In much of the scholarly discussions about moral philosophy and the "goals" of human life in ancient India, two Sanskrit expressions loom large: trivarga and purusartha. The first refers to a "triple set," a group of three concepts: dharma, artha, and kama, whose precise meanings are ambiguous and, as we will see, are defined differently by different authors. The second, often mistranslated as "aims of man," refers to the same set of three, sometimes with the addition of a fourth, moksa or final liberation. Even scholars who are careful to distinguish the two, giving logical and chronological priority to the former, do not take the next step to ask how that expression came to be transformed into or equated with purusartha. They assume that the two are synonyms and often use purusartha as a shorthand for these three (or four) concepts. (1) Second, there is a debate as to whether the concepts subsumed under these categories are, in fact, presented as goals to which human beings should aspire or simply a categorization of major human activities and pursuits (Davis 2004). This study attempts to answer these questions--very tentatively--using a philological scalpel to expose the complicated semantic history of these terms. It will show that many of the scholarly assumptions, with some notable exception, are, from a historical perspective, quite simply wrong.
THE INVENTION OF TRIVARGA
I take it as established that trivarga is the older and the more original formulation, and this study will further demonstrate it. The compound simply means "a group of three," and the group or set may consist of any three things in some way interrelated. This expression is quite old. The Samavidhana Brahmana (1.4.8 ) says prathamas trivargah, which, according to the commentator Sayana, refers to the three Samans originating from the verse agna a yahi vitaye (SV 1.1 = RV 6.16.10). In the context of a Vedic sacrifice, the Katyayana Srautasutra (8.6.11) uses the term to refer to the two sets of three roofs constructed on the northern portion of the sacrificial hut (sadas): trivargau cottaratah. Likewise, the Latyayana Srautasutra (4.12.8), after instructing the sacrificer to give a milk cow and so forth to various priests, says that he should give to the Udgatr priests all of the above made into trivargas (sarve trivargah), that is, each gift is tripled and the gifts are thus made into sets of threes. Clearly, in this early period of its usage, the term referred to any set of three objects.
The meaning of varga as a grouping or cluster is made clear by a statement in Kautilya's Arthasastra about official royal documents: "A varga should be made with a minimum of one and a maximum of three words so as not to create an impediment to meaning of the other words" (KAS 2.10.21: ekapadavaras tripadaparah parapadarthanuparodhena vargah karyah). As I have pointed out in my translation of this passage (Olivelle 2013: 525-26), here varga refers to a clustering of words in a document; after a varga there is a stop (virama) probably made by leaving a "white space" between one varga and the next (note that in ancient Indian inscriptions and manuscripts there are no white spaces between words; they run together). This is evident in the oldest documents we have from India, the Asokan inscriptions. As Klaus Janert (1973: 142-43) observes: "In the versions of the edicts under discussion spaces within the lines are frequent and occur particularly after groups of two or more words. It is my conclusion that this spacing can scarcely be anything other than a form of notation for pauses made during recitation of the edicts and which the scribes each recorded in this fashion." Each varga is expected to have words that are syntactically related and form a naturally meaningful unit.
The term varga thus can mean a category and could refer to any grouping of like objects, individuals, or concepts. Kautilya's Arthasastra, one of the earliest texts to use trivarga in its technical meaning with reference to dharma, artha, and kama, also uses varga to refer to other groups of three or more. In dealing with royal vices, for example, Kautilya (KAS 8.3.4; see also MDh 7.52) says that a group of three vices originate from wrath and a group of four originate from pleasure (kopajas trivargah kamajas caturvargah). Once again, no special or technical meanings are attached to either trivarga or caturvarga. Nevertheless, these usages reveal that these terms must have a referent, implicit or explicit, outside of them: a set of three makes little sense unless we know to which broader category these three refer. In the above examples, the implicit reference is vyasana, vices that afflict kings in a special way. We have an explicit referent in the extended compound satrusadvarga (group of six enemies) at KAS 1.6.11-12; 12.2.1. In all these expressions, varga refers to a group of similar things. So, at KAS 2.15.14-21 we have various kinds of sweets, salt, juices, spices, vegetables, and the like all referred to as varga (ksaravarga, lavanavarga, etc.); at KAS 2.15.63 types of slaves and laborers (dasakarmakaravarga); at KAS 9.6.56 upayacaturvarga (group of four strategies); and at KAS 2.17.4-12 the categories or types of various forest produce, plants, medicinal herbs, poisons, and the like (kupyavarga, venuvarga, vallivarga, valkavarga, ausadhavarga, visavarga). In dealing with seducible factions in an enemy's territory, Kautilya (KAS 1.14.2-5) refers to different categories of such people: kruddhavarga (angry people), bhitavarga (frightened people), lubdhavarga (greedy people), and manivarga (proud people). Likewise, in the Ramayana (2.40.2) we have suhrdvarga, people who are friends, and (2.73.17) silpivarga, those who are artisans.
It is within this context of the varga semantics that we must locate and understand the technical use of trivarga referring to dharma, artha, and kama. Given the other usages of varga, we should expect a priori that (1) these three are in some way similar, and (2) there is some other category or concept to which this particular trivarga refers, of which these three are subcategories. You simply cannot have a free-standing trivarga. A passage in the Arthasastra (9.7), which is closely followed by Vatsyayana in his Kamasutra (6.6), gives us an insight into the thinking of the Arthasastric tradition, in which this expression was probably coined, (2) with regard to the referent(s) of the trivarga. The context is Kautilya's discussion of a king preparing to march into battle. In such a situation, the king has to be cognizant of artha (advantages, benefits), anartha (disadvantages, material losses), and samsaya (doubt, uncertainty) with regard to things happening during the military expedition. An example of artha is capturing the rear enemy, and of anartha is giving troops and money to the neighbor of an enemy. But Kautilya knows that what is a seeming advantage may turn out to be a disadvantage, and vice versa: this is doubt.
That is the context within which Kautilya presents the threefold classification of these three possible scenarios (KAS 9.7.60-64):
artho dharmah kama ity arthatrivargah--"artha, dharma, and kama: that is the trivarga of artha." anartho 'dharmah s'oka ity anarthatrivargah--"anartha, adharma, and sorrow: that is the trivarga of anartha." artho 'nartha iti, dharmo 'dharma iti, kamah soka iti samsayatrivargah--"Is it artha or anartha? Is it dharma or adharmal Is it kama or sorrow? That is the trivarga of doubt." After each definition of the first two trivargas, Kautilya says: "Of that, it is better to encounter each preceding one than each following" (tasya purvah purvah sreyan upasampraptum), while after the trivarga of doubt he says: "Of that, it is better to encounter the first alternative after subduing the second" (tasyottarapaksasiddhau purvapaksah sreyan upasampraptum). A few points are worth noting here. The discussion is carried out within the most royal of a king's pursuits: waging war, underscoring the Arthasastric provenance of the trivarga doctrine. Second, the technical term trivarga is ascribed both to artha (what is beneficial) and to its opposite, anartha (what is detrimental). Third, in the enumeration of the three, artha occupies the first place, signaling its centrality within the trivarga doctrine of Kautilya, who states explicitly (6.7.61) that for the king it is better to encounter artha than dharma. Finally, the trivarga (artha, dharma, kama) is presented as subcategories of artha, presenting an interesting bifurcation of the meaning of artha. In the expression arthatrivarga, the term artha appears to signify an advantage in general or something beneficial and good, while within the trivarga itself, in conjunction with dharma and kama, the term refers more clearly to material, political, and/or military advantages. Another significant point in Kautilya's discussion is that the opposite--the anartha--of kama is not akama but soka: sorrow or grief. As we will presently see, at 1.7.3 the opposite of kama is given as nihsukha, the absence of sukha or pleasure, a concept very similar to soka. This conception of kama as the opposite of grief is interesting in light of the use of prlti (joy, pleasure) in Sahara's discussion of purusartha discussed below. These three are "similar" in that they all are "beneficial" (artha in the first sense) and conducive to a person's happiness (priti).
In a very interesting twist, the Kamasutra (6.6.5-6) cites the definitions of arthatrivarga and anarthatrivarga verbatim from the Arthasastra, and also deals with the topic of doubt (samsaya), all within the context of a courtesan and her activities.
Another discussion of the trivarga that gives us an insight into Kautilya's views occurs in KAS 1.7 in the context of secret tests (upadha) administered to ministers and high functionaries of the state to test their honesty and loyalty to the king. The king uses three kinds...