Page #1
--------------------------------------------------------------------------
________________
PATANJALI
commonly claimed that these distinctive styles of argument, recently uncovered, served to dissolve many traditional philosophical problems. The enthusiasm with which philosophers embrace a new style of argumentation is not always matched by their readiness to examine its credentials, and this book played an important, if sobering, role. See also: AUSTRALIA, PHILOSOPHY IN
List of works Passmore, J.A. (1951) Ralph Cudworth, Cambridge:
Cambridge University Press. - (1952) Hume's Intentions, Cambridge: Cam
bridge University Press. --(1957) A Hundred Years of Philosophy, London: Duckworth.
- (1961) Philosophical Reasoning, London: Duckworth.
(1970) The Perfectibility of Man, London: Duckworth.
(1974) Man's Responsibility for Nature, London: Duckworth.
- (1985) Recent Philosophers, London: Duckworth. - (1991) Serious Art, London: Duckworth.
1 Patanjali and philosophy Patanjali is the author of the Mahābhāsya (Great Commentary), which comments on Panini's famous grammar called Astādhyāyi (c. fourth century BC) and on Katyāyana's Vārtikas; the latter have
the latter have only survived as part of the Mahābhāsya (Great Commentary) (c. second century BC). Patanjali is one of the few authors of early India whose approximate date is probably known. He lived around the middle of the second century BC in the north of India during the incursions into the country of the Graeco-Bactrian king Menander, to which he refers. This early date is confirmed by the absence of influence from the classical systems of Indian philosophy on his thought, with the possible exception of Sarvāstivāda Buddhism.
The Mahabhäsya is not primarily a philosophical treatise. It addresses some philosophical questions in its introduction (whose title is Paspašāhnika) while commenting upon certain grammatical rules (sutra) of Pāņini. Interestingly, Patañjali (who often sides in this respect with Kātyāyana) does not always share Pāņini's point of view.
References and further reading Brown, R. and Rollins, C.D. (1969) Contemporary
Philosophy in Australia, London: Allen & Unwin. (Articles on sundry philosophical topics by Australians, with an introduction by Alan Donagan on the state of Australian philosophy and its background.)
FRANK JACKSON
2 The linguistic units that have meaning Rule 1.2.45 of Pāņini's grammar states that 'what is meaningful, but is not verbal root or affix, is nominal stem' (arthavad adhātur apratyayah prātipadikam). If we use the term 'stem' to refer to both verbal roots and nominal stems, this sūtra makes clear that for Panini only stems and affixes really have meaning. Combinations of stems and affixes, that is, words and sentences, have at best meanings that are derived from those of the constituent stems and affixes. For Patañjali the situation is the exact opposite of this: words and sentences, not stems and affixes, are meaningful. This is clear from his discussion of rule 1.2.45 where he observes that this sūtra would assign the designation 'nominal stem' (prātipadika) to words and sentences. This contingency is avoided with the help of some far-fetched and unconvincing arguments. Next Patañjali turns to the question of how stems and affixes can be thought to have meaning, which he considers problematic, unlike Pāņini. In Patanjali's opinion the nominal stem vrkṣa on its own expresses no meaning. Cardona (1967-8) delineates the following solution offered by Patanjali: the method of concurrent occurrence of meaning and linguistic unit (anvaya) and absence of these two (vyatireka) shows that vrksa, which is the common part of vrkşas, or 'one tree and vrk sau, or two trees, must have the meaning, 'tree'. In this way, Patanjali derives the meaning of the stem vrksa from the 'real meanings 'one tree' and 'two trees'. The pair pacali,
PATANJALI (c. 2nd century BC) The grammarian Patanjali lived in the second century BG before the appearance of the classical systems of Indian philosophy. The aspects of his thought that we would call philosophical are concerned primarily with questions of meaning and meaning-hearers in language.
1 2 3 4
Patanjali and philosophy The linguistic units that have meaning Do phonemes have meaning? The meaning of a word
248
Page #2
--------------------------------------------------------------------------
________________
PATANJALI
The cooks and pathati, 'he recites', he similarly points out (sūtral.3.1) allows us to assign a meaning to the common part -ati. Speaking generally, the meanings Pof stems and affixes are derived from the meanings of zoomplete words. * Patañjali's deviation from Pāņini is most easily Paplained by the hypothesis that he was influenced by Elbe Sarvāstivāda Buddhists, who had reified phoPoemes and words (and perhaps sentences), but not stems and affixes, into existing 'real' elements dharma). This is confirmed by the fact that Patanjali, too, accepts phonemes and words, but not stems and Muffixes, as independently (and eternally) existing entities. In connection with these independently Existing phonemes and words, Patañjali uses several times the expression sphota, which plays a major role in the discussions of later grammarians. Unlike them, Patañjali does not look upon the sphofa as a meaningbearer, according to Joshi (1967). He speaks, for example, of the sphora of individual phonemes. This sphopa is different from the sound (dhvani) which manifests it. The word or phoneme (sabda) is the sphora and the sound is a property of the word or phoneme. If one person speaks slowly, another quickly, the sphosa is the same, only the manifesting sound is different (see LANGUAGE, INDIAN THEORIES OF 881, 3).
4 The meaning of a word What is the meaning of a word according to Patañjali? In the second ähnika four kinds of words are distinguished according to what they refer to: words that refer to a genus (jätisabda), those that refer to a quality (gunaśabda), those that refer to an action (kriya sabda) and arbitrary proper nouns (yadrcchā sabda). The first three of these clearly designate nouns, adjectives and verbs respectively. But in the first ähnika Patañjali enumerates (twice over) the following four classes of words: nominal words (näman), verbs (akhyāta), preverbs (upasarga) and particles (nipāta). This enumeration is taken from Yāska's Nirukta (c. third century BC (1967: 1.1]), where the meanings of these different types of words are elaborately discussed. Nominal words (nouns and adjectives), for example, are there described as 'having entity as their predominant notion' (sattvapradhāna), verbs as 'having being as their predominant notion' (bhāvapradhāna) (Kahrs 1986: 117). Patañjali was probably aware of the meanings assigned to words there.
Elsewhere (especially in sūtra 1.2.64) Patañjali distinguishes two possible meanings of words: the form (akrti) or the individual object (dravya). The former of these two positions is associated with the name of Vajapyāyana. Vyādi, on the other hand, held that words denote individual objects. Here, it seems, the discussion concerns itself with nouns primarily. In fact, there is reason to believe that for Patanjali 'form (akrti) and 'genus (jani) were synonyms. Patanjali himself held that both form and individual object constitute the meaning of words. See also: MEANING, INDIAN THEORIES OF $81-2
3 Do phonemes have meaning? Although reticent with regard to the meaningfulness of stems and affixes, Patanjali pays great attention in his second chapter (ähnika) to the question of whether Individual phonemes have meaning. He argues that certain stems and affixes consist of just one phoneme. The verbal root i, for example, means 'go'. Certain words which are identical but for one phoneme, express different meanings: kupa means 'well', sūpa soup', yüpa 'sacrificial post'. Also, the removal of one phoneme can change the meaning: vrksa means 'tree, rksa 'bear'. Finally, if phonemes had no meaning, collections of phonemes could have no meaning bither, one hundred blind people cannot see more than what one blind person can see. Patanjali tubsequently rejects these arguments. Sounds normally have no meaning, because they can be modified, Alided, or change position in a grammatical derivadon. Since only collections of phonemes have meanng, nothing can be concluded from sets of similar Words like küpa, süpa and yupa, nor indeed from the air vrksa and rksa. Collections of phonemes can have catures which the constituent phonemes do not Sossess: a chariot, too, can perform functions which ts parts cannot.
List of works Patañjali (c. 2nd century BC) (Vyakarana-) Mahāb
hāşya, ed. F. Kielhorn, 1880-5, 3 vols; 3rd edn, ed. K.V. Abhyankar, Poona: Bhandarkar Oriental Research Institute, 1972. (Comments on Pāņini's Astādhyāyi.) -- (c. 2nd century BC) Mahābhäsya, trans. S.D. Joshi and J.A.F. Roodbergen, Poona, 1968. (Eleven volumes have appeared between 1968 and 1990. They continue to be published.)
References and further reading Bronkhorst, J. (1987) Three Problems Pertaining to the
Mahābhāşya, Poona: Bhandarkar Oriental Research Institute. (Especially chapter 3, "The Ma
249
Page #3
--------------------------------------------------------------------------
________________ PATERNALISM habhasya and the development of Indian philo- sophy':) * Cardona, G. (1967-8) 'Anvaya and vyatireka in Indian grammar', Adyar Library Bulletin 31-2: 313-52. (Discusses the use of this pair of conceptual tools.) - (1980) Panini: a Survey of Research, Delhi: Motilal Banarsidass. (A full bibliographical breakdown of editions and translations of the Mahabhasya.) Joshi, S.D. (1967) The Sphotanirnaya of Kaunda Bhatta, Poona: University of Poona. (The introduction contains a discussion of Patanjali's views on the nature of words.) * Kahrs, E. (ed.) (1986) 'Durga on bhava', in Kalyaami traraganam: Essays in Honour of Nils Simonsson, Oslo: Norwegian University Press, 115-44. (Discusses the possible interpretations of sattvapra dhana and bhavapradhana.) * Panini (c. 4th century BC) Astadhyayi, ed. and trans. 0. Bohtlingk, Leipzig, 1887; repr. Hildesheim and New York: Georg Olms, 1977. (Panini's famous grammar book.) * Yaska (c. 3rd century) Nirukta, ed. and trans. L. Sarup, Delhi: Motilal Banarsidass, 1967. (The meanings of different types of words are discussed in detail.) 1 The concept of paternalism One behaves paternalistically if one treats an adult as though one were a parent dealing with a child. One's behaviour shows concern for the welfare of the person and a presumption that one's judgment about what will promote it is superior. The paradigm of paternalism, and the focus of most philosophical discussion of it, is restriction of people's liberty against their will for their own good. Whether a restriction of liberty is paternalistic or not depends on its rationale. If most people want a law requiring that they wear seat belts when riding in cars for their own protection, and we enact a law in order to cater to this desire, enforcing this policy restricts people's liberty, but not against their will. In its application to the minority who do not want to be under this requirement, the law is still not paternalistic - though it may be unfair - if its rationale is administrative convenience and not the aim of restricting people's freedom against their will for their own good. JOHANNES BRONKHORST PATERNALISM Restriction of people's liberty of action is paternalistic when it is imposed for the good of those whose liberty is restricted and against their will. The argument in favour of paternalism is that, if one can prevent people fron harming themselves, there is no reason not to do so. Versions of the ethical creed of liberalism tend to oppose paternalism. One argument is that as a practical matter the policy of permitting paternalism tends to do more harm than good in the long run, or at least less good than a strict refusal to countenance paternalism would achieve. Another argument appeals to a right of autonomy which paternalism is held to violate whether or not its consequences on the whole are undesirable. Paternalist advocacy can be 'hard' or 'soft'; soft paternalism is the doctrine that paternalism can only be justifiable when the individual action that is being restricted was not chosen in a substantially voluntary way. 2 Utilitarianism for and against paternalism The case for paternalism is simple. If we can prevent people from harming themselves, why not do it? According to act utilitarianism, one ought always to do whatever would produce the most good (utility) for people and perhaps other sentient creatures) (see UTILITARIANISM). Whenever restricting someone's liberty to prevent harm to that very person is the utility-maximizing act, then act utilitarianism requires us to do it. Utilitarianism imposes stricter requirements of benevolence than most people accept (see HELP AND BENEFICENCE). One might hold that the better the cost-to-benefit ratio of a paternalistic act, the more obligatory the performance of the act. Or one might hold that so long as a paternalist imposition would do more good than harm, it is permissible even if not morally required. In On Liberty (1859), a classic statement of a liberal utilitarian antipaternalism, John Stuart Mill does not dispute the theoretical possibility that in particular circumstances restricting someone's liberty for their own good might be the best thing to do from a utilitarian standpoint. Mill nevertheless proposes a liberty principle, which asserts that in modern societies the liberty of a sane and cognitively competent adult should never be restricted except to prevent harm to other persons who do not consent to this involvement (see Mill, J.S. SS 12). Among its several implications, Mill's liberty principle forbids paternalistic restriction of liberty. Mill argues that we should adhere to this principle 1 2 3 The concept of paternalism Utilitarianism for and against paternalism The right to autonomy; hard and soft paternalism 250