THE TRIPADI BEING AN ABRIDGED ENGLISH RECAST OF PURVATRASIDDHAM (AN ANALYTICAL-SYNTHETICAL INQUIRY INTO THE SYSTEM OF THE LAST THREE CHAPTERS OF PANINIS ASTADHYAYI) BY H. E. BUISKOOL, PH D LEIDEN E J BRILL 1939Copyright 1939 by E. F. Brill, Lesden, Holland. Alf rights reserved, wncluding the right ta translate or to reproduce this book or parts thereof many form. Printed wn the NetherlandsCONTENTS Page SOURCES AND LITERATURE XII INTRODUCTION CiarPrer I The Astadhyayi Introductory observations and general analysts I General form and contents of the Astadhyayt Style instead of system? §1 t Il Nature of the Astadhyayi The Indran word § 2 A Vydkarana Cabdanugisana 4 B Word clements 1 Form 5 2 Function bee MII Methodical planning of the Astadhyay: Analysis synthests § 3 A General plan 12 B. The two parts of the Astadhyay: 15 IV General analysis of the two parts $ 4 A First part 17 B Second part 19 V Principles of Panins system § 5 A Grouping 21 B Junction Combmation Brevity of expression 23 c Neutralisation of sutras 25 D Logical mnemotechmical and asiddhatva principle 26 THE TRIPADI Carrer It Panbhasis and Adhikaras Their relation to the Tmpadi { Paribhasas § 6 A Panbhasis mentigned by Panini himself 1 Therr normal application also with reference to the sutras of the Tripadi 27 2 Interpretation of the sthantvat sutras 28 B Partbhasis which do not occur in the Astadhyay1 38Vul CONTENTS * Page I] Adhtkaras § 7 A Their validity also with reference to sutras of the Tripadi 44 B The adhikdra “Padasya (VIII 2 16) t The notion “pada a Word padas 45 é Stem padas 47 2 The genitive Padasya 48 3 Combining with adhikara sttras 50 CHAPTER UT The logical principle ¥ Subject matter of the Tripadi § 8 A Sphere of application of the sutras contained in I-VI + 52 B Subject matter of the Tnpadi Restoration of vowels and consonants 54 IY Division of the subject matter § 9 A Sphere of application of the stitras 55 B Dhaviston of the subject matter The Tripadi asystem of groups 1 Formation groups 58 2 Position groups 59 3 Thematical groups 59 CHarrer IV The asiddhatva principle I The notion asiddhatva § 10 a Primary asiddhatva 63 B Secondary asiddhatya 65 Ui The asiddhatva as a principle § 11 A The asiddhatva a methodical principle 66 B Existence of asiddhatva 73 1 Formation group I 74 2 Formation group IL . @ Position group A 86 & Position group & 8 ¢ Position group C 94 ¢ Asiddhatva tacitly ignored 9CONTENTS Carter V Parallelism of the logical and the astddhatva principle I Parallelism of the logical and the asiddhatva principle 1p the progressions § 12 Tl Parallelism of the logical and the asiddhatva principle in the arranging of the groups § 13 CuarTer VI Systematscal upbuilding of the Trpadi I Sub groups § 14 A Formation group I y Ac division 2 Hal division a Yam groups 3 Jhal-groups 3 Acdivision 2 B Formation group II 1 Wa Avasana forms 2 lis Samhita forms @ Position group 4 5 Position group B ¢ Position group € 1 Hal diviston 2 Ac division Ti Higher groups § 15 CONCLUSION 1X Page 193 107 112 113 14 116 127 131 132 138 141 149 151 153PREFACE While laying before the public the English version of my Purvatrasrddham, 1 consider it a privilege to express my gratitude to all those Sanskntists who, whether directly, or by means of ther favourable criticisms in various penodicals, have furthered this publication Tt was, mdeed, a great satisfaction to me when, even before the Dutch edition was out of print, Professor C C Uurevnecs, Professor Racuu Vira, and other distin guished scholars, advised me to translate my treatise into an international language I regret all the more that, owing to various circumstances, the MS of this work, which had already been fimshed some years ago, could not until now go to press This English edition is not a mere translation of the Dutch one, but rather an (abridged) recast While, on the one hand, the treatment of the detatls has, in general, remained the same, I have, onthe other hand, not hesitated to make some alterations in the order and the arrangement of some of the subjects The reader who compares “The Tripad:® with Parvatrastddham wall observe that the present recast gave me the opportunity to revise and even amend the argumentation of certain passages which had not been satisfactorily expounded in Parvatrasiddham T should hke to acknowledge my indebtedness to my guru, Professor B Fappecov, of the University of Amsterdam, Professor SteN Koxow, of the University of Oslo, andXIL PREFACE others to whose kind suggestions I owe some of these improvements In conclusion I wish to express my sincere thanks to Mr Tu Fotxers, Manager of Messrs E J Britt, and his staff, for the pains they have taken in dealing with the techmical part of the work Leyden, July 1939 HEBSOURCES AND LITERATURE Ballantyne, LK. = Laghukaumud: A Sansknt Grammar by Varadaraya, with an English version, commentary and references, by James R. Ballantyne, LL D , Benares MDCCCXCI Bechtel, Hauptprobleme = Die Hauptprobleme der indogerman schen Lautlehre seit Schletcher, von Fritz Bechtel, Géttngen 1892 Belvalkar, Systems = An account of the different existing sy- stems of Sansknt Grammar, by Shmpad Krishna Belvalkar, M A, Ph D , Poona 1915 Boer, Oudnoorsch Handboek = Oudnoorsch Handboek, door Prof Dr R.C Boer, Haarlem 1920 —, Oergermaansch Handboek = Oergermaansch Handboeh, door Prof Dr R C Boer, Haarlem ro24 Béhtiingk, Pawn = Panmi's Grammatik, herausgegeben tiber- setzt, erlautert und mit verschiedenen Indices versehen von Otto Bohthingk, Lerpzig 1887 Brugmann Vergl Gr = Vergleichende Laut-, Stammbildungs und Flexionslehre der indogermanischen Sprachen von Karl Brugmann, . TI, 4, Strassburg 1906 Tf, 3, 1, Strassburg 1913 Delbriick, Emleitung = Bibhotheh indogermamtscher Sprachen, Band IV von B Delbriick, Letpaig 1919 DkP = Dhatupatha, ed Liebich, Zur Einfuhrung II Faddegon, Studies on Panim’s Grammar Verhandeling der Ko- ninklyke Akademue van Wetenschappen te Amsterdam Afdeeling Letterkunde. Nieune Reeks, Deel XX XVIII, N° 1 Noord Holland- sche Uitgeversmaatschappy Amsterdam 1936 —~, Mnemotechmcs = The mnemotechmecs of Pamm's Grammars, I The Crva Sitra Paper read at the 1st International Gongress of Linguists on the (3th of Apnf rgz8 by B Faddegon, Acta Onentaha VII GP = Ganapatha ed Bohthngh, Pann Gelpke, Anantabhatta = Anantabhatta’s Padartha prakiga, Em Kanya Kommentar zum Vajasaneyt Praticakhya , Gottingen 1929 Goldsticker, Pamm = Pann, His place in Sansknt Literature, by Theodor Goldstucker, London, Berfin MDCCCLXIKag. = Kagika, A Commentary on Pani: s grammatical aphorisms by Pandit Vamana and Jayaditya, edited by Pandit Bala Castri, Benares 1898 Kaeg1, RV = Der Rigveda, die alteste Literatur der Inder, von Adolf Kaegi Leipzig 1881 Kielhorn, Relation = Katyayana and Patafiyali their relation to each other and to Panim, by F Kielhorn, Ph D, Bombay 1876 —, Grammar = A Grammar of Sansknt by F Kielhorn, Ph D, Bombay 1880 Liebich, Konkordanz = Konkordanz Pann: Candra, von Dr Bru no Lrebich, Indische Forschungen, 6 Heft, Breslau 1928 —, Zur Einfuhrung = Zur Emfuhrung m die Indische einhet musche Sprachwissenschaft, Sitzungsberichte der Heidelberger Aka demie der Wissenschaften I Das Katantra, von Bruno Liebich , Heidelberg 1919 II Historische Emfuhrung und Dhatupitha, von Bruno Liebich, Heidelberg 1919 WI Der Dhatupatha von Bruno Liebich, He:delberg 1920 IV Analyse der Candra Vrttr, von Bruno Liebich , Heidelberg 1920 ——-, KsT = Ksiratarangini, Ksirasvamin’s Kommentar zu Pamm's Dhatupatha Zum ersten Mal herausgegeben von Dr Bruno Liebich, Indische Forschungen 8/9, Breslau 1930 ——, Zwer Kapitel = Zwei Kapitet der Kagika, tibersetzt und mit emer Eintertung versehen von Bruno Liebich, Dr phil , Breslau 1892 Macdonell, Vedic Grammar == Vedic Grammar by A A Macdo nell, Strassbourg 1910 Afbh = Vyaharana-Mahabhdsya of Patafiyalr, edited by F Kielhorn, PhD, Volt I Bombay 1880, Vol 11, Bombay 1906, Vol HI, Bombay 31909 Monier Williams, Grammar = A particular Grammar of the Sanshmt language by Sir M- Monrer Williams, Oxford 1877 Pbhl = The Paribhasenducekhara of Nagojibhatta, edited and ex- plamed by F Kielhorn, Ph D, Part I, The Sanskrit text and. various readings , Bombay 1868 Pbhl, te = Idem, Part II, Translation and Notes, Bombay 1874. PhS = Phit Siitra,ed SdhK., Vol IT, p 112—125, Renou, Grammaire = Lous Renou, Grammaire Sansknite, Tome I, Phonetique, Composition, Derivation, Paris 1930SOURCES AND LITERATURE XV Ries, Was 1st Syntax? = Was ist Syntax? Em kntischer Versuch von john Ries, Prag 1927 SdhK = The Siddhanta Kaumud: of Bhattoj: Diksita, edited and translated ite Enghsh, Vol =I, by Criga Chandra Vasu and Vamana Dasa Vasu, Alla- habad, Vol II, Part I, by (rica Chandra Vasu B A, Allahabad 1906, Part II by Gricga Chandra Vasu and Vamana Disa Vasu, Allahabad 1907, Vol U1, by (nica Chandra Vasu, Allahabad 1905 Sk6ld, Papers = Papers on Panim and Indian Grammar in general, by Hannes Skold, Lund, Leiparg 1926 Speyer, Analecta = Indologische Analecta, von J S Speyer, Zeit schrift der Deutschen Morgenlandischen Gesellschaft, Band LXIV, Leipzg 1910 Surya Kanta,XT = Rktantram, A Pratigikhya of the Simave da, cntically edited by Professor Sirya Kanta Shastry, MA MOL, Lahore 1933 Thieme, Panim and the Veda = Pammt and the Veda Studies in the Early History of Lingurstic Science in India Allahabad 1935 Thumb, Handbuch = Handbuch des Sanshnt mit Texten und Glos- sar, Eme Emfihrung im das sprachwissenschaftliche Studiam des Altindischen, von Dr Albert Thumb, Heidelberg 1905 UnS = Unnad: Siitra, ed SdhK, Vol II, Part W, p 147-333 Wackernagel, Altimdische Grammatik = Altindische Grammatik von Jacob Wackernagel, I Lautlehre, Gottingen 1896 Whitney, Grammar = A Sanskrit Grammar, including both the classical language, and the older dialects, of Veda and Brahmana, by Willham Dwight Whitney, Lerpzig 1879 ——-, Roots = The roots, verb-forms and pnmary derivatives of the Sansknt Language by Willam Dwight Whitney, Lespzig 1885 Winternitz, Geschichte = Geschichte der indischen Literatur, von Dr M Wintermtz Dnitter Band, Letpug 1922 Wundt, Volkerpsychologie = Vélkerpsychologie, Eine Unter- suchung der Entwicklungsgesetze von Sprache, Mythus und Sitte, son Wilhelm Wundt Zweiter Band, Die Sprache, Lespzig 1912INTRODUCTION CHAPTER I. THE ASTADHYAYI INTRODUCTORY OBSERVATIONS AND GENERAL ANALYSIS I GENERAL FORM AND CONTENTS OF THE ASTADHYAYI. STYLE INSTEAD OF SYSTEM? $2 The concise style of Panint’s sitras has been highly praised by Western and vernacular scholars!) The older Indian gram- manians, as 1s well known, rejoiced at the saving of even the length of half a short vowel as greatly as at the birth of a son 7} *Wober”, Kaegi says*), “noch daran zu erinnern ist, dass ohne etnen Sohn, der die Sterbeceremontien verrichtete, em Brahmane nicht glaubte, in den Himmel gelangen zu konnen” Belvalkar‘), Winternitz‘) and others freely call the szra- patha a complete grammar (See, however, Wackernagel, Altindische Grammatik, p LXX seq and Liebich, Zwet Kapttel, p XXXII seqq, and comp § 2.) Yet Panini ts blamed for having sacrificed system for the sake of style, or even for having unnecessarily neglected it in hts manner of formulating and arranging the sttras, Belvalkar, who nevertheless yudges ®) that “sf Panini seems to depart from this [2 ¢. a complete programme] in places 1t 1s more for conve- mience of treatment than for anything else”, regards?) the division of the samdks prakarana (under VI r and VII 2—g} as a lapse in regular logical sequence which he 1s inclined to ascribe to *Panini’s anxiety to secure a maximum of brevity”. *[Die] knappe[n], im schwierigen Sutrastl gehaltene(n] Regeln” §) are still more reyected by Wackernagel *Die Moghchkest, Worte zu ersparen”, he supposes"), “[méchte] [Spatere] dazu fubren,2 INTRODUCTION Regeln aus ihrem naturhchen Zusammenhang herauszureissen und anderswohin zu versetzen” and he even speaks') of a “fur Pfanini) egentumliche Art von Unordnung” Winternitz, too, thinks?) “dass dem ,Panini Kurze die Hauptsache [sei] und Satras uberall eingeschoben [wurden], wo sie sich der Kurze wegen am besten einfugen hessen, wenn auch dadurch oft nicht Zusammengehoriges neben einander gestellt [wurde]’ Many years before him already Whitney expressed’) his disapproval of “the highly artful and difficult form of about four thousand alge- braie formufa-like rufes in the statement and arrangement of which brevity alone 1s had in view, at the cost of distinctness and unambiguousness’ And Aufrecht spoke 4) of the “brevitas et obscuritas, quibus Panint eyusque schola studuerunt™” Statements and criticisms which show that opinions have very little changed since Colebrooke wrote 5) *The studied brevity of the Paniniya siitras renders them in the highest degree obscure, even with the knowledge of the key to their interpre- tation, the student finds them ambiguous In the application of them, when understood, he discovers many seeming contra- dictions, and with every exertion of practised memory, he must experience the utmost difficulty in combining rules dispersed in apparent confuston through different portions of Panina’s eight Lectures The apparent simpheity of the design vanishes tn the perplexity of the structure The endless pursuit of exceptions and limitations so disyoins the general precepts that the reader cannot keep in view their intended connexion and mutual rela~ tion He wanders in an intricate maze, and the clew of the labyrinth ts continually sitpping from his hands” Obs — With reference to the Sidd/tanta-Kaumid: and and similar works tt which a different arrangement of Panini’s sitras has been attempted, Colebroohe says") But the sitras of Panini, thus detached from their context ase wholly aaeatelligeble, sethout the com- mentatot’s exposttion, they are indeed, what Sir William Jones has somewhere termed them, “dark as the darkest oracle’ Likewise BOhtlingk?) “Spatere Grammatiken, Y) Abid, p 63 Note 2) Winternits,Geschebte, p 394 7) Whitney, Groat, p x 4) Liebreh, Zur Eafuhrong I, pr, Note *) Ballantyne, Lathe Koumudi, Preface, pot ser %) Ib pu 2) Dohtlingk, loc cit, Fintenung p xebCHAPTER I GENERAL OBSCRVATIONS 3 die Panins’s Stra aus der unverruchbaren Ordnung gebracht haben, um alles dem Stoffe nach Zusammen- gehdrige an cinander zu reshen, sind ohne ausfubriiche Commentare, die stets anf etwas weit Vorangegangenes oder Folgendes Rucksicht nehmen mussen, ganz unver- standiich und als Missgriffe zu betrachten”. If then, as late as 1887, 1n flat contradiction with all that has been quoted above, Bohtlingk judges!) ‘Die Anordnung der Sotras kann uns hier und da befremden, ist aber streng durch- dacht und in bewundernswurdiger Weise durchgefuhrt”, he 1s answered by Faddegon?) ‘in iis laudation Bohtlingk 1s more assertive than argumentative ... The reader who reads Panini for the first time, will be pa:nfully struck by the fact, that the text of the first chapters totally contradicts the assertion of Bohtlingk. Numerous references to later passages are already needed here and are mdeed given by Bohtlingk himself in Tus translation And even these references are not sufficient in number. Already at the second chapter of the first book the reader 1s overwhelmed by sts grammatical intncacies, and when he has finished the eight books, he only possesses a chaotic impression of details”. (See, however, Faddegon’s *Studies on Pantar's Grammar”, which result in a high apprectation } System ws obviously not unanimously ascribed to Panent The question arises whether Pansns really avns at shortness of expression onty or, though making unmistakable concessions to this, Starts nevertheless from a thorough mental planning of the subject mater The following chapters are to be Jooked upon as an attempt to answer this question, as far as the most remarkable part of the Astadhyayt, the Tripadt, 1s concerned Rem. — As appears from several data, Panini was not the first Indian grammarian Highly instructive are in this respect the recent statements of Surya Kanta Shastrt3) While, thus, on the one hand, Panini may have borrowed from predecessors, 1t 1s on the other hand no less probable that later grammarians should have made their interpolations in his work‘) As to this, all we know Id 7) Faddegon, Mnemotechnies Act Orient VII, p 48seqq 5) Surya Kanta Shastei,&2,pfarnn Comp also the author s Purvatrasdéhams, Dutch ed y P 6seq 4). Cf Skold, Papers, p 304 INTRODUCTION 1s that the Astadhyay? must have had tts present form already in the sth century A D') A terminus ad quem cannot be given Throughout the following pages since they do not concern the historical growth of the work, we shall speak of “Panini”, Il NATURE OF THE ASTADHYAYI THE INDIAN WORD $2 A VYAKARANA GABDANUGASANA Panini’s sasira 1s always and emphatically termed a vyakarana In view of the fact that in the pratgakhyas the complete word 1s the object of investigation, one is inclined to comprehend a vyakarana as a word analysis, Thus 1s, indeed, implied by its name wyakriyate 'nenets vyakaranam}, Gold stucker says?) “Vyakarana means ‘undoing’, + ¢ analysis, and Panini’s grammar is intended to be a linguistic analysis, it undoes words and undoes sentences which consist of words, it examines the component parts of a word, and therefore teaches us the properties of base and affix and all the linguistic pheno- mena connected with both, it examines the relation, in sentences, of one word to another, and hkewise unfolds all the Iingusstre phenomena which are inseparable from the meeting of words” Analysis, therefore, 1n its usual sense, “Synthetische Darstellung”, says Liebich 4), *bis zur logischen E:nheit, dem Satze”, and he believes that “man fragen konnte, ob nicht das panineische System in seinen Grundzugen noch heut mit Vorteil zu ser- wenden ware” The name 1s of no consequence. As regards the subject-matter the descriptive formulas of the Astidhyayt, save a few rare exceptions, contains the grammar of the word, by which, according to Ries’ clear statements), that part of grammar 1s to be understood which deals with both the internal and the external mflection of the word (taken in the sense as applied to 1t by Wundt %)) ) Liebich, Konkordanz, p 49 2) MBA on 4, t, Varet 12, Vol Tp 11 3) Goldstucker, Papim,p 196 ")Liebich, Zwei Kapitel, p xxau — ) Ruesy Was ust Syntax? p 67 — €) Wundt, Volkerpsychologie I, pr seqCHAPTER Y GENERAL OBSERVATIONS 5 The stsfadhyayt 1s therefore nghtly called ') a caddanusasana, @ grammar of the (sentence word as a symbol for a notion When Patafiyal: asks himself of what elements the word is composed, he finds?) that, what has been said, t22. that word == (A) word form (¢adda), (8) signification (artha), and (C) relation (sambandha), 1s nght This conception of the word notion, in skipping the Greek grammar, has been a pattern to the later European grammar and may, therefore, conveniently be judged from a modern standpoint’) Thus Panins distinguishes J. (For m) (A) auditive articulatory elements 1 Form of the root, 2 Form of the affixes (in the broad sense as adopted by Brugmanan ‘)), 3 Accen- tuation, 4 Mutual order of these elements, 5 (¢ ¢) Auxilaries, IT (Function) noumenal elements (ZB) Significations, (€) Relations How this 1s to be conceived, may be traced as follows B WORD ELEMENTS i Form Grammatically Panini distinguishes words as verbs and (primary and secondary) nouns (Thus already K atyadyana §)) Indeclinables are regarded as nouns with dropped case-endings Hence all words are exther verda fensta or casus (with possible zero-ending) Both form categories are called (word ) padas (cf § 7) Hence the question may be put in this way In what manner are these padas to be constructed out of the five elements mentioned above (J (4) 1-5)? 1 Roots By roots Panins understands ¢ (orginal or primary roots) the forms, deprived of anuéandhas, enumerated by him in the 3) MEL, pt 2) MBA T, p 6,1 14 seq Kathom punar stam ShagavatsA Paarner acaryarsa lohjanams pravrtiam ? (Varttska ) «Sidhe cabdirtha sambandhe }) Faddegon, Woorden Zn 4) Brugmann, Verg] Grammatk I, p 8 5) Vay Prat 1, 276 INTRODUCTION Dhatupatha, 2 (derived or secondary roots) formations which arise by the addition a) of the so-called suffixes of the secondary conjugation to primary roots according to III x § seqq, 22seqq, 25, 20d pt.seq 28 seqq , b) of certain suffixes to nominal formations (see Obs 5) 2. Pratyayas The name pratyaya 1s applied to the word parts (affixes) treated of mn adhyayas TI—V As appears from the manner in which 1t 1s indicated (I 7 46 seq), an affix 1s sometimes conceived as an augment, ¢ ¢ an increase, hence part of the preceding or following portion of the word Thus the subjunctive. and optative precative-signs are regarded as an increment prefixed to the personal endings {IM ¢.94 seq, 102, seq, 107) similarly the ¢ of *zéchat: (VI 2 73) (> wchat: VITI ¢ 40) 1s regarded as an addition to the of the root wh ( curtats > coriat: (VII 3 86) > coreats (VIL 3 84) >corayate (VI 1 78)} cur 1s (original) root, as well as stem (with regard to mzc), as well as part of the derived root curt (cori) On account of their relation to the root, the pratyayas may be divided into two groups x. One group 1s composed of such suffixes as, according to Panini, relate directly to the primary or secondary root They are @ those by whose affixing to an anga (primary root, pratz Padika, cast) secondary roots onginate(I17 5 —~33), 4. the wharanas (II z 33—90), which, as appears from the locative in the stitras in question, are placed before the suffixes sub ¢ and d, hence, facto, directly after the root, ¢ and d the a: and &rt suffixes, both of which suffixes (III r 96—end of the adhyaya) are put after the root in the sense of Ill 7 gt All these suffixes may conveniently be called root suffixes, in the face of 2 the other group which comprises all those suffixes which are separated from the root by a “root” suffix, since they are attached to a pratspad:ka or to a feminine stem the case, motion and faddheta suftixes (IV, V) Ideahter, 2 ¢ im disregarding syncopation or epenthesis pheno- mena (II 4 72, Ill 2 97, VI 3 67 seqq and the like) and apart from the Unnad: suffixes (cf Obs 2), we find, therefore, that the following suffixes partake of the formation of wordsCHAPTER J, GENERAL O8SERVATIONS. 9 7. Root-suffixes (if at all — c and d— sepa- rated from the root by a vékarara only). @, Suffixes of the secondary conjugation and denominative-suffixes. .. . . . HI x 5—32 4b. Vikararas (affixed directly to the root; see survey below}. . 2... - . 33—90 Kere-suffixes immediately after the root with the exception of cases such as IH 2 124 seqa. (¢ét-suffixes which are separated from the root by a zkarana) 96— 2 104, 106—~109 t24—end of the pada; 3 (1—3), 10-12, 14, 16—130, 158, 163, 167, 169 seqq., 1743 & 9—68, 70-~-76, e a. Tik-suffixes (directly attached to the root in the perf. and prec. only; in the other /a-karas preceded by a vika- rapa, which may be syncopated again 4 78—~112 (with mode-signs - assigned acc. to 92, 94 seq., 102 seq , (107). 2. Remote suffixes. ¢, Case-sufixes ... 6... “lL. Wa? J. Motion-suffixes . 2... 6. 2 we 4-81 & Taddhita-sufixes 2... 6... an 7- V 4 —Vend of the . adhyiya. Among the root-suffixes (group z) such as are separated from the root by a vikarana (the ti-suffixes — except the endings of the perfect and the precative —, and all those suffixes which are marked with the anudandha ¢ (III 4 113)} are distinguished from the others as sa@rvadhatska- and ardhkadhatska-suffixes respectively, A survey of the vikaranas, in the order in which they occur in Pagim’s work, yields the following aspect.10 INTRODUCTION zr drt and rn and duit sya and tas I r 33 2 det s (stp) 34 3 AL am 35-42 ¢ tun cha ssi) 44 b sa{hsa) 45—47 e a{can and an) 48—$9 d 2(ez2) (before medial te which is syncopated VI ¢ 104) 60-—66 (cf 8) 5 before a sartadhateke sufix denoting the dhava or the Aarman (t¢ in tm personal and passive verbs cf also 4d) xa(xak) 67 6 before a sarvedhatuka sufix denoting the agent 68—84. (Verbs 2 a(gap) regularly except in cases below b-h (cf Obs 8) « 1st and roth class (cf Obs 6) 68 70~—72 75 seq B 2nd class(ef Ig72 duk) 68 y 3rd class (cf Il ¢75 ¢éx) 68 b yalryan 4th class) 69-72 c nu(enu sth class) 73-76 82 d d(¢a 6th class) 7? e na(enam 7th class) 78 f (8th class) 79 80 § nalcna) ana (sanac) aya (yapac} gth class}, ef c 81—84) 7 aclr Im a{an) 86 & 1m reflexive verbs + (em) sa(sek) ya texan) 87-89 90 Obs 8 — In the Veda there 1s an interchange of vikaranas according to Ill r 85 3 Accent As a rule only one of the syllables of a nord bears theaccent (WI r 158) It usually falls as an ndatta exther on the first sHable of the suffix (IIX 7 3}or on the last sy Hable of the root (VI 7 162) in compounds it falls as a rule on the last syllable (VI 7 223)CHAPTER I GENERAL OBSERVATIONS i Obs 9 The accent 1s dealt with that of the autonome word m VE1158—z, end, in three places of the Tripadt and passim, that of the word t syntax: VIII x 18—end On Vedic intonation especrally I 2 34 seqq On Pantat's conception of accent see §9 Obs 6 4. Place The pratyaya always follows the anga (III z 2) As composita (II 7 3 seqq}, despite of their being accented according to VI ¢ 198, are considered by Indian grammarians as junctions of single words (ef however Obs 10) the As{adhpayi Must contain, besides a grammar of the word, a grammar of the composition Obs to — Compounds whose members are pure in flections arc, indeed, distinguished from those in which the mere stem occurs (whether or not modified secondarily) but the latter are considered as having dropped the ending by fuk (§ 7, Obs 2), accordingly such a member ts nevertheless treated as an onginal simplex — Composita with a finite verb as a last member are not treated as such by Panins (VIL g asamase'} The mutual arrangement of the members of the compositum 3s dealt with in II 2 30 seqq 5 Auxihanes Auxiliaries in the usual sense of the word are not mentioned im Paninis system In a wider sense loose words, as a means of expressing either temporal- or mood islations or both and several ufafadas (Bohtlingk, PSnint p 212°")) might be comprehended as such Such words are the anuprayogas tr (640, as), when put after the perfect formations in am (IL 7 35 seqq } the particle sma (IIE 2 118 seq, 122, 7 165, 176), wa in md Ohm, etc iT Functson To all the above forms noumenal functions correspond They are mentioned in each particular cas¢ and bear erther on the Semasiological notion of the word or on relations 1) For a deta led stady of the notion wpapats see Faddegon, Studet on Ps Cry} 3412 INTRODUCTION The former (the notions) are traced in the D#atupatha, the latter (the relations) are mentioned partly along with the forms — thus most of the suffix relations —, partly in separate places In the latter case the functions are summed up successively (thus among the grammatical categones ¢. g the fersonae verde and the case-relations) or, for hardly ever apparent reasons, in several scattered places (thus the — not yet separated! — modus and tempus- (/a-kara-) and some other relations Cf p 18 Obs 11, — According to I 2 56, one of the five siitras which Goldstucker called") the key stone of Panini's work, Panini would haye abstained from stating certain functions Although this stra might be an interpolation as well?), Liebich on the ground of the MBA I 31, Vartttha 2, states‘) that the DAatupatha must, indeed, originally have contaimed the diaius themselves only and that uno éenore. If such be the case, it must be assumed that Panini presupposes those functions to be known to the reader A special question referring to compostta exclusively, 1s the mutuaf relation between the members of composita and their relation to the whole These relations are described in I 7 6—2 29 Obs, 12 — Dionysios Thrax, as is well known, does not mention these relations at all 4) Il. METHODICAL PLANNING OF THE ASFADHYAYI ANALYSIS SYNTHESIS $3 A GENERAL PLAN The functions of the word-clements referred to above (§ 2) are gwen by Panini at the first opportunity. The forms themselves of these elements, however, and still more, the whole word forms, anse gradually, as Panini causes them to pass through several intermediate stages. Example. — Aéjavat consists provisionally of the dhate bhi (fanction and provisional form according to DAP. 11 Bk sattayam), the thematical vowel @ {function and form according to II} z 2, 68, I 79) and the personal ‘) GoldstGcker, Pagini, p 163,hole 7) Wackernagel, Alund Gramm I, PLXVIL,ote3 3) Liebich, Zur Ein! 11,§63 4) De/brucd, Eislettung, p &CHAPTER ¥ GENERAL OBSERVATIONS. 13 ending ¢ (function and form according to I 4 99, HI 7 2, 2 111, ¢ 78, 100) The 3rd p sing praet “*Shfart thus onginated, gets the augment « according to VI ¢ 71, with udatia according to the same sitra and VI 7 158 According to VII 7 84 uw becomes o and by VI 1 78 0 becomes av In this way adfazat evolves, in which form, further, 11 samtdhs a@ may become > zero (by VI 7 109} and in samdhz and also m pausa, ¢ may change to ¢ according to VIII 2 39, 4 $3 and 2 30, 4 §6 respectively Owing to this circumstance morphology occupies a large place in the sword description of the Astadhy a3, and thus the method applied to st by Panini becomes a characteristic feature of the whole work This method, which might be called a developing one,con sists in Panrnt’s provisionally concerving all words — inflection forms, § 2 — as considerably simphfied, wz all of them as verba finita, resp cases which have been built up according to a fixed prnciple Out of these *regular forms the real ones are gradually constructed im the same way as to some extent ts done an modern grammar too See instance above The simplifications referred to are mainly the following Tand Il Vocalism and Consonantism Accentuation 1s for the present neglected, vowel and conso nant samdhi are set aside ace sing vrkam =trka 4+ ant (suffix 4ak, UnS Ul 41, Panini WE ¢ 4, 1V 7 2, VI « 107), instr pl tagohih == tacékis (Vis 168, VI 2 30, 66, 3 15} The varsous vicissitudes caused by the alternation of vawels and consonants 1s done away with the vowel of the root usually shows the weak grade {but, for instance, rac by the side of ucyate'), the stem suffix guna (ragan with suffix kenin == an, UnS 1157) the final consonant of a root occurs im the form which st assumes tn the conjugation before vowels sem vowels and nasals (1aéts = vactt, cf tacmi) i Garloraing of coastrvction All words consist of fixed elements each in a fixed form A verbum finitum 1s composed of a root, possibly aztkarana anda personal ending, a casus 1s built up from a root, at feast one suffix (together with the root = stem) and a case ending In connexion herewith suffixes are tntroduced and removed, redu plication and augmentation are discarded (comp 1 $ perf4 INTRODUCTION tuthda << tudmi, 2 8 imper pdca < pacasr, str. pl dedit re Example devaR 67 contains two ritual terms and one mythological term, nomina- tives with A rn and the other as s>d group The junction 1s afforded by the sutras 72 and 73, which in the first progression do duty as negations, in the last as a rule B JUNCTION COMBINATION BREVITY OF EXPRESSION The concatenation just mentioned 1s very frequent in the Astidhzays Thus VI s, 223—2, end (accent of composita) 15 the connecting link between VI r 158—222 (accent of sumplicia) and VI 3 (larapade) With the former group 1t has in common that it bears on accentuation, with the latter that it treats of composita In order to bring about 2 concatenation between the first mentioned and the connecting group, VI r 223 has, to the neglect of the division into adijayas, been put in VI 7 {so as to give prominence to the anuvrtt: of anto (udatiah) (220) in 223) Sutra VUIz 108 shares the substitution of (for) + (z) with the preced ing sitra, the position with the following one The wpasargas pra ete (14 58,59) are mipatas like ca etc (57), but they are also gates