and that we can hardly learn anything new about this most famous of all Buddhist doctrines at this late stage, let me assure them that in this case I learned something new or I wouldn't be writing about it. Most of what we learn about Buddhism in the present is only loosely correlated to the ancient texts and in this case there are major discrepancies.

In Part II, I will take the usual step and discuss paṭicca-samuppāda in terms of the nidānas (or bases) and what is often called the Spiral Path or upanisās. In particular, I will show, contrary to the received wisdom, that it is inconsistent with the nidānas, that the two describe very different kinds of conditionality. Unexpectedly, paṭicca-samuppāda turns out to be exactly consistent with the conditionality described in the upanisās. This is a major new observation.

Finally, in Part III, I will return to the issue of metaphysics and argue that paṭicca-samuppāda has nothing to do with metaphysics, but was employed as a description of subjective experience arising and passing away. Attempts to make it a metaphysical doctrine resulted in the kind of nonsense epitomised by Nāgārjuna's Mūlamadhyamaka-kārikā. Relieved of the necessity to make sense on the level of metaphysics, we are in a better position to see what the early Buddhists were getting at.

Let us begin at the beginning:

Dependent Arising

The classic formulation of paṭicca-samuppāda (Skt. pratītya-samutpāda) can be found in the four phrases found scattered through the Nikāyas:
imasmiṃ sati idaṃ hoti
imass' uppādā idaṃ uppajjati
imasmiṃ asati idaṃ na hoti
imassa nirodhā idaṃ nirujjhati
Since there is no regular pattern of metre we cannot think of these as verses. This is prose and, since each phrase has its own finite verb and there are no conjunctions, we can say that they are grammatical; four separate sentences, presented on separate lines to aid discussion. The formula occurs just a few times in Pāli: MN i.263, ii.32, iii.63; SN ii.28, 65, 70, 78, 79, 95, 96, v.388; AN v.184; Ud 1, 2.

The usual Sanskrit version is:
yaduta asmin satīdaṃ bhavaty asyotpadād idam utpadyate | yaduta asmin asatīdaṃ na bhavaty asya nirodhād idaṃ nirudhyate ||
Although in Sanskrit this often be abbreviated to just the first two sentences. And in Chinese the phrase is typically:


For reference, the verbs here are 有 "being", 生 "arising", 無 "non-being", and 滅 "ceasing". And Chinese does not have the rich grammar of the Indic languages so the structure is the same in each of the four phrases. It comes out sounding like something from the 道德經 Dàodé Jīng and this may not have been an accident since Daoism was a strong influence on Chinese Buddhism.

However, we will stick with a grammatical analysis of the Pāli. The two phrases imasmiṃ sati idaṃ hoti and imasmiṃ asati idaṃ na hoti use a locative absolute construction with present participles to indicate the relationship between the conditioning factor (paccaya) and the conditioned factor (vipāka). This kind of construction is used to indicate an action that is simultaneous with the main action of the verb. The main clause is idaṃ hoti "this is". The deictic pronoun idaṃ is used for an object present to the speaker and hoti is a dialectical variant on the verb bhavati (√bhū) "to be, become". The dynamic sense of "becoming" is probably better since it parallels uppajjati (ud√pad) "arising", though the difference in this case is minimal.

The "absolute" clause is imasmiṃ sati or imasmiṃ asati. The (irregularly formed) present participle sata is from the verb atthi (√as) and is in the locative case, while the pronoun is from the same base as idaṃ and also declined as locative. The meaning is: "this exists", but the locative absolute construction makes it "when this exists" or "while this exists"; and negatively "while this does not exist". Note that the deictic pronoun is used for both condition and effect; i.e., both are present to the speaker. However, if we translated literally it would be ambiguous, so most translators substitute this/that for this/this.

With this in mind I read these sentences as, “while the condition exists, the effect comes into existence” and “while the condition does not exist the effect doesn’t come into existence.” or more briefly: "This being, that becomes" and "This not being, that doesn't become".

The phrases imass’ uppādā idaṃ uppajjati and imassa nirodhā idaṃ nirujjhati combine an action noun (uppāda, nirodha) in the ablative of cause with a present indicative verb from the same root (ud√pad, ni√rudh). These mean: “from the arising of this [condition], that [effect] arises” and “from the cessation of this [condition], that [effect] ceases.”

So the sentences may be translated as:
This being, that becomes.
From the arising of this, that arises.
This not being, that doesn't become.
From the ceasing of this, that ceases.
While the condition exists, the effect exists.
From the arising of this condition, this effect arises.
While the condition does not exist the effect doesn’t exist.
From the cessation of this condition, this effect ceases.
This reading is supported by the influential Sri Lankan Buddhist writer, Kulatissa Jayatilleke, who expresses the relation as "Whenever A is present, B is present ..and... whenever A is absent, B is absent." (1963: 449). He notes that Buddhaghosa also saw it this way in the Visuddhimagga:
uppajjamāno ca saha samā ca uppajjati na ekekato no pi ahetuto ti sampanno (Vsm 521)
"Arisen (sampanno)" means arising together and arising equally, not one at a time and not for no reason.
There is not a lot of discussion about this, but my understanding is that the condition is both necessary and sufficient for the arising of the effect. The necessity part must be true, but the sufficiency is debatable. Conditionality might be underspecified and, indeed, in one way of talking about conditionality, multiple conditions are required to give rise to the effect (see Part III). We may ask if, in this standard case, the necessary condition could be present and not give rise to the effect? My reading of the formula is that this could not happen. Therefore the condition is sufficient. I deduced from Jayatilleke's translation and exposition that he also takes this to be the case.

To labour the point, the condition must be present for the entire duration of the effect, and as soon as it is not present, then the effect ceases. To put it another way, we could say that the effect and condition must coexist. This is one way to understand the world samuppāda, although more literally it means "co-arising".

Now that we know what the formula says and why, we can begin to explore the implications:

The Logic of Presence

The early Buddhist theory of conditionality says that an effect can only arise when the appropriate condition is present and that it must cease when the condition is absent; and thus we can say that the condition must be present for the entire duration of the effect. Note that Buddhaghosa himself has described arising as na ekekato "not one at a time" or "not from itself".

The doctrine of momentariness (focused on mental events) asserts that events can only happen one in each moment of time (the one citta rule). Under the conditions of momentariness, a condition can never coexist with its effect and therefore no effect can ever arise. Dependent arising simply does not work under these conditions. So the doctrine of momentariness fails, on its own terms, to explain karma (or anything). This puts the one citta rule in the spotlight, because this rule vitiates any attempt to link consequences to actions via dependent arising since it requires the two to always coexist (samuppāda).

There is a further profound consequence of the necessity for the coexistence implied by the paṭicca-samuppāda formula. Let us say that we have a number of events that (co)exist in conditioned relations as defined by dependent arising. We can call them a precondition, a condition, an effect, and an aftereffect (upanisā, paccaya, vipāka, and anuvipāka).

Each one is the basis (nidāna) for the arising of the next. If the precondition exists, then the condition arises. Once the condition arises, then the effect will arise, and the aftereffect will follow. And, of course, the system is not closed, but open-ended.

Let us say that the condition ceases, the effect immediately ceases, and thus the aftereffect also ceases. There is no backwards conditionality, so the cessation of the condition does not affect the precondition. This is good news for soteriology because if we can destroy the precondition then the whole edifice comes down. In Buddhism, ignorance (avijjā) is said to be the precondition for the whole miserable mess (kevala dukkhakkhanda), i.e., of rebirth, sentience, and suffering.

Another way to look at this is to begin with an event and trace back the conditions. Let us say that we observe the aftereffect and we analyse the conditions for that. We know that if the aftereffect exists, then the effect must exist at the same time. But if the effect exists, then the condition must exist, and if the condition exists then the precondition must exist. And so on. So if the aftereffect exists (i.e., if we perceive it) then all the preceding conditions must also exist at the same time.


→ → → → →
→ → → →
→ → →
→ →

In logic notation, ≡ stands for if and only if, thus the logic of this relation is:

aftereffect ≡ (effect ≡ (condition ≡ precondition))

We can generalise this as: For any system with N elements:

In order for anything to exist now, all the conditions for it must be in place stretching back in time. And at any point in the future, this must always be true. This was effectively the view of the Sarvāstivādins, although their process of inference was slightly different; they arrived at the same conclusion: in order to be consistent with paṭicca-samuppāda we are forced to conclude that everything exists all the time. See also my essay: Sarvāstivāda Approach to the Problem of Action at a Temporal Distance (02 May 2014).

There is a further problem here. This is a workable explanation of existence or becoming, but how could anything cease under these conditions? In order for something that is present to cease, the condition would have to ceased, and the condition for the condition, right back to infinity. But if all the conditions right back to infinity cease, then it would seem that all conditions whatever must cease. Thus if anything ceases than everything ceases. Though, of course, this is not what we experience, so it must be wrong.

One way around this would be to argue for a distinct thread of conditions for every phenomenon. However, in order for something new to arise, the conditions would have to stretch right back to infinity. In fact, if we follow the logic of paṭicca-samuppāda, nothing can come into existence and nothing can come out of existence. And since this is not the universe we observe, even on a superficial level, then something is wrong with our theory. Dependent arising does not describe the world at all. It cannot be thought of as a metaphysical theory. 

We also need to consider the implications of paṭicca-samuppāda for how Buddhists understood time:


The precondition cannot be the condition for itself. Any event that is the condition for itself can only be always existent or always nonexistent. If it presently exists and is the necessary and sufficient condition for its own existence, then the condition is present and it must continue to exist forever. If it presently does not exist, then the condition does not exist for it to come into existence and never will.

There are only a limited number of scenarios that can explain the situation:
  1. Time is linear and infinite towards the past. There is an infinite and constantly expanding stream of conditions which allow the present to exist.
  2. Time is linear and finite in the past. This would lead to a first condition which must always exist for anything at all to exist.
  3. Time is circular. This reduces to the case of an event being the cause for itself.
To clarify the problem with circular time. If a condition occurs in its own timeline, then it becomes a condition for itself. In the simplest case, two events A and B condition each other A ⇄ B. A conditions B, which conditions A. If we take an arbitrary slice of this stream and lay it out flat, we would see a series of conditional relations:

→ A → B A B

If we spell this out:
  • If A is present then B is present and then A is present...
  • If A is not present then B is not present and then A is not present ...
  • In other words, If A is present then A is present; if A is not present , then A is not present .
  • If A is not present, the only way for it to be present is if B is present, but B can only be present if A is present. Therefore A is never present.
  • If A is present, then the only way for it to cease is if B ceases. But the condition for B is A which is present, thus A is present. Therefore A is always present.
In logic notation, for any system A,B where the relation is dependent on presence:
A ⇒ B • B ⇒ A
¬A ⇒ ¬B • ¬B ⇒ ¬A
∴ A ⇒ A • ¬A ⇒ ¬A
which is equivalent to:
(A if A) and (¬A if ¬A)
If B stands for a relation such as (P ⇒ Q) then we can see that for any arbitrarily long chain of similar relations, circular conditionality with obligatory presence logically reduces to: (A if A) and (¬A if ¬A).

Traditionally, Buddhists opted for the idea that time was linear and infinite towards the past, but they combined this with epicycles of evolution (samvaṭṭati) and devolution (vivaṭṭati). Strictly speaking, it would not matter if the universe were spatially finite, or had a finite future, but under dependent arising time must be infinite in the past to avoid an eternally existing first condition.

Summary of Part I

The paṭicca-samuppāda formula describes a dynamic in which the condition must be present for the effect to arise and the effect ceases when the condition ceases. It says that the condition must be present for the entire duration of the effect. It has always said this, so if you learned something different then, I'm sorry, but your teacher misled you. Historically, only the Sarvāstivāda understanding of conditionality is consistent with paṭicca-samuppāda.

The requirement for presence means that the condition must be present for the effect to arise; and it means that the condition for the condition must also be present. And so on back through time. To avoid an eternal initial cause, Buddhists have to adopt a worldview in which time is infinite in the past. To avoid having conditionality collapse into something being a condition for itself, time must be linear, although within this linear time, Buddhists accepted the Vedic myth of epicycles of evolution and devolution. However, having explained presence this way, we struggle to explain ceasing.

I am sticking to the internal logic of paṭicca-samuppāda in this essay, but I cannot help but point out that early Buddhists were wrong on two counts: time is continuous rather than discrete; and time is finite in the past. As far as I can see paṭicca-samuppāda explains nothing on its own terms and it explains nothing on modern terms.

This concludes Part I. Part II will look at the relationship of paṭicca-samuppada to the concepts of nidāna and upanisā.



Jayatilleke, K. N. (1963). Early Buddhist Theory of Knowledge. Motilal Banarsidass, 2010.
Related Posts with Thumbnails