World Sanskrit Conference (WSC) (17th : 2018)

Svārthika Suffixes vs. Endocentric Taddhita Derivatives Wielińska-Soltwedel, Małgorzata 2019

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


70440-Wielińska-Soltwedel_Malgorzata_Svarthika_suffixes_2019.pdf [ 704.07kB ]
JSON: 70440-1.0379850.json
JSON-LD: 70440-1.0379850-ld.json
RDF/XML (Pretty): 70440-1.0379850-rdf.xml
RDF/JSON: 70440-1.0379850-rdf.json
Turtle: 70440-1.0379850-turtle.txt
N-Triples: 70440-1.0379850-rdf-ntriples.txt
Original Record: 70440-1.0379850-source.json
Full Text

Full Text

 Svārthika Suffixes vs. Endocentric Taddhita Derivatives  Małgorzata Wielińska-Soltwedel Proceedings of the 17th World Sanskrit Conference, Vancouver, Canada, July 9-13, 2018, Section 3: Vyākaraṇa.  Section Convenors: Malhar Kulkarni and Peter Scharf
General Editor: Adheesh Sathaye Published by the Department of Asian Studies, University of British Columbia, on behalf of the International Association for Sanskrit Studies. DOI: 10.14288/1.0379850.
URI: Suggested Citation Format: MLA:
Wielińska-Soltwedel, Małgorzata. “Svārthika Suffixes vs. Endocentric Taddhita Derivatives.” Proceedings of the 17th World Sanskrit Conference, Vancouver, Canada, July 9-13, 2018, Section 3: Vyākaraṇa. Edited by Malhar Kulkarni and Peter Scharf, 2019. 
DOI: 10.14288/1.0379850. APA:
Wielińska-Soltwedel, M. (2019). Svārthika suffixes vs. endocentric taddhita derivatives. In M. Kulkarni and P. Scharf (eds.) Proceedings of the 17th World Sanskrit Conference, Vancouver, Canada, July 9-13, 2018, Section 3: Vyākaraṇa. DOI: 10.14288/1.0379850. Chicago:
Wielińska-Soltwedel, Małgorzata. 2019. “Svārthika Suffixes vs. Endocentric Taddhita Derivatives.” In Proceedings of the 17th World Sanskrit Conference, Vancouver, Canada, July 9-13, 2018, Section 3: Vyākaraṇa, edited by Malhar Kulkarni and Peter Scharf. 
DOI: 10.14288/1.0379850. Proceedings of the 17th World Sanskrit Conference, July 9-13, 2018 University of British Columbia, Vancouver, CanadaCopyright © 2019 by the author. Content is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International license (CC BY-NC-ND 4.0).   17TH    WORLD   SANSKRIT  CONFERENCEVANCOUVER, CANADA • JULY 9-13, 2018वैधुसव ्मकबुंटुकअ ारा यसं तृा यनसमवायःINTERNATIONAL ASSOCIATION OF SANSKRIT STUDIES THE 17TH WORLD SANSKRIT CONFERENCE, VANCOUVER, CANADA, JULY 9-13, 2018 Svārthika Suffixes vs. Endocentric Taddhita Derivatives Małgorzata Wielińska-Soltwedel Institute of Indology and Tibetology, Ludwig Maximilian University of Munich Munich, Germany. Abstract Neither the term svārthika nor svārtha is taught or used by Pāṇini. The word svārthe appears for the first time in Kātyāyana’s vārttikas and svārthika in Patañ-jali’s commentary. Patañjali does not restrict the usage of these terms exclusively to the secondary suffixes as is usually the case in later Pāṇinian tradition. For him, the feminine suffixes ṭābādi are svārthika and he regards the intensive suf-fix yaṄ and the desiderative suffix saN as added svārthe, in which he partly fol-lows Kātyāyana. The big shift in understanding of both of these terms can be seen in the Kāśikā. While for Patañjali, apart from the samāsānta (and ṭābādi) suf-fixes, only single suffixes are svārthika (or added svārthe), Jayāditya states that the svārthika suffixes begin with the rule A 5.3.1. His view is shared by most of the later grammarians, even if the borders of various svārthikaprakaraṇas could suggest some difference of opinion on that topic. Having examined how the grammarians beginning with Kātyāyana and Patañjali characterize the svārthika suffixes and whether the suffixes taught in A 5.3. and 5.4. are really covered by these remarks, I compare this traditional description with the modern concept of endocentric derivatives as proposed, e.g., by Bhate 1989, Singh 1991 and Ran-garajan 1999. In particular, I discuss the question whether this new approach offers a better understanding of the suffixes prescribed in A 5.3. and 5.4. than the traditional term svārthika and, as a final point, why this is not the case. Keywords: taddhita, svārthika, svārtha, endocentric, exocentric, Bloomfield. 1. In later Pāṇinian tradition the term svārthika refers first of all to a particular subclass of taddhita suffixes. What exactly the word svārthika means seems not to be a matter of heated debate. Only late authors such as Nāgeśa or Vāsudeva Dīkṣita bother to paraphrase this term at all. According to them svārthika suffix-es are svīyaprakṛter arthe bhavā ity arthaḥ (LŚŚ on A  5.3.1) or svīyaprakṛtyarthe Proceedings of the 17th World Sanskrit Conference, Vancouver, Canada, July 9-13 2018, Section 3: Vyākaraṇa, edited by Malhar Kulkarni, Hideyo Ogawa, and Peter Scharf, 2019. DOI: 10.14288/1.0379850. WIELIŃSKA-SOLTWEDEL 2bhavāḥ (Bālam on SK 1947 = A 5.3.1),   i.e. they occur in the sense of their own 1bases.  Thus, a svārthika suffix is added svārthe, literally “in its own meaning,” which is, however, rendered as “the meaning of its own base.”  Svārtha is not 2used by Pāṇini but it is already attested in Kātyāyana’s vārttikas  and is much 3more frequent in grammatical texts than svārthika.  4The elucidation of these two terms suggests nothing that makes them ap-plicable exclusively to taddhita suffixes, so it is not surprising that in the opinion of Kātyāyana the intensive suffix yaṄ as well as feminine suffixes are used svārthe.  Patañjali follows Kātyāyana on that but adds the desiderative suffix saN 5to the group of suffixes which are used svārthe.  Significantly, such remarks are 6quite rare in later grammatical texts. As far as I can see, the Kāśikā and the  Cf. a very similar passage in Bālam on SK 282 (A 6.4.84): svasyāḥ prakṛter arthaḥ svārthaḥ, 1tatra bhavaḥ svārthikaḥ. On the meaning of svārtha see Wezler (1980) and Cardona’s (1983, 101-116; 2004, 149) re2 -action to his opinion. Similar to Wezler, Cardona (1983, 114-115) interprets svārtha as a karmadhāraya compound, but he sees its uses as falling into three slightly different categories: A) as “used of whatever meaning is proper to an item in question,” B) “[i]n a more specialized sense [...] with reference not to a word’s meaning in general but to a property that inheres in every referent of a term” and C) as “used of a meaning associ-ated with an affix but said properly to belong to the element with which the affix oc-curs.” Svārthe or svārtha- are found in vt 5 on A 3.1.22, vt 1 on A 3.3.19, vt 1 on A 3.4.67, vt 5 on A 4.1.3, vt 6 3on A 4.1.163, vt 1 on A 4.2.55, vt 2 on A 5.1.13, vt 1 on A 5.1.57-58, vt 2 on A 5.1.59, vt 1 on A 5.2.37, vt 3 on A 5.3.7, 10, vt 1 on A 5.3.45, vt 3 on A 5.3.55, vt 3 on A 5.3.74, vt 2 on A 5.4.4 and in vt 4 on A 8.1.12. Compare e.g. the Mahābhāṣya, where Patañjali mentions the word svārtha, compounded 4or uncompounded, 87 times (including 60 occurrences of svārthe) and svārthika – 33 times, the latter mostly while describing some general features of these suffixes. See vt 5 (na vā nānārthatvāt kartṛkarmaṇoḥ hi lavidhānam kriyāviśeṣe svārthe yaṅ) on A 3.1.22 5and vt 5 (siddhaṃ tu striyāḥ prātipadikaviśeṣaṇatvāt svārthe ṭābādayaḥ) on A 4.1.3. See svārthe yaṅ on A 3.1.22 vt 5 (Mbh II 30.6). The phrase svārthe ṭābādayaḥ occurs in the 6commentary on A 2.2.24 vt 9 (Mbh I 423.2-3) and on A 4.1.1 vt 21 (Mbh II 195.16, but here it is just a quotation of vt 5 on A 4.1.3); svārthikāṣ ṭābādayaḥ appears in Mbh on A 4.1.1 vt 21 (Mbh II 195.17, but it is the repetition of Patañjali’s commentary on A 4.1.3 vt 5), on A 4.1.3 vt 5 (Mbh II 199.20) and on A 4.1.78 (Mbh II 231.10). Svārthika Suffixes vs. Endocentric Taddhita Derivatives 3Bhāṣāvṛtti, for example, explicitly  refer only to the causative suffix ṆiC as added 7svārthe and not yaṄ, saN or ṭābādi.  82. Contrary to Patañjali, who occassionally calls feminine suffixes svārthika (see fn. 6) the grammarians following him employ this term only in regard to the par-ticular group of taddhita suffixes. However, it appears to be slightly unclear ex-actly which group it is. This initial impression seems to gain more strength if we have a look at some dictionaries of Sanskrit grammatical terms.  For example, Renou (1957: 349) sees differences of opinions between vari-ous grammatical treatises. The svārthika section begins – in his view - with A 5.3.1 according to Jayāditya, with A 5.3.26 according to Puruṣottamadeva, and with A  5.3.96 according to Bhaṭṭoji Dīkṣita. Contrary to this, Abhyankar (1961: 411) does not mention any such differences; instead, he maintains that svārthikas are taught in rules A 5.3.36 up to 5.4.67. He adds that suffixes termed samāsānta can be also called svārthika, which is obviously inspired by Patañjali’s dictum samāsāntā api svārthikāḥ.  Finally, for Roodbergen (2008: 476) all the suffixes pre9 -scribed in the last two pādas of the fifth adhyāya belong to svārthikas.  What is the reason for this disagreement? Is there actually any controversy over svārthika suffixes among Sanskrit grammarians as suggested by Renou? On closer inspection this question must be answered in the negative. To be sure, there are different approaches to svārthika suffixes, but the main line of division lies between Patañjali and Jayāditya. While Kātyāyana and Patañjali occasionally remark that some particular suffix or – seldom – several suffixes occur svārthe or are svārthika, the grammarians beginning with Jayāditya state that svārthika suffixes start with A 5.3.1 (ataḥ paraṃ svārthikāḥ). This opinion is found in almost all commentaries, with the slight difference that Puruṣottamadeva puts this re-mark just before A 5.3.1. Interestingly, he repeats the very same sentence in the commentary on A  5.3.25. The next sūtra is the last one prescribing suffixes termed vibhakti, but as it deals with the Vedic language it is not commented  Cf., however, Kāś on A 4.1.3: strītvaṃ ca pratyayārthaḥ prakṛtyarthaviśeṣaṇaṃ ca ity ubhaya-7thāpi prayujyate (“The feminine gender is used in both ways: as the meaning of the suf-fix and as the qualifier of the meaning of the base.”). See curādibhyaḥ svārthe in Kāś on A 3.1.25. This remark is also repeated in Bhv on the 8same sūtra. See Patañjali on A 4.1.1 vt 21 (Mbh II 195.20).9 WIELIŃSKA-SOLTWEDEL 4upon by Puruṣottamadeva.  Consequently, ataḥ paraṃ svārthikāḥ appearing in 10the Bhāṣāvṛtti on A 5.3.25 should rather be treated as an introductory remark to A 5.3.27 and not to rule 26 as it was obviously understood by Renou.  Renou’s (1957: 349) claim that according to Bhaṭṭoji Dīkṣita svārthika suf-fixes are taught in rules beginning with A 5.3.96 has, most probably, very simple grounds. Renou (1957: ix) used the Lahore edition of the Siddhāntakaumudī, pub-lished in 1935, presumably the one prepared by Mādhavaśāstrī Bhāṇḍārī.  In11 -deed, in this edition it is A 5.3.96 which opens the svārthika section. However, if we compare various editions of the Siddhāntakaumudī we notice that different editors put different limits to this section.  Consequently, to avoid the danger of 12being misled by a particular editorial decision, it seems to be more appropriate to study what Bhaṭṭoji actually writes about svārthika suffixes.  At the beginning of the section which precedes the svārthikaprakaraṇa, i.e. in the Siddhāntakaumudī on A 5.3.1 we discover the familiar remark ataḥ paraṃ svārthikāḥ. This makes it clear that for Bhaṭṭoji svārthika suffixes begin with rule A 5.3.1 and not with A 5.3.27 or even 5.3.96. To find out where, in his view, the svārthika section should end and especially whether he considers samāsānta suf-fixes to be svārthika it would be particularly rewarding to consult his commen- Strictly speaking, this sūtra, as many others describing the Vedic language, is followed 10by the remark chandaḥsūtram (“This is a sūtra concerning the Vedic language”). The fact that Puruṣottamadeva announces the beginning of the treatment of svārthika suffixes under the rule A 5.3.25 could suggest that the comment chandaḥsūtram does not stem from him but rather has been added by some diligent scribe. At least this is the only edition published in Lahore in 1935 which I have been able to find.11 For example, in Vasu 1982 the svārthika section comprises rules A 5.3.27-95, in Nene 121977 this section begins with A  5.3.27 but ends with A  5.4.67; in Sastrigal 1910 and Bhāṇḍāri 1935 the svārthikaprakaraṇa consists of rules A 5.3.96-5.4.67. Svārthika Suffixes vs. Endocentric Taddhita Derivatives 5tary on A 5.4.7, where he lists obligatory svārthika suffixes.  As he mentions here 13suffixes prescribed in different rules scattered in the third and fourth pāda of the fifth adhyāya, including also samāsānta suffixes, dealt with at the end of the fourth pāda, we have to accept that Bhaṭṭoji also shares the opinion of Jayāditya that suffixes taught in the last two pādas of the fifth adhyāya are called svārthika.  This means that there is no significant difference between the opin14 -ions of Jayāditya, Puruṣottamadeva and Bhaṭṭoji Dīkṣita. With respect to Abhyankar’s (1961: 411) view it is obvious that A  5.3.36 dakṣiṇād āc, which he mentions in his dictionary, must be just a misprint. Firstly, this sūtra lies in the middle of rules in which rule A 5.3.27 is understood by anu-vṛtti. Secondly, there is nothing special about it suggesting the possibility that it could open the svārthika section. However, it is more difficult to decide which rule Abhyankar actually had in mind: A 5.3.26, or more correctly 27, or perhaps rule 96. Whatever his intention, it is clear that it is rather Roodbergen (2008: 476) who is right when he says that svārthika suffixes are prescribed in the last two pādas of the fifth adhyāya.  But what kind of suffixes are actually taught there? Do they indeed have anything in common? 3. Certainly, the sūtras in the last two pādas of the fifth adhyāya are not typical for the taddhita section, but more in agreement with the rest of the Aṣṭādhyāyī. These two pādas lack the usage of pronouns to express the meaning of deriva- See SK 2079: anye ’pi kecit svārthikāḥ pratyayā nityam iṣyante. tamabādayaḥ prāk kanaḥ. 13ñyādayaḥ prāg vunaḥ. āmādayaḥ prāṅ mayaṭaḥ. bṛhatījātyantāḥ samāsāntāś ceti “Others maintain that some svārthika suffixes are always required. Those beginning with tamaP (A 5.3.55) up to kaN (A 5.3.95), Ñya (A 5.3.112) etc. up to vuN (A 5.4.1), āmU (A 5.4.11) etc. up to mayaṬ (A 5.4.21), [the suffix kaN added] after bṛhatī (A 5.4.6), [the suffix cha added] after what ends in jāti (A 5.4.9) and samāsāntas.” Some editions (e.g. Nene 1977 and Bhāṇḍāri 1935) give the source of this quotation (...ceti bhāṣyam). The similar pas-sage is indeed found in Mbh on A 5.4.7 vt 2 (II 431.15- 16: ke punar nityapratyayāḥ. tamā-dayaḥ prāk kano ñyādayaḥ prāg vuna āmādayaḥ prāṅ mayaṭo bṛhatījātyantāḥ samāsāntāś ceti), with the crucial difference that Patañjali speaks of nityapratyayas and Bhaṭṭoji of svārthikas which are nitya. Taking into consideration Bhaṭṭoji’s commentary on A 5.3.1, 27 and on 5.4.7 it is diffi14 -cult to agree with Joshi and Roodbergen (2006: 114) that “[t]he SK does not acknowl-edge svārthika suffixes as a separate group within the taddhita section.” WIELIŃSKA-SOLTWEDEL 6tives, usage so characteristic of other taddhita rules.  Sūtras of the type tasyā15 -patyam (A 4.1.92) or sāsya devatā (A 4.2.24) are practically absent in the svārthika section.  Instead, the meaning condition is given – as it is normally the case in 16the Aṣṭādhyāyī – in the locative. On the whole, the class of svārthikas seems to comprise rather varied suf-fixes. However, there are two main groups we can distinguish among them, groups which have already been recognized by Pāṇini himself. These are vibhakti suffixes, taught at the very beginning of the third pāda up to A  5.3.26, and samāsānta suffixes dealt with at the end of the fourth pāda, starting with A 5.4.68. Already these two groups show how divergent svārthika suffixes can be. Vibhakti suffixes taught in A 5.3 occur optionally after a nominal termination, also termed vibhakti, before which the nominal termination is deleted, and the derivative built with them are termed avyaya. Samāsānta suffixes, on the other hand, are mostly  obligatory and they usually form declinable derivatives. How17 -ever, both kinds of suffixes deserve to be called svārthika, as they generally do not change the meaning of their bases. Tatra is an equivalent to tasmin or tasyām, the stem mahārāja has the same meaning as *mahārājan, with the difference that *mahārājan does not appear in real language, it always requires the addition of the suffix ṬaC. What can we say about the suffixes taught between these two groups? They are for sure very heterogeneous. Some of them, although not termed so, are similar to vibhaktis. They are treated directly after vibhakti suffixes up to A 5.3.46, but also in the fourth pāda, namely, in A 5.4.42-49. Still another group of svārthikas forms comparatives, another one belongs to the suffixes called tadrāja. Three other suffixes deserve a separate mention. These are CvI, trā and  For the analysis of meaning-rules within the taddhita section see Wezler (1975: 108- 135) 15and Bhate 1987. As already pointed out by Wezler (1975: 126-127), pronouns in the simi-lar function, i.e. acting as a kind of variable, are also present – although very seldom – outside adhyāyas 4 and 5. There are only two examples of somewhat similar usage of the pronoun tad in this por16 -tion of the Aṣṭādhyāyī, i.e. in A 5.4.21 tatprakṛtavacane mayaṭ and in A 5.4.54 tadadhīnava-cane. The pronoun tad in A 5.3.77 nītau ca tadyuktāt and in A 5.4.36 tadyuktāt karmaṇo 'ṇ has a different function as it refers to the meaning mentioned in the previous rules and not to bases to which the suffixes in question are added. Although Patañjali lists samāsāntas among obligatory suffixes (see fn. 13), some of 17them are optional, see e.g. A 5.4.109 napuṃsakād anyatarasyām or A 5.4.121 nañduḥsu-bhyo halisakthyor anyatarasyām. Svārthika Suffixes vs. Endocentric Taddhita Derivatives 7sātI, added to bases and used mostly with verbs kṛ, bhū and as, building forms like śuklīkaroti or udakasād bhavati. Last but not least, there is a small class of somehow unusual “lexical” suffixes like e.g. pāśaP (A 5.3.47), caraṬ (A 5.3.53), rūpya (A 5.3.54) and deśya (A 5.3.67).  Two svārthikas are even no suffixes at all: akAC 18(A 5.3.71) is an infix and bahuC (A 5.3.68) a prefix. As far as the meaning of derivatives is concerned, most svārthika suffixes can be broadly classified into two groups. Some of them do not cause any change in the meaning of their bases. Devatā means the same as devaḥ, avikaḥ as aviḥ. These suffixes are called by Nāgeśa atyantasvārthika.  Others seem to add some 19extra meaning or at least a connotative nuance.  Aśvakaḥ is not just a “horse,” 20but “a horse whose owner is unknown” (A  5.3.73) or “a poor horse” (A  5.3.74); vṛkṣaka is “a small tree” (A 5.3.86) and tailaka, “a small quantity of oil” (A 5.3.85). 4. How do commentators deal with the second group? Obviously, once the status of svārthika is given to these suffixes, the locative in the sūtra cannot be inter-preted as pratyayārtha – “meaning of a suffix.” Instead, it is said to be prakṛ-tyarthaviśeṣaṇa – “qualifier of the meaning of the base.” Thus, when a prātipadika expresses the meaning or is qualified by the meaning given in the sūtra, the par-ticular suffix taught there can be added. This is already the way Patañjali ex-plains the locatives occurring in sūtras A  5.3.66, 67 and 74 (see Mbh II 418.8,  However, the suffix rūpya is also taught in A 4.3.81 hetumanuṣyebhyo ‘nyatarasyāṃ rūpyaḥ.18 See the introduction to paribhāṣā 113 in PŚ.19 On the question whether all suffixes, i.e. also all svārthikas, convey meaning see Car20 -dona (1983, 81-101). WIELIŃSKA-SOLTWEDEL 8419.12, 424.23)   but Jayāditya employs this or a similar kind of interpretation 21much more frequently.  225. There are, certainly, good grounds to be sceptical about this kind of interpreta-tion or – to be less sympathetic – about this trick of the commentators. It is also more than understandable that we might feel inclined to look for some other possible description of the suffixes taught in these two pādas. The logical step in that case would be to have recourse to modern linguistics. This, however tempt-ing and somehow natural it may seem, is not without risk since using the termi-nology of one system for another which is completely different can easily turn out to be quite thorny. Still, it is certainly worth trying, although only on condi-tion that the new proposition should yield better results than the traditional one.  These are occurrences of this compound in the context of taddhita suffixes. Patañjali 21applies also the shorter form prakṛtiviśeṣaṇa: twice while explaining the usage of tad-dhita suffixes, i.e. in the commentary on A 5.2.27 (Mbh II 375.16) and on A 5.2.126 (Mbh II 400.14). See e.g. prakṛtyarthaviśeṣaṇam caitat (Kāś on A  5.3.55, 67, 69, 74), prakṛtyarthasya 22viśeṣaṇaṃ caitat (Kāś on A 5.3.66) or prakṛtiviśeṣaṇaṃ caitat (Kāś on A 5.3.53, 85). These statements are sometimes followed by the further clarification -viśiṣṭe 'rthe vartamanāt (prātipadikāt) (svārthe) ... pratyayo bhavati (Kāś on A 5.3.53; similarily on A 5.3.55, 66, 69, 85), with the variant -upodhike 'arthe instead of -viśiṣṭe 'rthe (Kāś on A 5.3.74). This kind of elucidation is often used on its own (e.g. Kāś on A 5.3.42, 68, 73, 75, 86; 5.4.21, 41), not seldom with the phrase -viśiṣṭe 'rthe / -upodhike 'rthe being replaced by the locative of meaning, e.g. yāpye vartamānāt prātipadikāt svārthe pāśap pratyayo bhavati (Kāś on A 5.3.47, similarly Kāś on A 5.3.95, 5.4.6, 9, 30, 31, 33, 35, 37). Rarely Jayāditya employs a compound with -vācin to paraphrase the locative in the sūtra, e.g. ekaśabdād asahāya-vācinaḥ svārthe ākinic pratyayo bhavati (Kāś on A 5.3.52, similarly Kāś on A 5.3.92). An-other common practice of Jayāditya is to add gamyamāne (gamyamānāyām) or dyotye (dyotyāyām) to the locative of the sūtra (Kāś on A 5.3.43, 56, 76, 88-90, 101; 5.4.1, 3). On the whole, the Kāśikā treats the locatives in rules at issue much more meticulously than later commentaries. Compare e.g. Kāś’s explanation of A  5.4.85 (parimāṇāpacaye al-paśabdaḥ. prakrt̥iviśeṣaṇaṃ caitat. alpatvaviśiṣṭe arthe vartamānāt prātipadikāt yathāvihitaṃ pratyayo bhavati. alpaṃ tailam tailakam. ghrt̥akam. sarvakam. viśvakam. uccakaiḥ. nī-cakaiḥ. pacataki. jalpataki.) with the short commentary of the Bhv (alpārthe kākacau syātām. alpaṃ ghṛtaṃ ghṛtakam. uccakaiḥ. viśvake. pacataki.) or even shorter of the SK (alpaṃ tailaṃ tailakam.). Svārthika Suffixes vs. Endocentric Taddhita Derivatives 9Is that true of exocentric and endocentric derivatives or constructions which are sometimes found besides asvārthika or prakṛtyarthabhinnārtha and svārthika suffixes?  23Do these modern concepts really give us a better understanding of taddhi-ta suffixes? This question seems to be inevitable, especially if we notice that the term endocentric derivative generally corresponds to a derivative built with a svārthika suffix.  This looks just like putting a modern label on an old concept 24with the only difference being that while Indian grammarians speak of pratyayas we speak rather of derivatives.  256. To decide whether the concept of an endocentric derivative is superior to that of a svārthika pratyaya, let us have a closer look at the terms exocentric and endo-centric. Both of them are met with in linguistics, but they pertain mainly to syn-tactic phrases. In fact, they were introduced by Leonard Bloomfield in his famous work Language. Bloomfield 1933 defines these terms firstly in relation to phrases (p. 194) and then extends their application to compounds as well (p. 235), but – sig-nificantly – not to secondary derivatives, which he treats directly after com-pounds.  It is for sure not just a blunder on his part, but there are good reasons 26 See e.g. Bhate 1989, Singh 1991, Rangarajan 1999.23 Rangarajan (1999: 131) states explicitly that “the constructions which are formed with 24svārthika type of suffixes may be brought under the endocentric construction” and that “[t]he construction which is formed out of the addition of asvārthika type of suffixes [...] falls under the purview of the exocentric construction.” This is close to the opinion of Bhate (1989), except for some inconsistences in her argumentation. For example, on p. 14 she maintains that “the first sub-section [of the A] covering rules between P. 4.1.82 and P. 5.3.72 deals with exocentric derivatives, whereas the second sub-section comprising rules from P. 5.3.73 to the end of the fourth pāda deals with endocentric derivatives,” but already on p. 15 we read that “[t]he whole tad. section is basically di-vided into two parts, one dealing with the rules deriving exocentric words (P. 4.1.92 - P. 5.2.140) and the other dealing with rules deriving endocentric words (P. 5.3.1 - P. 5.4.160).” P. 4.1.92 is probably just a misprint, but hardly rule 5.3.1. On Singh’s ap-proach (1991), which is different from that of Bhate and Rangarajan, see fn. 29. To be precise, the terms used in this context are construction and derivative. Apart from 25quotations or close references to the opinions of others I prefer to use here the term derivative instead of construction. Bloomfield deals with compounds in § 14.1-14.6 (p. 227-237) and with secondary deriv26 -atives in § 14.7-14.9 (p. 237-246). WIELIŃSKA-SOLTWEDEL 10for this omission and they are to be sought in the definition of exocentric and endocentric phrases or compounds propounded by him. According to Bloomfield (1933: 194, 227) both constituents of a phrase or a compound have to be free forms in the first place. A resultant phrase or a com-pound is then called exocentric when it belongs to another form-class than its constituents and is called endocentric when it belongs to the same form-class as one of its constituents (Bloomfield 1933: 194, 235). The classical examples would be John runs for an exocentric construction and poor John for an endocentric one. Two aspects have to be emphasized here, aspects which – in my opinion – make the application of the terms exocentric and endocentric with regard to sec-ondary derivatives most problematic. Firstly, it is the condition that the con-stituents of a phrase or a compound should be free forms. This is certainly not the case with secondary derivatives. Even if following Bloomfield (1933: 225), we agree that stems are to be treated as if they were free forms, this obviously can-not be said of any kind of affixes which are typical bound forms. Supposing we decided to ignore this condition – this would mean that e.g. the endocentric de-rivative should just belong to the same form-class as its stem. And this is the second problem we encounter. For Bloomfield (1933: 185) linguistic forms belong to the same form-class when they have the same func-tion, in other words, they appear in the same positions. This makes form-classes somehow similar to the traditional parts of speech, with the important differ-ence that the form-classes are defined exclusively on formal and not on semantic grounds.  In consequence, we would have to accept that e.g. apatya derivatives 27are to be treated as endocentric ones since – if we eliminate every trace of se-mantics from our analysis – Gārgya can fill in every position that Garga can pos-sibly take. At the same time, some derivatives built with svārthika suffixes have to be interpreted as exocentric ones, because they belong to different form-class than their stem. Strictly speaking even devatā, although semantically equal with deva, should be treated as an exocentric derivative since with the change of gen- See Bloomfield (1933: 185, 196, 265-266).27 Svārthika Suffixes vs. Endocentric Taddhita Derivatives 11der it changes the form-class of its base.  This is, however, not what most au28 -thors using these modern terms would be willing to accept.  297. No wonder all definitions of exocentric and endocentric derivatives I have come across do not make use of the term form-class. Instead, they are formulated rather semantically than syntactically. Thus, we read that the endocentric derivative (or construction) “exclusively refers to the meaning of the constituents” (Rangarajan 1999, 130-31) or “to one of its constituents” (Singh 1991: 75) or that “the stem and its derivative refer to one and the same object” (Bhate 1989: 14) or that the suffix “mod-ifies the meaning of the base” (Rangarajan 1999: 131). This is a radical change compared to the way Bloomfield understood the terms endocentric and exocentric. In fact, agreeing to these definitions would mean taking a path he would never have chosen. For in his opinion “[t]o accept defini-tions of meaning, which at best are makeshifts, in place of an identification in formal terms, is to abandon scientific discourse” (Bloomfield 1933: 266). Leaving aside the general question of how far we can go in adapting the terms and their definitions to our needs or how much is left out of Bloomfieldian definition if we get rid of free forms and form-class, the problem with the se-mantic definitions of endocentric derivatives is that they do not provide us with a method or an explanation easy to apply and revealing the genuine nature of suffixes taught in the two last pādas of the fifth adhyāya. Neither can these defi-nitions offer us a better line of division for taddhita suffixes. For they turn out to  See Bloomfield’s examples (1933: 236).28 This does not hold true for Singh (1991), who does not necessarily equate endocentric 29derivatives with those built with suffixes taught in A 5.3.1-5.4.160. However, as his ex-planations are scarce I admit that I do not quite understand his point. For him “[d]erivatives referring to one of its constituents are termed endocentric (i.e. having their center of reference within) and exocentric otherwise” (p. 75), but examples given by him are rather surprising. Derivatives such as bhārgava “an offspring of Bhṛgu” (A  4.1.92), ghaṭika “one who crosses (a river etc.) with the help of a pitcher” (A 4.4.7), vaiyākaraṇa “one who studies grammar” (A 4.2.59), grāmya “born in a village” (A 4.3.25), mayūra “a flock of peacocks” (A 4.2.44) are regarded by him to be en-docentric while vāstra “wrapped with cloth” (A 4.2.10), prāsthikī “(a cattle [sic]) that con-tains one prastha (a measure)” (A  5.1.52), gavya “(grass) good for cows” (A  5.1.2), prāsādīya “a palace may be possible to construct (on this ground)” (A 5.1.16), gomān “one who possesses cows” (A  5.2.94), etc., are supposed to be exocentric (translations ac-cording to Singh 1991: 75). WIELIŃSKA-SOLTWEDEL 12be ambiguous or simply wrong, that is to say, leading to results which are gener-ally not wished for by the authors propounding them. How can we, for example, understand the claim that in the case of the en-docentric derivative the suffix modifies the meaning of its stem? How far can this modification go? Or what does it mean that the endocentric derivative ex-clusively refers to the meaning of its constituents? In a way, an apatya derivative, mentioned already, or perhaps even each derivative, refers somehow to the meaning of its stem. Taken in the sense of “having the same referent” this defini-tion is close to the statement that “the stem and its derivative refer to one and the same object,” which is much more accurate and at the same time much easi-er to prove and – to criticise. Actually, not a few derivatives built with suffixes taught in the last two pādas of the fifth adhyāya do not refer to the same object as their stems. Just to mention two examples, the word pādya, “the water used to wash feet” (A 5.4.25), does not refer to the same object as pāda, “foot,” and which teacher (ācārya) would want to be called ācāryadeśīya, “an unaccomplished teacher” (A 5.3.67)? Strictly speaking, this definition suits only atyantasvārthika suf-fixes. 8. Indeed, the possibility that the concepts of exocentric and endocentric deriva-tives can satisfactorily replace the traditional classification of taddhita suffixes seems to be very small when we consider their likely origin. Even if these terms appear in Bloomfield’s work firstly in the context of syntactic constructions and then their usage is extended to compounds, there are good grounds to believe that for Bloomfield himself it worked the other way round. As Wujastyk (1982: 182) has already shown, it is highly probable that Bloomfield was inspired by In-dian grammarians and applied the terms endocentric and exocentric first to com-pounds and then to syntactic constructions.  Thus endocentric would be an Eng30 - Bloomfield was not only a great admirer of Pāṇini, but he also had considerable 30knowledge of the Aṣṭādhyāyī. See Wujastyk (1982: 179) on that. In fact, while discussing types of morphological constructions Bloomfield (1933: 235, 237) explicitly refers to the Indian system of compounds. Svārthika Suffixes vs. Endocentric Taddhita Derivatives 13lish equivalent of svapadārtha (-pradhāna), and exocentric would correspond to anyapadārthapradhāna.  31This would mean that to describe Sanskrit secondary derivatives we would use English equivalents of Sanskrit concepts developed not for secondary deriva-tives but for compounds and then change them so that they suit better other languages, first of all English, and as a final step we bring these concepts into line with Sanskrit secondary derivatives. This looks like a task with very little prospects for success, to say the least. 9. Most importantly, the question arises why we should do such a thing at all. Why can we not use the original Sanskrit terms, even if they are perhaps not free of faults? Or we could take one step further: forget the opinions of the commen-tators for a while and ask what the original intention of Pāṇini was. With rule A 5.3.1 did he really want to start a new svārthika section, without naming it so? And in that case it would be essential to ponder the concept of artha he possibly had.  Or did he rather treat in rules preceding this sūtra those secondary deriva32 -tives which have easily defined meanings, meanings which can be expressed in phrases such as tasyāpatyam, tasya vikāraḥ, etc., gathering in the last two pādas those suffixes which for some reason do not follow this scheme? This would make out of these two pādas a kind of śeṣa, or a waste bin section, and this would also explain why it is so difficult to find a satisfactory definition covering all the suffixes taught there.   Patañjali employs anyapadārthapradhāna several times, first of all, when he classifies 31compounds into four categories according to the fact which meaning is predominant in a compound: whether it is the meaning of the first constituent (pūrvapadārthaprad-hāna), of the last (uttarapadārthapradhāna), of both (ubhayapadārthapradhāna) or of some other word not being a part of a compound (anyapadārthapradhāna); see Patañjali on A 2.1.6 (Mbh I 378.24-379.3), repeated on A 2.1.20 (Mbh I 382.7-10) and on A 2.1.49 (Mbh I 392.7-9). Anyapadārthapradhāna occurs also in the commentary on A 2.2.6 (Mbh I 410.9- 10), but elsewhere in the context of compounds Patañjali speaks of anya-padārtha, occasionally of anyapadārthatva, anyapadārthatā or of anyapadārthāśraya, sometimes contrasting those with svapadārtha (on A 2.1.20 [Mbh I 382.16], on A 2.1.69 vt 5 [Mbh I 404.7], vt 6 [Mbh I 405.2] and on A 2.2.24 [Mbh I 420.4-56, 12]) or with sva-padārthāśraya (on A 2.1.20 [Mbh I 382.16]). The compound svapadārthapradhāna is not found in the Mbh. Cf. the remark of Joshi and Roodbergen (2006: 116) regarding Patañjali’s viśeṣaṇa view 32of svārthika suffixes: “Obviously, the concept of meaning adopted here leaves several questions unanswered.” WIELIŃSKA-SOLTWEDEL 14Acknowledgments I wish to thank Professor Albrecht Wezler for critical comments on the pre-final version of this paper. paper. I am also deeply grateful to Professors Peter Scharf and Malhar Kulkarni for their inspiring remarks and – last but not least – to Ms. Ann Roche B.Sc. (Hons.) for correcting my English. Bibliography Abbreviations A Aṣṭādhyāyī. Edited by Otto Böhtlingk. Hildesheim and New York: Georg Olms, 1977. Reprint of the Leipzig edition 1887. Bālam Vāsudeva Dīkṣita. Bālamanoramā. Edited by Bhāṇḍāri (1935). Bhv Puruṣottamadeva. Bhāṣāvṛtti. Edited by Srish Chandra Chakravarti. Raj-shahi: Varendra Research Society, 1918. Kāś Jayāditya and Vāmana. Kāśikā. Edited by Aryendra Sharma, Kshanderao Deshpande, and D. G. Padhye. Sanskrit Academy Series, 62. Hyderabad: Sanskrit Academy, 2008. Reprint of the 1969-70 edition. LŚŚ Nāgeśa. Laghuśabdenduśekhara. Edited by Narahari Śāstri Pendse. Re-Edi-ted by Gopāla Śāstri Nene. The Kashi Sanskrit Series, 5. 2 vols. 2nd ed. Va-ranasi: Chaukhambha Sanskrit Sansthan, 1987. Mbh Patañjali. Mahābhāṣya. Edited by Kielhorn (1880–1885). PŚ Nāgeśa. Paribhāṣenduśekhara. Edited by Franz Kielhorn. Bombay Sanskrit Series, 2. Bombay: Indu Prakash Press, 1868. SK Bhaṭṭoji Dīkṣita. Siddhāntakaumudī. See Vasu (1982). vt Kātyāyana’s vārttika. See Kielhorn (1880-1885). Works Cited Abhyankar, Kashinath Vasudev. 1961. A Dictionary of Sanskrit Grammar. Baroda: Oriental Institute.  Bhāṇḍāri, Mādhava Śāstri. 1935. Śrīmadbhaṭṭojidīkṣitaviracitā vaiyākaraṇasiddhān-takaumudī. Śrīmadvāsudevadīkṣitaviduṣā viracitayā bālamanoramākhyavyā- Svārthika Suffixes vs. Endocentric Taddhita Derivatives 15khyayā samalaṅkṛtā, vaidikabhāge śrīmajjayakṛṣṇapaṇḍitaviracitayā subodhinī-vyākhyayā sanāthīkṛtā, liṅgānuśāsanabhāgebhairavīvyākhyopetā, ante gaṇapāṭhādibhiḥ saṃyojitā. Pūrvārdham. Lahore: Motilal Banarsidass.  Bhate, Saroja. 1987. “The Meaning-Adhikāras in the Taddhita Section of the Aṣṭādhyāyī: An Analysis.” Indo-Iranian Journal 30: 81-92.  ———. 1989. Pāṇini’s Taddhita Rules. Publications of the Centre of Advanced Stu-dy in Sanskrit, Class B, 10. Pune: University of Poona.  Bloomfield, Leonard. 1933. Language. New York: Henry Holt. Cardona, George. 1983. “Bases, Affixes, and Meanings.” In Linguistic Analysis and Some Indian Traditions, edited by George Cardona. Bhandarkar Oriental Research Institute Post-graduate and Research Series, 20 [Pandit Shripad Shastri Deodhar Memorial Lectures (first series)]. Poona: Bhandarkar Oriental Research Institute, 37-116.  ———. 2004. Recent Research in Pāṇinian Studies. Second Revised Edition. Delhi: Motilal Banarsidass.  Joshi, S. D. and J. A. F. Roodbergen. 2006. The Aṣṭādhyāyī of Pāṇini with Translation and Explanatory Notes. Vol. XII (7.3.1-7.3.120). New Delhi: Sahitya Akademi. Kielhorn, Franz. 1880–1885. The Vyâkaraṇa-Mahâbhâshya of Patanjali. 3 vols. Bom-bay: Government Central Book Depôt. Nene, Gopāla Shastrī. 1977. Vaiyākaraṇa Siddhāntakaumudī of Bhaṭṭojidīkshita. 5th ed. The Haridas Sanskrit Series 11. Varanasi: Chowkhamba Sanskrit Series Office. Rangarajan, G. 1999. Word-Formation in Sanskrit. [With Special Reference to Taddhita-formation]. Delhi: Sharada Publishing House.  Renou, Louis. 1957. Terminologie grammaticale du sanskrit. Paris: Champion. Re-print of the 1942 edition [Bibliothèque de l’Ecole des Hautes Etudes, fasc. 280-282].  Roodbergen, J. A. F. 2008. Dictionary of Pāṇinian Grammatical Terminology. Publi-cations of the Last Decennium of the B.O.R.I. Centenary 1. Pune: Bhan-darkar Oriental Research Institute. Sastrigal, S. Chandrasekhara. 1910. The Siddhânta Kaumudî of Sri Bhattoji Deekshita with the commentary Srî Bâlamanôramâ of Sri Vasudeva Deekshita. Pûrvârdham. Trichinopoly: St. Joseph’s College Press.  WIELIŃSKA-SOLTWEDEL 16Singh, Jag Deva. 1991. Pāṇini: His Description of Sanskrit. An Analytical Study of the Aṣṭādhyāyī. New Delhi: Munshiram Manoharlal.  Vasu, Śrīśa Chandra. 1982. The Siddhānta Kaumudī of Bhaṭṭoji Dīkṣita. edited and translated into English. 4 vols. Delhi - Varanasi – Patna: Motilal Banarsidass. Reprint of the Allahabad edition, 1906. Wezler, Albrecht. 1975. Bestimmung und Angabe der Funktion von Sekundär- Suffixen durch Pāṇini. Wiesbaden: Franz Steiner.  —. 1980. “Studien zu Patañjalis Mahābhāṣya I: Der Term svārtha und Paribhāṣā CXIII.” In Festschrift Paul Thieme zur Vollendung des 75. Lebensjahres darge-bracht von Schülern und Freunden. Edited by Georg Buddruss and Albrecht Wezler. Studien zur Indologie und Iranistik 5-6: 279-309. Wujastyk, Dominik. 1982. “Bloomfield and the Sanskrit Origin of the Terms ‘Exocentric’ and ‘Endocentric.’” Historiographia Linguistica 9: 179-184.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items