31 January 2014

Dharma-niyama in the Vyākaraṇa-Mahābhāṣya

The term dhamma-niyama (Sanskrit Dharma-niyama) has taken on increased significance in the Triratna movement over the last few years since Saṃgharakṣita, through Dharmacārī Subhūti, informally published his more recent thoughts on the five niyamas (pañcavidha niyama) as an intellectual framework for understanding Buddhist practice and doctrine (See Revering and Relying Upon the Dharma July 2010).

In our discourse one now frequently hears reference to "the dhamma-niyama" in the place that used to be occupied by phrases drawn from German Idealism, such as "the Transcendental" and "the Absolute". Saṃgharakṣita himself has backed away from use of these phrases and suggests that his use of them was misunderstood. In which case it seems that in reifying the term dhamma-niyama (indicated by the use of the definite article) we have once again misunderstood him.

Just before Saṃgharakṣita published his new ideas on the niyamas, however, a side discussion was started up in the order by my friend and colleague Dharmacārī Dhīvan. Dhīvan circulated a long essay "Sangharakshita, the Five Niyamas and the Problem of Karma" (2009), which argued that Saṃgharakṣita's use of the term niyama was in fact an innovation and not based, as was claimed, on a traditional interpretation. He showed how the idea of the niyamas developed from the 5th century commentarial literature where it first occurred, through the interpretative lenses of Ledi Sayadaw and particularly C. A. F. Rhys Davids. The latter was an influence on Saṃgharakṣita in many ways. Dhīvan argued that though Saṃgharakṣita was largely drawing on Rhys Davids, his doctrinal innovation was both justified and necessary, and indeed successful, in responding to the concerns of his followers. The niyamas teaching is authentically Dharmic, just not traditional. One of the main differences of opinion was the meaning of the word niyama.
"However, according to my understanding of the Pāli language and the Theravādin commentarial tradition, the word niyama does not mean what Sangharakshita or Subhuti take it to mean, and Sangharakshita’s list of five niyamas is a creative re-interpretation of Mrs Rhys-Davids’ creative mis-interpretation of what the commentators say." Dhīvan 2013.
Saṃgharakṣita, following Sayadaw and/or Rhys Davids takes niyama to mean 'order of conditionality'. The set of five niyamas are said to outline five "orders of conditionality" i.e. five hierarchical domains in which conditionality operates. But niyama simply does not and cannot mean 'order', it means 'limit, restriction, inevitability'.

I joined this discussion in 2012 when I began circulating my samizdat translation of all of the Pali texts that mention niyama, particularly the previous untranslated commentarial literature on the pañcavidham niyamam or fivefold niyama. See also my blog: The Fivefold Niyama. In producing these translations it became clear that Dhīvan's comments regarding the relation to the tradition were spot on. I believe he and I are still the only two members of our Order to have read the source texts in Pāli (and only a handful of others would even be capable). My translations made most of the texts available in English for the first time, save one which was translated by Sayadaw. In The Niyama-Dipani: The Manual of Cosmic Order Sayadaw translates in such a way as to support his modern reinterpretation of the niyamas and often with no reference to the actual Pāli usage. Problems with Sayadaw's translation are dealt with in my translation notes and in Dhīvan's essay and article (see bibliography). Unfortunately Subhūti is uncritical of Sayadaw - referring to this work as "the translation of the Atthasālinī" (p.15). Far from being "the" translation, it is "a" translation and a highly idiosyncratic, not to say tendentious, translation.

While the word niyama (or niyāma, the two spellings are interchangeable in Pāli despite deriving differently) occurs in sutta texts it is not until the 5th century CE that Buddhaghosa takes up the word to produce the five categories of restriction on change. The translation that follows shows how the word dharma-niyama was used in Classical Sanskrit by the grammarian Patañjali commenting on Pāṇini's descriptive grammar, the Aṣṭādhyāyī. The Vyākaraṇa-Mahābhāṣya or Major Commentary on Grammar is usually dated to ca. 150 BCE, though this date is somewhat uncertain. In the passage concerned Patañjali is in fact commenting on some glosses on the Aṣṭādhyāyī found in the Vārttika by Kātyāyana. This whole section from the introduction to the Mahābhāṣya concerns the relationship between meaning (artha) and words (śabda).

This translation is really intended for my own use and ought to be treated with some suspicion, and at least read in conjunction with the published translation by Joshi and Roodbergen. My translation relies on the published translation and comments made during our class reading of the text. It must be emphasised that this is my translation and all errors and infelicities are due to my limitations.

  • Numbers in square brackets refer to the page numbers of the Sanskrit edition by Keilhorn (3rd ed. 1962).
  • Numbers in curly brackets refer to section and page numbers in the translation by Joshi & Roodbergen.
  • Passages in bold are Patñjali's citations from the Vārttika by Kātyāyana.

Vyākaraṇa-Mahābhāṣya, Paspaśāhnika (p.7-8)
[7] {80} But how is it known that the connection between a meaning and a word is established (siddha)?
From the world (lokataḥ) [i.e. from people in the world].
{81: 198} In the world, having acquired [in the mind] a thing meant (artha) the words (śabdān) are uttered. They make no effort in accomplishing this. However, an effort is required to accomplish a thing that needs to be made. For example: wanting to do a job with [or requiring] a pot, he goes to the house of a potter as says “Make a pot (kuru ghaṭaṃ), I require it for a job” . On the contrary one who will be using words doesn't go to the house of a grammarian and say [8] “Make words, I will utter them.” Right away having acquired the thing meant, he utters the words.
{82} If, then, the world is an authority (pramāṇa) in this [matter] what is the use of grammar (śāstra).
Where the use of a word is connected to the meaning from the world [as authority] grammar provides a restriction for the sake of religious merit (dharmaniyama).
{83: 200} Where the use of a word is connected to the meaning from the world [as authority] grammar provides a restriction for the sake of religious merit. What is dharma-niyama? It is a restriction for Dharma (dharmāya niyamaḥ); or, a restriction for the purpose of Dharma (dharmārthaḥ vā niyamaḥ); or a restriction aiming at Dharma (dharmaprayojanaḥ vā niyamaḥ)
Just as [in the case of] secular and Vedic [precepts].
{84: 202} The southerners have preference for taddhita compounds. So they say ‘laukikeṣu’ and ‘vaidikeṣu’ [in what is related to the world and what is related to the Vedas] instead of ‘loke’ and ‘vede’ [in the world and in the Vedas].
Or rather, the meaning of the taddhita is appropriate, i.e. just as the precepts (kṛtānta) found in secular and Vedic texts. So far as the world is concerned it is said “a domestic rooster is not to be eaten; a domestic pig is not to be eaten.” And that which is to be eaten is taken for the purpose of removing hunger. And one is also able to remove hunger by eating dog meat. In this case a restriction (niyama) is made: this is to be eaten; this is not to be eaten.
In the same way there is desire for women because of sexual arousal. And satisfaction of sexual arousal may be gained equally from available and unavailable [women]. In this case a restriction is made: she is available; she is unavailable.
{85: 207} Indeed in the Vedas also it is said “a Brahmin takes the vow (vrata) of milk (payo), a king the vow of gruel (yavāgū) and the merchant the vow of curds (āmikṣā)” And that “vow” is taken for the purpose of taking food (abhyavahāra). It is possible to take a rice (śāli) or meat (māṃsa) vow etc, as well. In this case a restriction is made.
Similarly it is said “The sacrificial post (yūpa) should be made of bilva or khādira wood. “Sacrificial post” is taken to mean what the [sacrificial] animal is tied to. And by this an animal might be tied to any bit of timber, erected or not erected. In this case a restriction is made.
Similarly the potsherds (kapālāni) are placed by the fire and the mantra is chanted “bhṛgūṇām aṅgirasām gharmasya tapasā tapyadhvam” [be heated by the heat of Bhṛgu and Aṅgirasa]. Even without the mantra, fire whose action is to burn would heat those potsherds. In this case a restriction is made: “done this way it leads to bliss (abhyudaya) [i.e. to heaven].”

{86:208} Thus here also the understanding of meaning may equally be expressed by correct words (śabda) and incorrect words (apaśabda) a restriction for the purposes of religious merit is made. “The meaning is only to be expressed by corrects words not by incorrect words. Done this way it leads to bliss.”


Now in this text it seems most likely, according to the commentaries ancient and modern, that dharma is being used in the sense of puṇya 'religious merit'. The idea that doing things in the way constrained by the injunctions or precepts (kṛtānta) will be a "causer of bliss" (abhyudayakārin) confirms this. Artha may have the sense of 'referent' (thing referred to by a word) or 'meaning' (the definition of a word) and it's not always clear if Patañjali makes this distinction.

The audience for this text lived their lives according to many religious and secular constraints. From the text we can see that some of them make sense on face value and some of them don't. Under most circumstances it is clear, for example, who is an available sexual partner and who isn't, even in our rather wanton society. In ancient India it was probably even more obvious since a person's spouse was the only sanctioned sexual partner (though that said prostitutes also plied their trade).

It might not be so obvious why a domestic pig was not appropriate food. The precepts allows for wild pigs to be eaten. And this is partly the point. A negative precept that says 'don't eat domestic pigs' is specific. We might be tempted to take the generally corollary that everything else is OK to eat. We might for example decide that dog meat was OK. But in India, as in the modern west, there was an unspoken understanding that dog meat was not for human consumption. There is no natural reason that this is so. Dog meat is consumed in some parts of the world and is presumably no more prone to disease or no less nourishing that any other kind of meat. But we just don't eat dog, and may even feel a sense of disgust at the thought. There is an implied restriction in the background to the specific restriction.

As mentioned above, one of the points of controversy in Dhīvan's initial essays on the niyamas was over the meaning of niyama. In this text there is no doubt that it simply means 'restriction'. One might eat anything, but there are various kinds of restrictions on what one may eat. One might have sex with anyone, but in practice one has a limited choice of partners. However the specific term dharma-niyama means a restriction for the sake of religious merit. That is to say that it is an injunction whose authority stems from the Vedas and is ultimately aimed at a good rebirth or at liberation through the correct performance of religious rituals. 

Fundamentally this argument is about restrictions on what is a correct word (plain śabda) and what is an incorrect word (apaśabda). Pragmatically Patañjali has to admit that many non-standard words are in common use. He is arguing that despite the many choices of words, that some are better than others. In particular he is arguing for what we call the Classical Sanskrit forms sanctioned by Pāṇini as correct and dialectical variations as incorrect. Here he points out even though we always have many choices of how to behave, that various kinds of restrictions apply: secular or worldly restrictions (laukikā) and religious restrictions (vaidkikā) found in the Vedic texts. So too words are restricted by secular and religious usage.

This is not so different to our time and place. Most people would use a different mode of speech when having fun with their friends than they might at a job interview. For English speakers in Britain the issue of local dialect words and expressions is a common one - and at present the mood seems to be going against allowing children to use dialect at school for fear that they won't be able to distinguish different contexts as adults and might use the wrong mode of speech. Or in other words that those with dialects that differ from standard (i.e. receive pronunciation, English as it is spoken in the South East) will be socially disadvantaged.

For the Buddhist who is interested in the idea of the niyamas the import is clear. Niyama means "constraint, restriction, limitation or inevitability". It is about restricted choices, vows made, and precepts imposed. In the Pali texts it refers to the restrictions on how change occurs. Thus is cannot mean a kind of "order" or "level" of conditionality, but only a constraint on how conditionality plays out. Things change, but not randomly. Plant a rice grain and it can only grow into a rice plant (and no other kind of plant). Perform an evil action and it must inevitably ripen as a painful vedanā.

In the case of dhamma-niyama it is used by the Buddhist tradition to explain the series of miraculous events that accompany the birth of a Buddha. There is a restriction on the universe related to the life history of a Buddha. As it says in the Sumaṅgalavilāsinī (DA 2.431)
"The shaking of the 10,000 world system when the bodhisatta enters his mother’s belly and other such phenomena [associated with the life story of the Buddha as told in the Mahāpadāna Sutta], this is called the inevitability of natures (dhamma-niyāma). Inevitability of natures is understood as consisting in this."
Such miracles as occur are bound to occur; they are what is required for the life story of a Buddha. I might well have translated dhamma-niyama here as "a restriction imposed by religion". In other words this is simply something that Buddhists believe, and, like the audience for Patañjali, they believe it because it is said in a sacred text. 



Dhīvan. Sangharakshita, the Five Niyamas and the Problem of Karma. 2009. (See also Dhīvan's website for some other related bits and pieces)
Dhīvan. 'The Five Niyāmas as Laws of Nature: an Assessment of Modern Western Interpretations of Theravāda Buddhist Doctrine.' Journal of Buddhist Ethics. Volume 19, 2012
Dhīvan. 'The ‘Five Niyamas’ and Natural Order.' [Blog Post] 5 June 2013. http://dhivanthomasjones.wordpress.com/2013/06/05/the-five-niyamas-and-natural-order/
Jayarava. Source Texts for the Five-fold Niyāma (pañcavidhaṃ niyāma). 2012
Joshi, S.D. & Roodbergen, J. (Ed. Tr) Patanjali's Vyakarana-Mahabhashya. Paspasha-Ahnika. Poona, 1986. Online: scribd.com
Kielhorn, F. (Ed) The Vyākarṇa-mahābhāṣya of Patañjali. 3rd Ed. rev. by K. V. Abhyankar. Vol 1. 1962.
Ledi Sayadaw. (1978). ‘ The Niyama-Dipani: The Manual of Cosmic Order,’ in The Manuals of Buddhism, trans. Barua, B.M, Rhys Davids, C.A.F., & Nyana. Bangkok: Mahamakut Press (orig. publ. 1965). Online: http://mahajana.net/texts/kopia_lokalna/MANUAL04.html [includes Sayadaw's correspondence with Rhys Davids showing how her interpretation is dependent on his]
Subhūti. Revering and Relying Upon the Dharma: Sangharakshita's approach to Right View. 1st Published in Shabda. July, 2010. Online: sangharakshita.org

24 January 2014

Origins of Myth: The Other Evidence

image credit: intrepidtravel.com
A few weeks ago I reviewed Michael Witzel's book Origins of the World's Mythologies. In that review I focussed, as Witzel does, on the evidence from comparative mythology. It's fairly obvious that if we share a grand narrative into which our myths fit, that there ought to be other evidence that follows a similar pattern. And Michael Witzel devotes a chapter of Origins exploring this evidence.

It must be said that none of the evidence is unequivocal and much of it is still rather ambiguous. More information is being added all the time. For example in the main text the book claims that there is no evidence of interbreeding between Homo sapiens sapiens and Homo Neanderthalensis. However between writing and publishing just such evidence was found indicating that all humans outside of Africa, New Guinea and Australia share a small number of genes with Neanderthals and Witzel acknowledges this in the forward. In the meantime further examples hybridisation have been discovered. (See Evolution: Trees and Braids)

To briefly recap, Michael Witzel sees a shared grand narrative in the mythologies of Europe, Asia, Oceania and the Americas (what Witzel calls "Laurasia") that is distinct from the grand narratives found in Sub-Saharan Africa, New Guinea and Australia ("Gondwana"). The Laurasian narrative involves creation of the universe from nothingness (void, chaos) via an egg or giant; the emergence of the earth from an abyssal ocean; the birth and lives of gods who fight amongst generations; the pushing apart of (father) sky and (mother) earth; the genesis and age of humanity, with an heroic age followed by our more mundane times; and finally the destruction of this universe, sometimes followed by the creation of a more perfect one. Local variations exist in abundance, but the overall story-arc seems to follow this broad outline. Gondwana mythology, by contrast, places no importance on creation.

Comparative Linguistics

The science of comparative linguistics is Witzel's home turf. He has, for example, studied the regional vations in Vedic Sanskrit and mapped the geographic areas that can be associated with various Vedic texts. He has also extensively studied loan words in Vedic showing that Munda may well have been the substrate language in Northwest India where the Vedic speakers first became firmly established. The early successes of comparative linguistics in the 18th and 19th centuries were impressive. It initially became clear, for example, that Greek, Latin and Sanskrit sprang from a common ancestor language. Germanic, Celtic, Iranian, and Slavic languages were soon added to the family, now most commonly called Indo-European. Systematic changes (such as /f/ for /p/ in Germanic as compared to Latin, part of Grimm's Law) across whole languages make it certain that they share a common ancestor and that language can be reverse-engineered on the basis of its surviving transformations. The reconstructed ancestor language is called Proto-Indo-European. The theory for example predicted three laryngeal sounds (related to our /h/) for PIE, which were not found in any living language, but were subsequently discovered in a written form of Hittite.

More recently the effort has been to try to create superfamilies by trying to locate systematic relationships across families or in reconstructed proto-languages. One result of which is a super-family called Nostratic (= "our language"). Nostratic includes Indo-European, Semitic, Sino-Tibetan, Austric (South-east Asia and Polynesia) and American Indian languages. The reconstruction is yet to find universal support amongst linguists. Witzel thinks this is in part due to artificial limits placed on the possibility of language reconstruction, but also due to the inherent difficulty of comparing so many languages at once. After all, few have the linguistic skill to do so. If we accept the proposed Nostratic language superfamily then we are immediately struck by the fact that its range is almost identical to the Laurasian mythology. For Witzel this is no coincidence, though he concedes that more work is required to establish Nostratic as a reality. But the tantalising conclusion is that the Laurasian mythology might have originally been framed in something like a Nostratic language. We are not there yet, but we have fine null hypothesis to try to disprove.


Where language provides tantalising hints at an underlying unity in the form of a common ancestor tongue, the field of genetics provides further insights into the relatedness and movements of peoples around the world. Two main subjects make up this evidence: phylogenies (or family trees) of mitochondrial DNA which is passed from mother to daughter; and phylogenies of Y chromosomes which are passed from father to son. Both types of DNA change only slowly, but at a rate we can estimate. However of all the evidence, the genetic evidence is most difficult to follow. The results of experiments are somewhat confused at times and the use of acronyms is intense.

In outline genetic studies show that all modern humans are related and that our ancestors lived in Sub-Saharan Africa. Anatomically modern humans emerged ca. 150 kya (1000's of years ago) plus or minus about 50 kya. More than one migration event seems to have taken place, but the one that succeeded in populating the earth seems to have happened about 65 kya. There are competing models for exactly how this was accomplished, but most include a small group of between 1000 and 10,000 travelling along the coastline eastwards. Sea levels were between 50m-150m lower, the figures cited vary wildly even within Origins, so evidence for this migration is mostly now covered by the ocean. But modern humans arrived in Australia (having crossed the open ocean) by about 45 kya for which we have good archaeological evidence. They continued North as well settling in China between 42-39 kya. Across Eurasia, modern humans encountered other species of hominids, but in every case survived, probably at the expense of the predecessors (and probably also interbred with them to some extent). The image below shows an up-to-date outline of the migrations based on mitochondrial DNA.

Migrations: approximate routes and times from Guha et. al (2013)

Theories on how the rest of Eurasia was settled are much less clear. There are two most likely scenarios. Firstly a second wave of migrants left Africa ca. 45 kya and went north into Western Asia and spread from there. Or secondly part of the first wave, perhaps based somewhere in West Asia, were the source of the expansion (this is what the image above shows). Eurasia being backfilled from China is also a possibility. In any case modern humans entered Europe 40-50 kya where they met, and to some extent interbred with, Neanderthals. From about 20-11 kya successive waves of migration occurred from Siberia into the Americas which were very quickly settled all the way to Tierra del Feugo. From about 5 kya inhabitants of Taiwan began the epic ocean voyages that peopled the islands of Polynesia, reaching New Zealand ca. 800 CE, but not before making contact (directly or indirectly) with South America or people from there. These dates are broadly supported by archaeological and anthropological evidence.

The dates for the settling of the Americas are important in dating the Laurasian mythology. Since the mythology is shared between all of the Americans and Eurasians the main outlines must have been in place before the first American migrations across the Beringia land bridge ca. 20 kya. This is long before any evidence of civilisation in the form of agriculture or large-scale permanent settlements. However, if Witzel is right about the implications of shared narratives then we have to accept that the narrative was in place by 20 kya at the latest.

Beringia Land-bridge from Balter (2013)

Recently a complete genome was sequenced for a child who died some 24 kya in Mal'ta, southern Siberia (Balter 2013). This boy is closely related to Amerindians, but also, surprisingly to populations in west of the Altai mountains. "Before 24,000 years ago, the ancestors of Native Americans and the ancestors of today's East Asians split into distinct groups. The Mal'ta child represents a population of Native American ancestors who moved into Siberia, probably from Europe or west Asia. Then, sometime after the Mal'ta boy died, this population mixed with East Asians. The new, admixed population eventually made its way to the Americas."

And just as with language studies the broad outlines of this evidence is consistent with Witzel's hypothesis. If a people, probably (initially) sharing a language or group of related languages, spread through Eurasia then we would expect to see evidence of relatedness in their genes. The best fit to Witzel's myth and language data involves two out-of-Africa migrations. The first, beginning ca. 65 kya, along the southern coastal route to Australia took with it the Gondwana mythology and certain mitochondrial genes. They left behind a string of languages with no connection to the languages of Laurasia. The second began around 45 kya and pushed first north and then both east and west populated Laurasia. These people spoke languages unrelated to the first migration, had a new, or at least different, mythology, and shared variants of mitochondrial genes not common amongst the first wave.


It is often commented on that although anatomically modern remains are found by about 150 kya, other features we associate with ourselves - burial, complex art, music - are first seen only about 40 kya. Witzel notes that more recent research indicates a slow build up to this so-called explosion of culture. But none-the-less there does seem to be a turning point. Most of the complex cave art begins around this time. The first evidence of musical instruments in the form of bone flutes are found. And burials with valuable items or indications of a belief in an afterlife also date to around this same period.

I've already cited the migrations to America as a latest date by which the Laurasian mythology can have been known in a more or less complete form. Since we know that the Gondwana mythology was unknown in Africa, New Guinea or Australia ca. 45 kya we have a upper limit for it's existence. It would seem then that the creation of the Laurasian mythology broadly coincides with the expansion of culture into Europe and Asia ca. 40 kya., but not later than 20 kya. 

Thus, with many caveats and hedges, we can draw out from the evidence a coherent picture in which ca. 40 kya a change took place amongst the ancestral Laurasian population that they subsequently spread, along with their genes and their language, across all of Eurasia, the Americas and the Pacific. Perhaps it is no coincidence that this story itself fits the Laurasian mythology, what we might call the Laurasian worldview. It is a peculiar feature of human beings that we are constantly seeking out new frontiers, while at the same time obsessing about our origins.

It's important to note that in the case of genetics and language we see evidence of hybridisation - shared genes across species on one hand, and loan words and regional language features on the other. The image we tend to have in our minds is a tree structure branching out from a singularity. This singularity almost certainly never happened. If you view railway lines going off to the horizon they appear to converge due to parallax error. I think we need to be aware of the historical equivalent of this. Just because we can find common factors underlying present complexity, does not mean that everything converges. History is complex at whatever magnification or scale we choose.



Balter, Michael. (2013) 'Ancient DNA Links Native Americans With Europe.' Science. 25 October 2013: Vol. 342 no. 6157 pp. 409-410. DOI: 10.1126/science.342.6157.409.
Guha P, Srivastava SK, Bhattacharjee S, Chaudhuri TK. (2013) Human migration, diversity and disease association: a convergent role of established and emerging DNAmarkers. Frontiers in Geneticsdoi: 10.3389/fgene.2013.00155. eCollection 2013.  Aug 9;4:155. 
Witzel, E. J. Michael. (2012) Origins of the World's Mythologies. Oxford University Press.

Note 27 Jul 2015

A recent large scale genetic study (scopeblog.stanford.edu) strongly suggests a single movement of people from the Siberia region into the Americans ca. 23,000 year bp.
The new genetic analysis suggests that the first immigrants to America left Siberia no more than 23,000 years ago, and then lived in isolation on the grassy plains of the Beringia land bridge for no more than 8,000 years. Those plains disappeared beneath rising seas 10,000 years ago. 
Once in the Americas, ancient Native Americans split into two major lineages about 13,000 years ago. One lineage populated both North and South America and one stayed in North America.

17 January 2014

Unresolvable Plurality in Buddhist Metaphysics?

image: Indiwall
In discussion over my forthcoming article describing changes in the metaphysics of karma, I raised the problem of the moral force of karma in the absence of personal continuity. An interlocutor responded that the tradition had resolved this problem, but I'm convinced this is not the case. I think this problem is decidedly unresolved. Most writing on the subject of karma assumes that a unified view is intended and can be discovered in the early Buddhist texts, but I see plurality that seems unresolvable.

One specific problem is this. The moral force of karma derives from the notion that we must live with the consequences of our actions and that even death is no barrier to the consequences being visited on us (or someone linked to us in a way we care about). It is fundamentally fear of negative consequences, particularly a bad rebirth, that pushes the unawakened Buddhist to be ethical; and the prospect of liberation from repeated death that pulls them along. This implies that the person who lives out the consequences of my present actions must in some way still be me. I must feel a sense of ownership over my actions and their consequences. In other words karma implies some kind of personal continuity or it doesn't make sense in human terms.

I'm well aware that this is specifically denied in texts such as the Milindapañha. According to tradition the one who experiences the results is not the same as the one who acted, but not different either. That person is dependently arisen (I'll come back to this). However if this was intended to make people behave according to Buddhist norms I can't help thinking that it's a rather poor attempt at motivating people. Theoretically the problem is solved, but practically it still disconnects the actor from both their actions and the consequences and thus can hardly motivate anyone to do anything.

Teachings on karma emphasise this implication by telling stories which explicitly link past and present lives. Such stories as the many hundreds of Jātakas (both in the two books of the Pali Jātakas and spread throughout the Nikāyas, Vinaya and other collections such as the Avadāna). In Theravāda countries the Jātakas are the main vehicle for teaching morality precisely because they emphasise living with the consequences of actions performed in part lives. However this is not to say that karma is only presented in these terms in Buddhist texts. Compare also such texts as SN 15.1 which describes saṃsāra in terms of ancestors stretching back through beginningless time; and SN 15.10 which by contrast describes one person (ekapuggala) wandering through saṃsāra leaving a mountainous pile of bones behind them. Karma is also said to be quite specific. My actions determine my rebirth, they do not determine your rebirth and vice versa. Similarly your actions do not give rise to my suffering except where they directly impact on me. But direct impact is unnecessary for karma generally since it is intention that determines outcome for the actor. 

Now contrast the metaphysics of paṭicca-samuppāda applied to the sense of selfhood. For the most part Buddhists seem to insist that, in reality, there is no self. There is a strong influence of the Two Truths teaching in such statements which use the language of existence or non-existence, i.e. the language of ontology. The Two Truths are a pervasive tool for dealing with the paradoxes  and contradictions thrown up in Buddhist ontology. However if there is no self, then there is no continuity over time and Buddhist ethics simply does not work. The language of ontology is carefully avoided in many early Buddhist texts that emphasise the application of paṭicca-samuppāda to experience only (the locus classicus being the Kaccānagotta Sutta SN 12.15) which is why I do not find the Two Truths teaching, with a foot in the camp of existence, useful (See Not Two Truths).

But even if we reject the language of ontology as belonging to the wrong domain (avisaya; cf. the Sabba Sutta SN 35.23) we are still left with a denial of personal continuity. The one who is reborn is not the same as the one who died, though not different either – they arise in dependence on causes. As Nāgasena says to king Milinda when asked about this problem: “It is not he, nor is it another” (na ca so, na ca añño. MP 41; c.f. S ii.18ff). The idea that vedanā arises because of oneself or another simply misunderstands how experience arises. Indeed the very question "who suffers?" is deemed unsuitable (no kallo) (SN ii.13). This is clear enough. But it's not clear how morality would work on this basis. If it is not me that suffers (or enjoys) the consequences of my actions then what is my motivation for practising virtue and avoiding vice? I don't believe that a morality based on such an abstract notion of responsibility is viable. And I would argue that the Buddhist tradition, in a tacit acknowledgement of this problem, does not teach morality in this way. Most Buddhists or whatever time and place teach some variation on "your actions have consequences for you and the people around you." Cf Buddhanet, The Budddhist Centre (1st sentence in both cases), SEP (§1 sentence 2). The theme recurs in many introductions to Buddhist ethics.

There is a fundamental disconnect between the metaphysics of karma and the metaphysics of paṭicca-samuppāda. I cannot see how to resolve these two while preserving the essential features of both. On the face of it this problem ought to have produced a crisis in Buddhist philosophy, though to the best of my knowledge it never has.

Why Do We Seek a Singularity?

Is it possible to step back from the content of this question and ponder why it is important to frame problem the way it is framed? In other words I want to ask if it is essential to attempt to reconcile the irreconcilable. I said above that writing on the subject of karma assumes that a unified view is intended and can be discovered in the early Buddhist texts. The central historical narrative of Buddhism is that it all springs from a single individual, the Buddha. The Buddha is presented as having first cut his ties to society (i.e. to all the conditioning of his early life from family, clan, class). Then over an extended period he pursues practices designed to break his identification with his body (practices which were historically associated with Jain ascetics). Then, completely cut off from his antecedents, the Buddha produced a new insight (prajñā) and founded a new lineage of instruction (anuśāsana). All Buddhist traditions take their teachings to be the direct or indirect words of the Buddha, if not in his historical manifestation then of the Dharmakāya.

Conventionally we expect that all the lines of the development Buddhist thought ought to converge at some point in the past. We expect the teachings to be unified and systematic. Many scholars of Buddhism declare that, from their point of view, unity implying a founder figure is clear in the texts. However even a relatively obtuse reader becomes aware that all is not unified. There are apparent discontinuities. The twelve nidānas are sometimes ten and sometimes eleven. And sometimes other numbers. I don't know how the tradition dealt with this, but in modern scholarship we have the handy model of evolution. If a teaching exists in various different forms then we can line them up chronologically (if only relatively) and argue that as the Buddha lived 80 years he must have refined his teaching as he went on and what we have preserved are various versions of the teaching from different periods of His "career"; or that it was developed by later disciples. Thus anomalies that might make us question the story of "unity" are used to support it using evolutionary models.

Lately I've been questioning the applicability of the the linear models that result from applying simple Darwinian ideas to the development of Buddhism (see Evolution: Trees and Braids). The tree structure, with its linear, binary diverging leads back in time to a singularity. A braid allows for divergence and convergence that is not unidirectional. Might it be that because the tree metaphor dominates our view of evolution or development that even when we find inconsistencies we still perceive them as springing from a single source? Such a reaction would also be consistent with my theory of how people with strong beliefs handle counter-factual information. 

The fact that we buy into the idea of an historical founder is also a cultural lens. It also predisposes us to see unity if we believe in a founder figure. But which comes first? Do we read widely with an open mind and discover a unity which demands that we accept the idea of a founder figure? Or does the idea of a founder figure cause us to read with confirmation bias and see only unity and ignore and explain away diversity? Is not Buddhism always presented as the teachings of the Buddha? And if our religious beliefs predispose us to believe very strongly in the historical reality? And if we identify him as someone who shares our religious, philosophical and cultural concerns as most modern Buddhists do?

If we extend the image of a braided river and look at the sources of the river we find that many tributaries contribute to the stream. It's not always clear which is the mainstream and which the tributary. If we follow one tributary to source it may resemble a singularity, but we must always keep in mind that there are many tributaries with equal claim. Indeed we can say that each source spring is reliant on a watershed, and that the hydrological cycle recycles water in complex ways. The metaphor is complex, dynamic, and undermined singularity thinking. And since the object is human culture, we require any metaphor to have these qualities.

In my writing about the Buddhist texts, over many years now, I have noted broadly Vedic influences, more specifically Brahmanical influences, Jain influences, animistic influences that I take to come from (one, many, or all of) the Austroasiatic, Dravidian and Tibeto-Burman speaking substrate populations inhabiting the foothills of the Himalayas (before Vedic speaking people even entered India). I've also noted some Iranian and/or Zoroastrian influences, some of which are certain and some speculative. Of course it is always possible that all these influences came together and were synthesised in the person of a founder. Possible, but not very likely. Cultures tend to be assimilated and synthesised by other cultures. And in our case this may well have extended both before and after the time of the putative founder. 

My theory about early Buddhism has two aspects in relation to this problem. Firstly the morality we associate with Buddhism is probably (broadly speaking) the mores of the Śākya tribe. In my writing on the Śākyans (published and unpublished) I have argued that the Buddha and his contemporaries should be seen as representing the culmination of a process of synthesis that began with a group of Iranian tribes migrating into India. They becoming naturalised and then were forced by climate change to migrate again and so ended up on the margins of the kingdoms of Kosala-Videha and Magadha. And, contra Bronkhorst, I argue that Kosala was the cultural centre of gravity of this time and place (it is the setting for the debates with Yājñavalkya in the Bṛhadāraṇyaka Upaniṣad for example). Importantly the pan-Indian rebirth eschatology met and was synthesised with the Zoroastrian idea of single destination—Heaven vs Hell—eschatology based on morality. The result was most obviously the various Śrāmaṇa faiths, but there was also considerable influence on Brahmanism outside the Āryavarta as well. As yet no plausible explanation has been put forward for why the Brahmins suddenly became dissatisfied with being reborn amongst their ancestors. However, if eastern, Kosala based, Brahmins were interacting with Zoroastrian influenced tribes, they might have been attracted to an afterlife in an eternal heaven and adapted the idea to their own uses. The earlier appearance of karma in Brahmanical texts simply reflects a predisposition to encapsulating religious ideas in texts that early Śrāmaṇa groups did not share until centuries later.

The Braid of Buddhist Metaphysics.

If there was an influence from Zoroastrian eschatology then it would emphasise the impersonal inevitability of post-mortem judgement, and would suggest post-mortem personal continuity. Personal continuity was already a feature of pan-Indian rebirth eschatology. Thus personal continuity ought to surface as an element in early Buddhist morality since it is present in the substrate belief systems on which Buddhism is built. And this is what we find in the Jātakas. Many aspects of the Jātaka literature, in particular some characters and moral themes, seem to cross sectarian boundaries and reflect shared culture. 

The other aspect of my account of early Buddhism, which is heavily reliant on Sue Hamilton's account of the khandhas, is that paṭicca-samuppāda was initially applied only to the process of having experiences. The basic description is vedanā (from √vid 'to know) arising from contact between sense object, sense organ and sense cognition, and the polarities involved in vedanā in turn giving rise to the mental processes by which we become infatuated and intoxicated with sense experience (papañca). My argument has long been that properly understood paṭicca-samuppāda only applies to this domain of experience. 

However, it is only natural that having discovered a principle like paṭicca-samuppāda that Buddhists would want to see what light it could shed on all aspects of their lives. And after all, as I have pointed out, the impermanence of the world and human life is quite universally acknowledged (Everything Changes, But So What?) My account of early Buddhism predicts that paṭicca-samuppāda applied outside the domain of experience ought to produce metaphysical problems. Thus the principle applied to the sense of self is a powerful lever that can shift our perspective on experience, but when we start to ask whether our self is real or unreal and try to answer in the same terms, we end up with nonsense. Asked if your self exists, later Buddhists are forced to answer both yes and no. And the yes/no answer is still being debated almost 2000 years after it was proposed by Nāgarjuna and his contemporaries (Not Two Truths). The metaphysics of the ontology of the Buddhist self are enormously complex and confusing. The result is some of the most convoluted discourses that end up degrading into insoluble paradoxes and are presented as representing the ineffability of the truth. It's notable that early Buddhist texts which apply paṭicca-samuppāda in the correct domain (visaya) never seem to resort to paradox or convolution in this way. Experience has three salient characteristics: it is impermanent, unsatisfactory, and lacking substance. The experience of selfhood is just another experience, has the same three characteristics, and is subject to the same limitations. Whether or not the self is real is completely irrelevant to the discussion of experience and how to manage it for our well-being. All other ontological questions seem also to be set aside as unanswerable and therefore irrelevant.

In this view the application of paṭicca-samuppāda to an ontological question such as personal continuity, particularly post-mortem, will be unlikely to produce entirely consistent answers. Experience is inherently discontinuous whereas the afterlife requires some kind of continuity. And this leaves us with an unresolvable problem in the area of morality. Considerable ingenuity must be employed to join the Buddhist morality to the principle of paṭicca-samuppāda. It is no coincidence that the question is one that troubles King Milinda. That the first answer never satisfied Buddhists can be deduced from the fact that it has been constantly revisited and reshaped. And even so when teaching morality it was and is common for Buddhists to embrace the idea of personal continuity, as in the Jātakas, for pedagogical reasons.


The kind of dependent continuity proposed by Buddhists as underlying Buddhist morality can hardly motivate people to practice virtue or avoid vice. Without the experience of actions having consequences for oneself, particularly in personal relationships, that Buddhist morality hardly makes sense. Without continuity between actor and the experience of consequence we have no motivation to alter our behaviour. As social animals were keenly attuned to the impact of our behaviour on others and theirs on us. If there were no personal continuity, or even if it were experienced as the kind of nominal continuity that Buddhist theorists propose, then we could not correlate actions and consequences in this life, let alone across lifetimes. Thus, ironically, it is the sense of our self as an entity continuous through time that underlies morality, at least in the unawakened.

Or we could see this as part of a pragmatic program. Morality is best taught with a strong grounding in personal continuity. Morality is concerned with our personal relations and without continuity the word "relations" is meaningless. Approaching morality in this way helps to eliminate major conflicts and prepare the mind for meditation. When it comes to reflecting on the nature of experience in religious exercises designed to liberate us from suffering, its best to point to discontinuity and to emphasise that the sense of self is no different from other experiences. To me this suggests the marriage of two very different activities and modes rather than a unified teaching. 

In fact what we see in the Pāli texts is a sea of partially integrated plurality, in crucial ways irreconcilable, and with a considerable amount of flotsam and jetsam from non-Buddhist systems of thought and practice. This isn't a problem if the Buddhist program is pragmatic rather than systematic. A lot of Buddhists are embracing pragmatism as an antidote to the idea of Buddhism as a systematic tradition. That said a surprising number of pragmatists are highly critical of, for example, teaching mindfulness to people who are anxious or in pain, precisely because it does not fulfil their criteria of Buddhism as a system. In the face of the plurality of doctrine, usually the best we can do is select a subset of the teachings that hang together and gloss over the discontinuities. A dense and complex jargon combined with an anti-intellectual discourse helps us to obfuscate such problems. Even those who study the texts more directly are doing so through cultural and historical lens that predispose them to see unity and continuity and to gloss over evidence of the opposite.


For a view on how the Buddhist tradition makes use of eternalism as an argument against nihilism and then mitigates eternalism with a specifically Buddhist argument. See:
Del Toso, Krishna. (2008) 'The Role of Puñña and Kusala in the Dialectic of the Twofold Right Vision and the Temporary Integration of Eternalism in the Path Towards Spiritual Emancipation According to the Pali Nikayas .' Esercizi Filosofici 3, 2008, pp. 32-58. Online academia.edu.
I have some reservations about this article. It is presented in terms of what Gotama taught, e.g. "Gotama makes a dialectical use of Eternalism as means to eliminate Nihilism." which I think is indefensible unless by "Gotama" the author means "the Buddhist tradition" and he does not seem to make this distinction. However the observations about the rhetorical uses of eternalism are very interesting.

Del Toso also refers to:
Hallisey, C. (1996) 'Ethical Particularism in Theravada Buddhism,' Journal of Buddhist Ethics, 3, pp. 32-43.
And in this article Charles Hallisey highlights the Mahāmaṅgala Sutta, which gives a list of desirable practices. Hallisey points out that the list of duties in the sutta and Buddhaghosa's exegesis argue "against any attempt to find a single metaethical principle that would make sense of everything on the list with an account of the occasion on which the canonical text was first taught." Hallisey's point is that although the audience for the text could agree on the auspiciousness of particular acts, they could not identify universal criteria of auspiciousness and Theravāda Buddhist ethics is thus particularist rather than generalist. 

Now see also Ethics and Nonself in relation to the Khandhas. (21.3.2014)

23 Mar 2015: See this short article by Thomas Metzinger on the value of the illusion of continuity for goals and rewards, which I take to be fundamental to morality also.
"As a philosopher, my conceptual point is that only if an organism simulates itself as being one and the same across time will it be able to represent reward events or the achievement of goals as a fulfillment of its own goals, as happening to the same entity. I like to call this the "Principle of Virtual Identity Formation": Many higher forms of intelligence and adaptive behavior, including risk management, moral cognition and cooperative social behavior, functionally presuppose a self-model that portrays the organism as a single entity that endures over time."

11 January 2014

RIP MIchael Dorfman

Michael Dorfman
There's a thread on Reddit saying that Michael Dorfman, a long time reader of and commenter on this blog, and a penpal of mine, died on Christmas Day 2013. He was 49, a year older than me. It seems he took his own life. 

As often happens with internet relationships I never really knew much about Michael personally, except that he knew a lot about Buddhism and thought a great deal more highly of Nāgārjuna (the subject of his MA thesis) than I did. He wrote to me quite often expressing his enthusiasm for my writing, but also not hesitating to criticise my mistakes and trying to make me think again. Lately we mainly did this behind the scenes through email rather than in comments. 

One of his last emails to me gives you a good idea of the kind of thing we talked about, and how generous he was with his praise and how keen he was to get things across to me (he was like this on the Reddit website as well):
I'm impressed you have the patience to argue with Sxxxx and Bxxxx -- they seem to be acting willfully obtuse, uncharacteristically so.  I would have thought it absolutely indubitable that (a) we have no direct evidence of the pre-Aśokan period, so anything we do is a reconstruction of some sort, and (b) the form this reconstruction takes is going to be governed in large part by what hermeneutical axioms we bring to bear on the indirect evidence. 
The Diamond Sutra piece is very much of interest-- I've been waiting for the library in Oslo to send me the volume of the Schøyen manuscripts with the Harrison piece, but they keep sending me other volumes instead, which is fascinating but a bit frustrating.  I think I've mentioned that the software experiments I am working on are aimed at Braarvig and his Bibliotheca Polyglotta. 
In any event, I'll try to write up a proper response to post on your blog-- I think your analysis is very interesting, and quite similar in its results to where I've been heading, albeit from another direction (i.e., not via philology in my case, but via philosophy.)  I think you are quite right that Nāgārjuna and the PPM literature are aimed at Abhidharmikas who have ontologized experience; unlike you, perhaps, I think that the notion of the "two truths" is a useful way of responding. 
If we look at Harrison's translation:

Subhūti said, “His personal presence would be substantial, Lord, it would be substantial, Blessed One. Why is that, Lord? The Realized One has described it as an absence. That is why it is called ‘a personal presence.’ For it is not a presence. That is why it is called ‘a personal presence.’”  
I'd highlight the "is called", i.e., dependent designation; it is conventionally 'a personal presence' because it is ultimately not a presence at all. 
But I think your problematization of the "because" here is very interesting, and something I'm going to have to think about.
 Michael Dorfman 15 Nov 2013
As one can see from this he is highly engaged in the discussion and we continued on at some length over the next few days exchanging thoughts on the Perfection of Wisdom tradition. The last time I heard from him was on 4 Dec 2012. 

He was also a good promoter of this blog, often posting my essays on Reddit.com under /r/Buddhism/ which typically generated a goodly amount of traffic from that site. Judging by the tributes there he was highly regarded by many Redditors. 

I always enjoyed getting emails from Michael and I will miss having his perspective on my writing. He understood the spirit of scholarship and had some sympathy with my approach, though did not always agree with either my methods or conclusions. In disagreement I found him a generous debating partner and able to argue the point quite vigorously without taking it personally. I felt we were kindred spirits and I'm sorry he's gone. 

Apart from a number of tributes on Reddit (link above) there is one other online memorial to Michael on The Endless Further. (if you know of others, please let me know). 


Michael's brilliant MA thesis published posthumously.

Dorfman, Michael. (2014) 'Putting the Madhyamaka Trick in Context: A Contextualist Reading of Huntington’s Interpretation of Madhyamaka.' Buddhist Studies Review, 31 (1): 91–124. [sub required]
In a series of works published over a period of twenty-five years, C.W. Huntington, Jr. has developed a provocative and radical reading of Madhyamaka (particularly Early Indian Madhyamaka) inspired by ‘the insights of post-Wittgensteinian pragmatism and deconstruction’ (1993, 9). This article examines the body of Huntington’s work through the filter of his seminal 2007 publication, ‘The Nature of the Mādhyamika Trick’, a polemic aimed at a quartet of other recent commentators on Madhyamaka (Robinson, Hayes, Tillemans and Garfield) who attempt ‘to read Nāgārjuna through the lens of modern symbolic logic’ (2007, 103), a project which is the ‘end result of a long and complex scholastic enterprise … [which] can be traced backwards from contemporary academic discourse to fifteenth century Tibet, and from there into India’ (2007, 111) and which Huntington sees as distorting the Madhyamaka project which was not aimed at ‘command[ing] assent to a set of rationally grounded doctrines, tenets, or true conclusions’ (2007, 129).

10 January 2014

Reasoning and Beliefs

In a desultory way I have been articulating a theory about religious belief over the last few years. As someone interested in factual accounts; as someone who's worldview has been changed by new facts on several occasions; and as someone who regularly spends a fair amount of time amongst credulous religious believers, I've been fascinated by the relationship between reason, factual information, and beliefs.

So for example, following Mercier and Sperber, I understand reasoning to be a function of groups seeking optimal solutions to problems. M & S argued that reasoning has confirmation bias as a feature when putting forward solutions to problems and that critical thinking, generally speaking, only works well when criticising someone else's proposed solution. The individual working in isolation to try to find the truth using reason is at a considerable disadvantage. Similarly, others have found that reason does not operate as predicted by mainstream accounts of it: "We’re assuming that people accept something or don’t accept it on a completely rational basis. Or, they’re part of a belief community that as a group accept or don’t accept. But the findings just made those simple answers untenable." (When it Comes to Accepting Evolution, Gut Feelings Trump Facts)

Furthermore I apply results published by Antonio and Hanna Damasio which suggest that emotions play a key role in decision making. (I outlined this idea in a blog called Facts and Feelings). Most real world problems are complex and making decisions about them, including deciding what we think is true, requires us to sift and weigh up a broad range of information. Before we can make a decision we have to assess the relevance of the information or category of information to the decision at hand, i.e. what is salient. Most of this process is unconscious and is based on emotional responses. Or in other words emotions function to help us decide what is important in any decision making situation. Decisions are then made by comparative weighing up of our emotional responses to the solutions we are aware of and have judged to be salient. Once a decision is made it is then rationalised to fit an existing personal narrative. This insight was also outlined at a library marketing seminar I attended almost 20 years ago.

The ability to unconsciously determine salience is what we often call our "gut feeling" or "intuition". This type of unconscious information processing seems to rely on pattern recognition and considering many options at once (parallel processing). The end result is decisions made with no conscious awareness of the process of thinking it through. Indeed the result often comes to us in a flash or after a period of sleep. The speed of this type of processing seems to contraindicate the usual cognitive processes of conscious problem solving, so that the answers that come via this route may not be fully integrated into the sense of self - spatially the answer comes from nowhere, or from outside us. Thus this kind of information process can coupled with views about metaphysical self that is not tied to the body and become "divine inspiration". (See Origin of the Idea of the Soul which relies on work by Thomas Metzinger).

The cognitive gap that opens up when we set aside information as non-salient is often filled by what neuroscientists call confabulation. Oliver Sachs poignantly described a man with no ability to make or retrieve memories (see Oliver Sach's Confabulating Butcher). Asked why he is present in the hospital or engaged in an activity he cannot say, but instead confabulates - he produces a plausible story and presents it as truth. There is no conscious lie, and the patient is not trying to deceive his interlocutors. He is presenting the most plausible account of himself that he has, despite being aware of inconsistencies, because he has no other account and the state of not being able to account for himself seems to be unacceptable at an unconscious level. Something similar happens whenever we have a flash of insight or intuition. The thought pops fully formed into our heads and then we confabulate a story about how it got there, and this generally speaking has nothing to do with how the mind or the brain works. Thus conscious thought is not a good paradigm for how the mind works. It is the just the tip of the iceberg.

Now this theory is still rather nascent and a bit vague. I'm still getting up to speed with the literature of evolutionary approaches to religion, though my views seem to have much in common with scholars like Ara Norenzayan. The theory does make an interesting prediction. It predicts that where people have strong existing views they will treat new contradictory information in a limited number of ways depending on how they feel about it. Where a view entails a major investment of identity and social status (e.g. a religious view) a person will tend to judge contradictory information as not salient and reason in such a way as to set aside the new information without having to consider the real implications of it. My idea was that this could be tested at some point on people with religious beliefs. On paper it does seem to account for some behaviours of religious people with respect to new information, for example the Christian fundamentalist confronting the facts of evolution.

With all this in mind I was fascinated to read an article by Steve Keen describing something very similar in the field of economics. Keen highlights a paper by Dan M. Kahan et al. "Motivated Numeracy and Enlightened Self-Government" Yale Law School, Public Law Working Paper No. 307. Keen, formerly Professor of Economics & Finance at the University of Western Sydney, is best known for his vehement polemics against the Neoclassical consensus in economics, epitomised in his book Debunking Economics. Neoclassical economics is what is taught to virtually all economics students at all levels across the world and has a monopoly over economics discourse that is disproportionate to its success as a body of theory.

Keen is one of a small number of economists who predicted the economic crisis that began in late 2007, and probably the only one who did so on the basis of mathematical modelling. One of his main criticisms of Neoclassical economists is that they ignore debt in their macro-economic models because aggregate debt cancels out: if I borrow £1 and a bank lends me £1 then the balance is zero. On the face of it this seems reasonable because our view of banks is that they lend out deposits. But in fact banks lend orders of magnitude more money than their actual deposits. When they lend they, in effect, create money at the same time. Problems occur when too much debt builds up and the repayments become a burden. For example private debt in the UK soared to 500% of GDP or five times the annual economic output of the whole country. Conditions may change and render debtors incapable of repaying the debt, which is what happened on a huge scale in 2007 and the sub-prime mortgage scandal. The levels of debt at that point meant that banks started to become insolvent as their income from interest payments plummeted and their own ability to service debts was compromised. From their the crisis spread like toppling dominoes.

Thus banks and debt are far from neutral in the economy. Perhaps in a post-crash world in which the role of banks in creating the crisis through a massive over-expansion of the money supply is public knowledge, the theory might be expected to change? But it has not. A global economic crisis has not caused any great soul searching amongst macro-economists who did not see it coming. Tweaking is the main result. 

Keen predicted the crash on the basis of the rate of change of debt. As we take on debt (private rather than public debt) growth ensues and , for example, employment levels grow. As shown in this graph there is a tight correlation (0.96) between changes in debt and the employment rate. 

Thus when the rate of change of debt began to fall sharply in 2006 it was a harbinger of collapse in the economy. In the USA employment levels fell from 95% to 90% and have yet to fully recover. 

This mathematical analysis ought to have been of interest to people trying to predict the behaviour of economies. Especially in post-crash hindsight it ought to be interesting to those whose job was to predict how the economy would perform and utterly failed to see the worst economic disaster in a century coming. As late as mid-2007, just months before sub-prime began to kick off, the OECD were still predicting strong economic performance in their member countries for the foreseeable future. However Keen's work, and the work of other economists who were successful in predicting a major recession/depression has been roundly ignored. Keen also argues that mainstream economic models are incapable of predicting a recession because it is not a possible state in those models, whereas his models do allow for recession. 

Not a man to mince words, Keen has been highly critical of the mainstream of economics. But now he puts that failure to react in the context of a theory of belief and decision making similar to the one outlined above. In the paper by Kahan et al, the participants were given tasks to assess their "ability to draw valid causal inferences from empirical data." The results were counter-intuitive and surprising. Numeracy – skill in understanding numbers – was a negative predictor of performance on these tasks if they conflicted with existing beliefs.
"It seems that when an issue is politically neutral, a higher level of numeracy does correlate with a higher capacity to interpret numerical data correctly. But when an issue is politically charged – or the numerical data challenges a numerate person’s sense of self – numeracy actually works against understanding the issue. The reason appears to be that numerate people employed their numeracy skills to evade the evidence, rather than to consider it." Steve Keen (emphasis in the original)
This is consistent with Mercier & Sperber's account of confirmation bias as a feature of reasoning. And it is consistent with my Damasio derived theory about the role of emotion in decision-making, if we read "politically charged" as signifying strongly held political beliefs, and associate that in turn with strong emotional responses to the issue. The authors tied this in with personal and social psychology:
Individuals, on this account, have a large stake – psychically as well as materially – in maintaining the status of, and their personal standing in, affinity groups whose members are bound by their commitment to shared moral understandings. If opposing positions on a policy-relevant fact – e.g., whether human activity is generating dangerous global warming – came to be seen as symbols of membership in and loyalty to competing groups of this kind, individuals can be expected to display a strong tendency to conform their understanding of whatever evidence they encounter to the position that prevails in theirsSteve Keen (emphasis added).
Kahan et al. extend the problem of salience of information to the social setting. Professed beliefs are often explicit markers of group membership and being well versed in group jargon and able to articulate group beliefs is part of what determines one's status in the group. In an economic setting, mainstream economists are able to ignore facts (such as a very high correlation of the rate of change of debt to the employment rate) that might change their worldview (particularly the way they view the role of debt in economics) because their status as members of a group requires them to conform to norms which are in part defined by holding a particular worldview. They are blind to facts which challenge their views. Keen points out that this is not a new observation and that some years ago that the great physicist Max Planck, who had struggled to have his work accepted by his peers, quipped that knowledge progresses "one funeral at a time".

This result reinforces the limitations of thinking of human beings in terms of individual psychology. It's a hard habit to break in the West. We are influenced by Freud, the Romantics, and the various revolutionary thinkers who championed the rights of individuals. Of course to some extent we are individuals, but not as much as we make out. Much of the inner life that appears to make up our individuality, is in fact determined by conditioning in various groups (family, peers, nation, religion, education) and by our place within these groups. We simply do not exist in isolation. 

Any philosophy of consciousness, mind, or morality which sees individuals as the main subject for study is of limited value. And the practice of trying to make valid inferences from the individual to the group is less likely to be accurate than the other way around. At the very least individuals exist in a series of overlapping gestalts with various groups. 

This view of people will most likely conflict with what we know. Most of us are convinced that we are individuals, who make our own decisions and think our own thoughts. We live in a society which highly values a narrative about "reason" and about what a reasoning individual is capable of. However, very few people are convinced by facts because that's not how reasoning works. That view of reason is isolated from other aspects of humanity such as emotions and our behaviour as social primates.

Those people who bombard us with facts fail to convince. By contrast the advertising profession has long understood that in order to change minds and behaviour one must change how people feel about the facts. Half the ads I see nowadays have almost no intellectual or factual content. It's all about brand recognition (familiarity) and a positive emotional response. One might say that for advertising the facts are now irrelevant. And so, liberals campaigning for protection of the environment, say, often fail to convince a sizeable proportion of the population or indeed anyone that disagrees with them to start with. Meanwhile advertising is a multi-billion pound industry, consumerism is rampant, and the environment is daily degraded in the direction of being unable to sustain human life.

Most people cannot be reasoned with, because neither people nor reason work the way they are popularly conceived to work. Those of us who want to make the world a better place must pay close attention to these issues of how people's minds actually work. If we want to convince people that we have a better solution it cannot be through facts alone. They must feel that what we say is salient in the context of their existing values. And even then, if what we say conflicts with strongly held beliefs then we can expect to be ignored. We tend to get so carried away in our enthusiasm for our own values that we fail to empathise with those whose minds we really need to change in order to change the world: i.e. political, military and business leaders. 


03 January 2014

If there is no self, who is responsible for what actions?

"I suppose a question I have is: if there is no 'self',
 how  can  we  be  responsible  for  our actions?"

A few days ago one of my colleagues asked this question in response to Dayāmati's essay What does one not have when one does not have a self? This is one of the crucial questions of Buddhist metaphysics. If there were no self, then there could be no moral agent or moral agency, and there would be no one who could be held responsible for actions. It would be an odd world. But modern Buddhists continue to argue that Buddhism involves not having a self on two levels: the self we currently have is an illusion; and once we are awakened we'll realise we have no self. 

This highlights a major discontinuity in the early Buddhist texts. I already have an essay ready and waiting which addresses another dimension of this problem, but I found the phrasing of this particular question interesting enough to dash off a reply. 

As far as the early Buddhist texts are concerned none of them ever comes out and says "you don't have a self". They seem to me to be saying something quite different and much more subtle. My starting point would be that early Buddhist texts focus on a particular domain (visaya) of knowledge. They say that all the knowledge we have is experiential: everything (sabbaṁ) is just the senses and the qualities of sense experience (solidity, resistance, volume, colour etc). Our perceptual world (loka) is the product of perceptual processes (khandhā). Sensations (vedanā) only arise when sense object and sense organ connect in the presence of sense consciousness. But we are intoxicated (pamāda) by the play of sense experience - and lost in the manifold interpretations of it (papañca). We mistake experience for something else. In Thomas Metzinger's terms we are "naive realists". We treat experience as if it is real; as though we are in direct unmediated contact with objects. Why we do this is another interesting question, but I'll take it as read for the purposes of this essay.

One of the important points made by early Buddhists was that dichotomies like "real" and "unreal" don't apply to experience - this is implied in the critique of the terms atthitā (existent) and nātthitā (non-existent) with respect to loka in the Kaccānagotta Sutta. Be it the immediate sensory experience or the cascades of impressions and associations that follow contact, the ontological status of experience is indeterminate. Experience is best described in process terms as arising when the conditions are in place and ceasing when they are not.

I'm not sure about other people but I don't think of myself as "having" a self, but in terms of "being" a self (I am). It's only in artificially abstract discussions that we begin to talk in terms of "having" a self. Most of the time the sense of being a self is transparent and unconscious. I have a locus of experience, a point of view (I look out through the windows of my senses), a sense of having a certain amount of control, a sense of  (and a desire for) continuity, and a sense of ownership over what I am aware of (they are my perceptions). These observations are the background against which I understand my world of experience. These qualities of consciousness are a bit like the "a priori judgements" in Kant's philosophy that structure experience. The language of "having" a self may well stem from the problem of the afterlife and continuity. From a Christian point of view we "have" a soul which survives our physical death. We can use the language of possession because a soul does not participate in the world of matter, it only provides continuity in the afterlife. I've explored this kind of duality of spirit and matter in some depth already so I won't go into it again here (See Metaphors and Materialism).

In the final analysis this sense of being a self is just an experience with the same characteristics as all experiences (anicca, dukkha & anāttan). We experience the world from a first person perspective but it is wrong to say "I am this experience" or "that experience is me" or "this experience is mine". Thus it would in fact seem to be wrong to say that we do or don't have a self. Or that we have a self that is then lost. The world of experience is what arises in our awareness when our body and mind interact with the world of objects (which is never in doubt in early Buddhist texts) including other people, who we can impute have a very similar experience of being to us through communication. Thus experience is neither subjective nor objective but exists in the overlap of the two - it is always both together. And this may be one reason the Buddhadharma is describes as a middle way.

The fundamental problem outlined by early Buddhist texts is intoxication with experience so that we mistake what is impermanent for something permanent and so on. Of course a great deal of ink is spilt on the consequences of this misidentification, but it all stems from being out of alignment (mithyā) with how experience really is. As we know ourselves through experience (since all knowledge stems from experience) then our self shares the characteristics of all experiences.

Though we cannot use the language of real or unreal with respect to experience, particularly the experience of being a self, "we" (the self we experience being) can still take responsibility for what this body, mouth and mind do. This is partly because experience has a physical locus (rūpa, first of the khandhās). Another aspect of having an experience is saṅkhāra or volition, in effect the urge to act on the stimulus. Volition in relation to experience is explained in terms of the mental & emotional responses (cetanā) associated with the different senses. Early Buddhism sees action (kamma) as resulting from responses to experience (cetanāhaṃ, bhikkhave, kammaṃ vadāmi AN 6.63). Motivation/intention/reaction with respect to experience does not require a sense of self. It is part and parcel of having an experience in this view. Since both saṅkhāra (the urge to act) and vedanā (the results of kamma) share a locus it can be observed that activities enacted from this locus of experience have consequences at this locus of experience. In other words as well as the observation that experiences arise in a conditioned fashion, we can in addition observe causal relations between actions and consequences. From a first person perspective this is interpreted as 'my actions have consequences for me'. And for most of us this is an important perspective. For if my actions did not have consequences for me, then surely I would be far less motivated to be ethical.

I've argued before that Buddhist morality is primarily focussed on interactions with other people. To speak of "actions" in the abstract can be misleading. By "action" I think we can say that early Buddhists meant our behaviour towards other people considered as other loci of experience. A Pāḷi verse which is repeated in several places suggests that our ability to attribute our kind of experience to other loci is the basis of morality. We understand that 'our' experience is much the same as 'their' experience, and thus we don't go about causing pain because we understand that pain is undesirable. "I" am responsible for actions initiated at this locus of experience if only because this is also where the consequences are experienced. 

The realisation of self qua contingent experience is liberating because it allows us to become sober (appamāda) with respect to sense experience. We are not caught up in grasping after experienced pleasure and averting unexperienced pain. We still experience both, but don't get hooked up on them. The Salla Sutta (The Discourse on Being Pierced) makes an important distinction between the unawakened and the awakened. The unawakened, struck by an arrow experience physical (kāyika) pain, but then they also experience a psychological (cetasika) reaction to the pain, as if they are pierced by a second arrow. The awakened experience the first arrow, but not the second. I wrote about this distinction in terms of pain, which we all have, and suffering which only the unawakened have. I imagine the experience of awakening to be characterised by contentment as we are not pulled this way and that by (habitual) reactions to stimulus. Our sensory world becomes less compelling and we can disbelieve it if we choose, or suspend disbelief and plunge in. (Gary Weber describes something like this). However as an embodied being we still have the same perceptual processes going on.

In terms of morality, post-awakening we are able to behave in ways that are consistent (Skt. samyañc) with how experience actually works instead of inconsistent (Skt. mithyā) and especially in relation to other people we don't behave selfishly because we see through our sense of being a self. Thus realising the contingent nature of self and not treating the self as a really existent entity allows for naturally skilful actions.

It's not that at some point we do "have" a self and then later we don't "have" a self. From an early Buddhist point of view we have experiences that we (unconsciously and transparently) interpret as us being a self. We proceed as though I, me, & mine are straightforward propositions. To the point where questioning I, me, & mine in any serious way is unusual and apt to draw blank stares from most people. When we learn to see those same experiences in the light of awakening then we no longer interpret them in the same way. The problem of agency in the light of not "having" a self doesn't arise because having or not having a self is not the case.

I can't speak to later developments. Buddhists seem to have very quickly got caught up in the pan-Indian conversation about ontology and the competition for influence and resources that accompanied being a large religious institution. They seem to have lost their way doctrinally. Once the Abhidharma is in place as canonical (around the beginning of the common era?), different solutions to the question were required and were supplied with varying success.

Related Posts with Thumbnails