Negative Criteria for Moral Decision Making in The Kālāma Sutta

IN THE KĀLĀMA SUTTA the Buddha provides a list of negative criteria for making moral decisions. These are quite interesting, but I don't think any of the mainstream translations really captures what's going on. We all seem to agree that the criteria form sets, but the translations offered don't seem to hang together as sets. What's more, some of the translated criteria seem counter-intuitive or confusing. This may be because the terms are vague, or in some way unusual. What follows is my attempt to combine etymology with historical and textual context to tease out more connected meanings and translations of the terms used so they makes sense on their own, and also form natural sets in English.

The general formula is mā X-ena - i.e., the prohibitive particle mā 'don't' with a word in the instrumental case, and no verb. The sentence then means 'don't use X to do something', and we are left to discover what the something is from the context. Clearly, the context shows that the something is making a decision about morality, about how to behave. I'll just mention that Buddhaghosa's commentary supplies the verb gaṇhittha which I take to be the second person aorist of gaṇhati ' to grasp, seize, take hold' - the combination of mā + a verb in the 2nd person aorist forms a strong prohibition. Though this raises the further question of 'don't seize what with X?'

The same set of criteria are used in the commentaries (DA iii.879 & SA ii.308) as part of the explanation of the phrase ekaṃso gahito (SN 47.12, and elsewhere). In SN 47.12 the Buddha is describing Sāriputta's declaration of trust (pasanna) that the Buddha is more knowledgeable (bhiyyobhiññataro) than anyone -- past, present or future -- on the subject of sambodhi. He says that Sāriputta's statement has ekaṃsa gahito 'grasped certainty ' (gahito is the past-participle of gaṇhati). Bhikkhu Bodhi translates this phrase as an adjective -- 'definitive' -- of the next phrase: "a definitive, categorical lions roar". (Connected Discourses, p.1641) Here the Buddha accepts Sāriputta's inferred knowledge of the Buddha's attainment as trustworthy which I'll come back when discussing nayahetu below.

Coming back to the Kālāma Sutta. The list of criteria in Pāli, with my translation, is:

mā anussavena, mā paramparāya, mā itikirāya, mā piṭakasampadānena, mā takkahetu, mā nayahetu, mā ākāraparivitakkena, mā diṭṭhinijjhānakkhantiyā, mā bhabbarūpatāya, mā samaṇo no garūti. (A i.189)

Don’t use revelation, don’t use lineage, don’t use  quotation and story, don’t use tradition; don’t use pure reason, or inference, or the study of signs, or speculation, don't just accept what seems likely; don’t use respect for a toiler.

We begin with anussava which comes from √śru 'to hear'. In this kind of context what is heard is religious teachings. The Vedas, for instance, are known as śruti 'the result of hearing' or more aptly 'revealed'. The suffix anu- means 'after, along, along with'. This and the next three terms can all be translated as 'tradition', PED has "hearsay, report, tradition", but each of our terms brings out a different aspect of tradition. Buddhaghosa merely supplies the gloss anussavakatha 'talk of tradition'. My sense of this word is that it reflects the origins of tradition in revealed truths, and that it not only forms a set with the next three items, but that they form a sequence.

Next, we have parampara: literally this means 'another and another', i.e., 'one after another', or 'a succession'. We might translate it as 'lineage.' This refers to the passing on of revealed truths from teacher to student generation after generation. In the parallel list in his commentary on SN 47.12 Buddhaghosa substitutes ācariya-paramparāya, i.e., a lineage of teachers. This kind of succession of teaching receives a sharp criticism in the Tevijjā Sutta (DN 13), where the Buddha suggests that what is passed on is only empty words (appāṭihīra-kathaṃ). He doesn't accept the original revelation because it is not based on personal experience (sakkhidiṭṭhi). Buddhaghosa, again, merely glosses paramparakatha 'talk of succession'.

Then comes itikirā, which is a (dvandva) compound of two words used to indicate quoted speech: iti & kirā. The two words together as a compound occur only infrequently (here, A ii.190, and in the commentaries on these texts). PED says that kirā is used in continuous story, whereas iti is used in direct or indirect speech. PED suggests 'hearsay', but I think the context makes this more specific and suggests to me the practice of quoting from spiritual teachers and telling spiritual stories. The contrast is, again, with personal experience. A more speculative translation might be 'aphorisms & parables'. This kind of thing is a step further removed from a revealed truth than the lineage of teachings - it is the teachings becoming popular culture.

The last term in this set is piṭaka-sampadāna: 'handing on of collections'. We call the three main sections of the Buddhist Canon -- sutta, vinaya, abhidhamma -- the tipiṭaka 'three collections'. The etymology is not clear, but it apparently means 'basket' with an agricultural application - e.g. vīhipiṭaka, 'a basket for rice'; kuddāla-piṭaka, 'hoe & basket'. It's not clear when this term came to mean 'a collection of writings' - the usage seems to me to be Buddhist, so for example, the Vedas use different terminology for collections of texts. Would the metaphor predate writing; i.e., predate the need for a physical container to place physical texts in? Or might it refer to the mind of the expert who memorised the texts before they were written? Buddhaghosa is no great help: ...piṭaka-tantiyā saddhim sametīti 'collections of sacred teachings (tanta = Sanskrit tantra) together with associations (PED sameti 'to come together, assemble' with a connotation of 'what is learnt'). I suggest that what is collected are the quotations and stories (itikirā) just mentioned. By the way, the word anthology has a similarly rustic origin: it comes from the Latin for 'a collection (logia) of flowers (anthos)'. The anthology is the museum of religious teachings: frozen in time, and devoid of the living context of revelation or personal communication of that revelation.

So the sequence is: 'revelation, lineage, aphorisms & parables, and anthologies'. Each step is further from the source of wisdom, but even revelation is not necessarily connected with personal experience and is therefore not a reliable guide to how to behave.

Having dismissed tradition in its various forms, the Buddha then moves on to deal with intellectual criteria. Firstly, takkahetu. PED gives 'ground for doubt, or reasoning'. Hetu, of course, is 'cause, reason, condition'; and takka is literally 'twist, turn' and metaphorically 'to turn something over in your mind, to think about'. For the Sanskrit tarka MW suggests 'reasoning, speculation, inquiry' or 'logic'. Buddhaghosa glosses: takkaggāhenapi mā gaṇhittha 'also don't grasp by seizing of reasoning'. The question then is: what kind of compound is this? So we may see this as a karmadhāraya: 'logically-caused'; or a bahuvrīhi: 'whose cause is logic'. The sentence seems to be saying 'don't reason'; but in light of what comes after we have to take this as referring to hypothetical reasoning, to what Kant called "pure reason", i.e., reasoning disconnected from experience, and especially from emotions and values.

The next term is similar in form: nayahetu. Naya is from √nī 'to lead' and means 'method, plan, inference; sense; behaviour, conduct'. 'Inference' fits the context nicely as a counterpart of logic. However, in SN 47.12 Sāriputta understands according to the Dhamma (api ca me dhammanvayo vidito) where dhammanvaya = dhamma 'nature, truth, the teaching?' + anvaya (anu- + √i) 'conformity, accordance; according to'. One of Buddhaghosa's glosses of this passage suggests that such understanding is anumānañāṇaṃ 'knowledge from inference' (where anumāna is a synonym of naya). So inference per se is not a bad thing, as long as it is based on dhamma.

Next we have ākāraparivittaka which is a bit more complex. Ākāra is from ā + √kṛ and means 'a way of making; a state or condition; a property, sign; a mode'; while parivitakka derives from takka with prefixes pari- and vi- and means 'thought, reflection' or 'meditation' (in the English sense). PED suggests 'study of conditions, careful consideration, examination of reasons' but these seem to be perfectly good ways of approaching moral decisions, and in keeping with the general trend of Buddhist approaches. Nyanaponika & Bodhi translate it "reflection on reasons" (Numerical Discourses, p.65). I'm not satisfied with this, especially in light of the Vīmaṃsaka Sutta (MN 47) where one places faith in the Buddha as teacher (sarathi pasīdaṃ) for the reason (ākāra) of having heard the dhamma. Turning to Buddhaghosa we get:

'sundaramidaṃ kāraṇan'ti evaṃ kāraṇaparivitakkenapi mā gaṇhittha

One should also not grasp thinking about obligation as 'this obligation is beautiful'.

So Buddhaghosa appears to relate ākāra with kāraṇa - both from the same root. I'm not convinced that this fits the context either. My feeling is that it might be a reference to seeking knowledge through interpreting (parivitakka) signs (ākāra), i.e., divination and reading omens. The extensive list of divination practices banned in the Dīgha Nikāya make it seem likely that many such practices were in use and popular. This is placed amongst intellectual approaches to making moral decisions because it provides a rationale for behaviour which is not related to experience, but not tied up with larger religious revelations, and therefore continues the theme.

After this comes diṭṭhinijjhānakhantiya; PED 'delighting in speculation'. It's a triple compound with diṭṭhi 'views' nijjhāna 'understanding, insight; favour, indulgence' and khanti 'patience, forebearance'. Nyanaponika & Bodhi suggest "accepting a view after pondering". Clearly khanti here suggests passivity so 'accepting' fits quite well. I can just about see how the compound could mean a view accepted after pondering, presuming that nijjhāna can mean 'pondering'. I wonder if it would be more straightforward to read it as saying 'accepting & indulging a view'; i.e., uncritically accepting an ideologically based understanding (reading nijjhāna-khanti as a dvandva; and then as a tatpuruṣa with diṭṭhi). The compound allows for this, I think, and it makes more sense to me. It also fits the context of decision making based on something other than personal experience.

Next comes bhabbarūpatā. PED has no suggestion for this compound though bhabba (a gerund from √bhū) means 'able, capable; fit for'. Rūpatā is from rūpa 'form' and means 'appearance'. Nyanaponika & Bodhi, apparently following Buddhaghosa, attribute the fitness (bhabba) to the 'speaker' (or bhikkhu in AA). I understand the quality of 'fitness' to relate to the idea, however. I think it means something that 'seems likely'. That is to say, intellectual laziness. Something seems likely so we just accept it uncritically. It would, therefore, relate to how I understood the previous four terms, and I place it with that set. Nyanaponika & Bodhi, however, takethe plausibility to be a quality of the speaker rather than the idea and so place this criteria with the next one. There is a symmetry to this - four criteria relating to tradition; four to intellect; and two relating to teachers.

The last criteria is samaṇo no garu 'the toiler is respected by us'. Here it seems that the samaṇa (literally 'one who toils' from √śram 'to toil') is the one mentioned at the beginning of the sutta who outlined a doctrine, but lashed out at other doctrines. The most common translation seems to be 'teacher'. Samana is one of the words that are difficult to translate. We find: ascetic, contemplative, recluse, etc., but none of these are accurate and all of them carry a heavy burden of connotation. My coining, 'toiler', is more literal and in this context has less baggage. Garu means 'weight', and by association 'respect'. A guru (from the same root) is someone with gravitas.

It may be that in aiming to fit these terms into sets I have done some violence to the original text. I hope not. I prefer to be creative in finding a translation that makes sense, rather than sticking rigidly to the dictionary definitions and producing something which is not coherent. All of my translations can be justified on etymological grounds, however. I did not pluck them out of the air. I used the dictionary as a starting point, and read each one in the context of its neighbours, as well as other texts which use the terms. In each case the aim was to produce a decision-making criterion divorced from experience, but one which makes sense on every level. For instance, it doesn't make sense to admonish people not to use their reason to think about their behaviour, but it might make sense to tell people to reason in the light of experience.