18. 60. Contrast Krishnamoorthy 1979b: 304.

60. Contrast Krishnamoorthy 1979b: 304.

61. KA 2.280. The rest of Dandin's illustrations are not so clear, though the commentator Ratnashrijnana certainly takes them this way.

62. A list of seventeen additional semantic figures of speech follows.

63. A citation from an unknown Mahabhārata play (a portion of the verse is interpolated after MBh 5.87.23; it possibly refers to Krishna's visit with Vidura in 5.89-90). Vidura is the wise half-brother of King Dhritarashtra and a well-wisher of the Pandavas and their counselor, Krishna, also known as Govinda. (The verse is cited also in KĀ 2.274, DhA 2, and many later treatises.)

64. On the example, see the headnote.

65. From an episode in the Mahabharata, where Karna's first shot at his enemy Arjuna misses; Shalya is Karna's (reluctant) charioteer. (Bhamaha conflates two passages, MBh 8.66.8 and 8.1104 *; I thank Y. Bronner for the reference.)

66. The three following tropes are provided only here, and ignored in KĀ.

67. The machinations of the jealous king Shishupala against Krishna are detailed in the Mahābhārata. The Ratnāharana

mentioned is not extant.

68. The Rajamitra is lost, and the episode mentioned is otherwise unknown.

69. Chanakya was the wily advisor of Chandragupta Maurya, and enemy of Nanda (the episode is not known in the extant literature). Moonstones ooze liquid when struck by moonbeams.

70. For the former, see Dimitrov 2014; for the latter, Pollock 2006: 344 (on p. 422 I misidentified him as a Jain; the incipit of his commentary shows him to be a Vaishnava).

71. Subsequent authorities will define the latter differently: for Bhoja, for example, the arthaguna mādhurya is when a character is described as showing gentleness even in anger (SP p. 502.16; Tarkabagisa on KĀ is wrong to say that this is merely the manifestation of the stable emotion, p. 49.7).

72. SKĀ 5.4-6 deals with rasa as sophisticated expression; "word rasa" is ignored.

73. KA 2.273. See Ratna's comment ad loc., and Bhoja's full exposition (ŚP p. 624.18-626.6; see also SKĀ p. 704.12).

74. ABh 1.266, where KĀ 2.279, 281 are adduced.

75. See Ratna's extended note on 2.279, below.

76. See chapter 2 headnote, and notes 17 and 73 there.

77. Presumably, their own poetry; and thus perhaps an echo of the "expression theory" of poetic creation (see the introduction, section 2), if the reference is to rasa taken generally.

78. priyatara. The examples below use the derived substantive prīti, which signifies both pleasure and affection.

79. yuktotkarsam. Here I follow the explanation of Ratna and Bhoja (contra Böhtlingk and Tarkabagisa ad loc., and Raghavan 1978: 420; compare KĀ 1.76: utkarșavān guņaḥ, "an intensified quality," and the Kannada Kavirājamārga 3.199, where the definition of ūrjita is given: ārūdhanijamanohamkārōtkarșaprakāśam ūrjitasadaļamkāram, "an intensification of the sense of self").

80. I read īyasunādibhir ... uktaḥ. yuktotkarșa iti pratyāmnāto 'bhivyaktyartham iti (with the Nepal ms., for ayam u[p]amādibhir ... ukto yuktokarşa iti). The "and so on" comprises the matup suffix in both rasavat and ūrjasvi.

81. BKA 3.5, above.

82. Ratna is correct in taking this verse as Dandin's explanation of the example, not a continuation of it. The following illustrative verse adheres to the same pattern.

83. The verse is addressed to Shiva, who is embodied in the eight forms listed.

84. A scion of the Raghu clan, according to Ratna (V reads ramavarmanah; the correct reading may, however, be rājavarmanaḥ,

referring to Narasimhavarman II, alias Rajasimha, very likely Dandin's patron).

85. The two examples are probably not meant to illustrate objective and subjective aspects of the "affectionate utterance" (how it affects the listener in the first example and the speaker in the second), since it is the speaker's emotion that should be at issue, as in all later discussions. It may well be that Dandin wanted simply to reference the two principal deities, Vishnu and Shiva. And given that both verses are addressed to a god, they clearly represent what in later discussions is called emotion poetry (first in Induraja on Udbhata 4.2, below), which typically concerns the stable emotion of desire for a god.

86. Bhoja provides an exegesis of Dandin in ŚP pp. 629-31, below.

87. Source unknown. King Udayana speaks in reference to his beloved, Vasavadatta, who had been thought dead but has just been revealed as present in disguise.

88. That is, the two male deities.

89. "Highly strengthened": (aty)upacaya, a term of art of the early commentators on the NS such as Bhatta Lollata.

90. "Inner principle": antastattva. I find no precise definition for the term elsewhere.

91. I read vācika- (with the Nepal ms., for nāțikā).

92. From a lost Mahabharata poem. Draupadi is the wife of the heroes, the Pandavas (one of whom is Bhima); Duhshasana is one of the antagonists, the sons of Dhritarashtra.

93. I read kāruņyākhyānaviśeșārayaḥ (with the Nepal ms., for kāruņyākhyarasaviśeșāśrayaḥ).

94. I read abhivyaktim (with the Nepal ms., for abhivyaktam). The term was not necessarily borrowed from Dhà (though Ratna knows the work); see above, n. 119.

95. I read mānāyā (with Vadijhangala and ŚP p. 679), for Ratna's mālāyā (where the girlfriend is amused at the nāyikā's boldness in the first days of her marriage, the uncrushed garland, māla, indicating the top position in lovemaking).

96. I read kāvyābhidheyasya vastunah (with the Nepal ms., for kāvyādheyavastunaḥ; correct the printed edition to sāgrāmyatā sabhyata; and in the final line, sambandhinī [sambandhina Nepal ms .; sambandhinah printed ed.]).

97. From a lost Ramayana play or poem.

98. Ravana is the demon king who abducted the wife of the hero, Rama.

99. This and the following example may well be Dandin's ad hoc creations; they are not found in extant literature.

100. Bronkhorst 2008.

101. "Only the small-minded ask whether one is kin or stranger. To the high-minded the whole world is family." The poem ascribed to Udbhata in Subhāșitavalī no. 498.

102. For Pratiharenduraja, see the colophon to his commentary; for Tilaka, Jacobi 1908: 291; Ramaswami Sastri (in his edition of KASS), pp. 39-41.

103. For a general reassessment of Udbhata, see Bronner forthcoming.

104. However implicitly in BKA 3.6.

105. The apparently redundant form preyasvat does not of course indicate ignorance of the analysis by which both the comparative (preyas) and the possessive (vat) signify atiśaya (see Ratnashrijnana on KĀ 2. 237, above), but its precise function is unclear to me (it may pertain to Udbhata's redefinition of the figure as referring to any emotion; thus, it seems, Krishnamoorthy 1979b: 306, and compare DhĀL p. 192).

106. Abhinava is the first we know of to add udaya, sandhi, and śabalatā to Udbhata's typology (DhĀL on 2.3). Bhoja will use different terminology (SKĀ 5.9-10).

107. A good example of Udbhata's combination of conservativism (honoring the old template of alankārśāstra) and innovation

(introducing a new affective one); see n. 152 below. I omit from the translation his verses on double entendre and two other rhetorical figures (n. 46).

108. KASS p. 54. If Induraja really thought Udbhata himself had resolved this question, he would have ascribed to him the verses he cites later anonymously (KASS pp. 80, 83; contrast Krishnamoorthy 1979b: 307). That Udbhata's framework of rasa analysis was taken as fully rhetorical was precisely what provoked the critique of Kuntaka (VJ 3.11-13).

109. For the early eighth-century playwright Bhavabhuti, see Pollock 2007: 42-44 (compare his verses cited in ŚP p. 624, below).

110. See "On Verbal Representation" (chapter 4.2).

111. By contrast Tilaka, the second commentator, inhabited a conceptual landscape radically discontinuous with this tradition (see for example his comment below, n. 112.

112. I omit from the translation below the disingenuous statement and double entendre.

113. The normal understanding of "emotion poetry" (see KP 4.35cd, and n. on Ekāvalī p. 103). The example (motherly love) supports Tilaka, though Udbhata's definition and Induraja's explication do not, and clearly represent an earlier conception.

114. Induraja lists them, citing NŚ 6.17-22 (with a v.l. that notably includes śama).

115. The reactions and the other three categories are usually said to intimate the eight (or nine) stable emotions (as below), not all fifty emotions (contrast Banhatti p. 98).

116. These are typically taken to be the four registers of acting, not four types of reactions. Yet note that the four kinds of anubhava, i.e., the arambhas, are correlated with the four kinds of abhinaya elsewhere (e.g., Agnipurana ch. 341). There may well be two parallel traditions, the one, abhinaya, referring to drama, the other, ärambha, referring to poetry.

117. The term visaya is usually used for the object of emotion, not as here āśraya, which refers to the subject.

118. sādhāranyena. This recalls Bhatta Nayaka, though for him the object of affect, not the affect itself, is "commonized" (Avaloka pp. 218-19, below); see also n. 122 below.

119. Udbhata cannot have known the theory of suggestion in aesthetics (Anandavardhana's innovation was some three generations later), though he might be said to have anticipated it.

120. svalaksaņasvabhāva-, like svalakșaņasvarūpa- later on, is not a dvandva (so Banhatti) but a bahuvrīhi. For Buddhists, language refers only to constructed universals, not unique particulars, but the idea is found elsewhere too (see Abhinava's IPVV v. 1, p. 117). And in Bhatta Mimamsa semantics, "proper terms" denote a universal.

121. I conjecture anubhava (suggested by L. McCrea, for anubhāva- in all printed editions).

122. sāmānyarūpatayā (or perhaps, in view of the preceding Buddhist terminology, "in the form of a constructed universal").

123. Note the absence of the stable emotion, which is required for the rasa to be clearly displayed and defines the figure as the "rasa-laden."

124. Source unknown, but probably from Udbhata's commentary on the NS (so with other Udbhata citations below). By "emotions" is presumably meant emotion poetry (see below); contrast the five different elements of rasa poetry (Induraja on 4.4, below).

125. Not all elements need be present; those absent are supplied by implication (thus many later thinkers, for example SD 3.16cd-17ab).

126. tayā (sc., ratyā) sāhacaryāt.

127. It is probably going too far to take this as "given that desire is the necessary accompaniment to all of them," and so seeing a prefiguration of Bhoja's view (ŚP p. 613; Bhoja seems to echo Induraja's account of preyah, see ŚP p. 625, below).

128. I read nirviśesa- (clearly indicated by P, so T and Jacob, for NSP, BSPS nirviśeșā-), and -vāllabhyāt (implied by P, so T, BSPS, and Jacob, for NSP -vātsalyāt).