1.33

येषामर्थे काङ्क्षितं नो राज्यं भोगाः सुखानि च
त इमेऽवस्थिता युद्धे प्राणांस्त्यक्त्वा धनानि च

Printed: yeSHAmarthe kANkSHitaM no rAjyaM bhogAHa sukhAni cha , ta ime'vasthitA yuddhe prANNANstyaktvA dhanAni cha .

kANkSHitaM n ... [8.3.23] mo'nusvAraHa
rAjyaM bh ... [8.3.23] mo'nusvAraHa
ime'vasthitA ... [6.1.109] eNaHa padAntAdati
no r ... [6.1.114, 6.1.87, 1.3.2, 8.2.66] hashi cha
bhogAHa s ... [8.3.15, 1.3.2, 8.2.66] kharavasAnayo visarjanIyaHa
avasthitA yuddhe ... [8.3.17, 8.3.22, 1.3.2, 8.2.66] bho bhago agho apUrvasya yo'shi
prANNANstyaktvA ... [8.3.4, 8.3.34, 8.3.15, 1.3.2, 8.2.66, 8.3.7] anunAsikAt paro'nusvAraHa
yeSHAmarthe ... [6.1.72] saNhitAyAm
ta ime ... [6.1.78, 8.3.19] echo'yavAyAvaHa

Underlying: yeSHAm arthe kANkSHitam nas rAjyam bhogAs sukhAni cha , te ime avasthitAs yuddhe prANNAn tyaktvA dhanAni cha .

COMMENTS:

• This is a complex sentence to parse correctly, as we will show in the analysis below.
• Notice that two elided copula verbs ('is' and 'are' respectively) must be 'read in' by the parser into Clauses A.2 and A.3, in order for the parse-tree to be correct. The parser inserts the elided 'copula verb' ('as:2:P:to be') into these clauses.
• Notice that if 'kANkSHitam' is treated as a Predicative Adjective, then there is nothing wrong with its underlying form being NOM-S (it is very common for Predicative Adjectives to be in a NOM underlying form). In fact, it may be preferrable because the Neuter Gender would then match that of the subject 'rAjyam' (unlike the ACC-S option). The parser marks 'kANkSHitam' as a NOM-S Predicative Adjective (as it also inserts the elided 'copula verb' into the clause). In [1], this word is treated as ACC-S, but in [2], it is treated as a NOM-S Predicative Adjective, as does the parser. Hence this is not a defective assignment by the parser.
• Another point to note is that the Subject of Clause A.2 appears at first glance to be a Coordinated Noun Phrase ('rAjyam bhogAHA sukhAni cha' i.e. 'kingdom, enjoyment/ profits, and pleasures'), but the Predicative Adjective 'kANkSHitam' has a Singular underlying form, suggesting that 'bhogAHa sukhAni cha' may have an appositional relation. Notice the comment 'Link_rAjyam' in the entry for the passive past participle 'NOM-S:kANkSHitam:Neut.:desired'. Further, the term 'LOC-S:arthe:benefit' has the comment 'Link_kANkSHitam' in its entry, denoting it as a complement of the passive past participle (i.e. '[that is] desired in whosoever''s benefit').
• Notice also that the Correlative Pronoun 'te' occurs after its matching Relative Pronoun 'yeSHAm', but does not match it in Case (i.e. is not 'teSHAm'). It must also be kept in mind that 'te' also occurs very commonly as a regular pronoun. In other words, in sentences that have a relative pronoun as well as one or more occurrences of 'te', some of the occurrences of 'te' may not be correlative pronouns. In fact, it is common for a relative pronoun not to have a matching correlative pronoun, as we will see in other stanzas of this text.
• Clause A.3 has both a correlative Pronoun 'te' (with a NOM-P underlying form), as well as another unexpected NOM-P underlying form ('ime'); perhaps the simplest explanation for this may be that 'ime' is a redundant pleonastic form for emphasis i.e. 'those very these are stationed in battle'.
• The Correlative Pronoun 'te' has a large number of alternate underlying forms (including Masc/Fem/Neuter, Singular/Dual/Plural, NOM/ACC/DAT/GEN/VOC) that can cause a great deal of confusion for the parser. However, if the parser deduces the elided verbs correctly, as well as the matching Correlative Pronoun, it can deduce that the Number of the Relative Pronoun should be Plural, thereby leaving only the NOM-P and VOC-P underlying forms to choose from (and the VOC-P form is most unlikely for a Relative Pronoun).
• Similarly, the short form pronoun 'naHa' has many underlying forms (ACC/DAT/GEN and Masculine/Feminine) to choose from and the parser will require a lot more information to figure this out correctly. ACC forms can be ruled out if it is clear that 'kANkSHitam' is a Predicative Adjective, but it is difficult to choose between DAT and GEN. A lot of preceding context (from preceding lines in the text) will be needed to figure out its Gender correctly. For instance, it will be necessary to figure out the participants in the preceding context, and then correctly map them to the pronominal forms in subsequent sentences. In fact, the word 'naHa' ('us') also requires an understanding that Arjuna is a member of the Pandava army that includes most of his family members (a fact that is known to a human expert, but is not necessarily mentioned in the preceding text). This is beyond the capabilities of our parser. Note that [MM2015] [2] treats the word as GEN-P, as the parser does.
• The Locative 'yuddhe' has an adjunctive role in this sentence, hence the parser must decide which clause to attach it to (i.e. whether A.1 or A.3) without an 'understanding of the world'. Unfortunately, the word is positioned exactly on the boundary between clauses A.1 and A.3. [MM2015] attaches 'yuddhe:in battle' to clause A.3 (i.e. '... are stationed in battle, having sacrificed their lives and wealth'). The parser also attaches it to clause A.3, relying on the presence of a passive past participle (avasthitAHa: stationed) that has 'Active' force. However, in the absence of such a cue, the parser would normally use a simple heuristic to decide between the two clauses.
• Following the above analysis, we should read the sentence as below:
• 'Kingdom of ours (enjoyment/profits and pleasures too) [that is] desired in whosoever's benefit, those very these are stationed in battle, having sacrificed their lives and wealth.'
• However, please note that a reordering of phrases, such as the above, was found to be too error-prone to be done mechanically without an 'understanding of the world'. It is best left to a human expert.

---

A: yeSHAm arthe kANkSHitam naHa rAjyam bhogAHa sukhAni cha asti te ime avasthitAHa santi yuddhe prANNAn tyaktvA dhanAni cha

A.2:

yeSHAm:GEN-P:yad:Masc.:Pronoun

arthe:LOC-S:artha:Neut.:Noun:Link_gov_kANkSHitam

kANkSHitam:NOM-S:kANkSHita:Neut.:Adj:past_participle_passive_kta_1U_kANkSH:Link_subj_rAjyam

naHa:GEN-P:asmad:Masc.:Pronoun

rAjyam:NOM-S:rAjya:Neut.:Noun

bhogAHa:NOM-P:bhoga:Masc.:Noun

sukhAni:NOM-P:sukha:Neut.:Noun

cha:Indeclinable

A.3:

te:NOM-P:tad:Masc.:Pronoun

ime:NOM-P:idam:Masc.:Pronoun

avasthitAHa:NOM-P:avasthita:Masc.:Adj:past_participle_passive_kta_1P_ava-sthA:Link_subj_te

yuddhe:LOC-S:yuddha:Neut.:Noun:Link_gov_avasthitAHa

A.1:

prANNAn:ACC-P:prANNa:Masc.:Noun

tyaktvA:-:tyaj:1:P:VerbGerund

dhanAni:ACC-P:dhana:Neut.:Noun

cha:Indeclinable

References