Defining a Meṇḍaka Question in the Questions of Milinda and Its Commentarial Texts

Abstract

The word meṇḍaka, a derivative of meṇḍa (“ram”), is generally translated as “made of the ram” or “about the ram” or “horned.” However, in the Pāli Milindapañha (Questions of Milinda), the word meṇḍakapañha, literally, a question about the ram, is also rendered as a logical conclusion that refutes an imaginary dilemma. Hence, in this treatise, the word meṇḍaka is a special logical term which means an imaginary dilemma that can be logically refuted. This raises the question as to why the word meṇḍaka has come to be associated with this logical technique. To answer this question, this paper examines various aspects of the word and its possible connections to a dilemma and its refutation. The discussion ranges from the meaning of this word in a tale in the Jātaka (Birth Stories), within the contextual usage in a meṇḍaka question, to a relatively recent commentarial text (aṭṭhakathā) which gives a different perspective on the etymology of the word. The Milindapañha is explicit in defining a meṇḍaka question as knotty, hard to penetrate, and difficult to resolve, some of which an opponent puts forth to undermine certain aspects of the Buddhist system. However, the way certain meṇḍaka questions are framed, though not directly stated, seem to utilize the principles of logic in a dilemmatic form of argument. With that, a meṇḍaka question, at least in the Milindapañha, could also mean “a dilemmatic expression put forth by a challenger to undermine an opponent, but which can be logically refuted.”

Introduction

The Milindapañha (Questions of Milinda) is a Pāli text consisting of a series of dialogues purportedly between an Indo-Greek ruler, King Milinda (Menander I, r. 155/165–130 BCE) and a Buddhist monk, Nāgasena. The general format of each dialogue is that the king asks questions or puts forward his arguments, Nāgasena replies to him, and then the king accepts the answers. Each dialogue of this form is essentially a reasoning, which has the following four steps: (i) question (pañha) containing a thesis in doubt; (ii) inference (anumāna) that proves or refutes the thesis contained in the question; (iii) illustration (opamma) that demonstrates a connection between the thesis and the argument from the inference; (iv) accepting or rejecting the thesis from the question in accordance with the inference and its illustration. The event of these dialogues is said to have taken place in or near the city of Sāgala (Eastern Punjab) (Mil 1). These dialogues were written down approximately a century later around the beginning of the first millennium (Bodhi in Mendis ed., 1993, p. 2) by one or more unknown authors likely in North-West India, most probably in Gāndhārī (Salomon, 2018, p. 26). Thus, the Pāli text that we have today would be an adapted version.

The Milindapañha is one of the most popular and authoritative Pāli works not only in the Theravāda Buddhist world but also in the modern academic circles. Bhikkhu Bodhi regards the text as “a literary monument exhibiting intellectual acuity, subtle humour, inspirational fervour, and bold imagination” (Bodhi in Mendis ed., 1993, p. 13). In Burma (Myanmar), some authorities have raised the state of text to be that of the Discourses of the Buddha (Sutta) by including it in the Khuddaka Nikāya (Minor Collection) of the Sutta Piṭaka (the Basket of Discourses).¹

The text consists of several parts, namely, the Pubbayoga (Past History); the Milindapañha (Questions of Milinda); the Lakkhaṇapañha (Questions on Distinguishing Marks); the Meṇḍakapañha (Questions about the Meṇḍa); the Anumānapañha (Questions [solved by] Inference); and the Opammakathāpañha (Questions on Talk of Similes) (Mil 2). The part on the Meṇḍakapañha is by far the largest section, constituting slightly more than half of the text.² It comes with more than eighty dialogues³ which revolve around puzzles, or knotty issues within the Buddhist system that are difficult to resolve. For example, the inconsistencies within the literature where statements attributed to the Buddha cannot be both correct apparently.

The compound ‘meṇḍakapañha’ is generally translated as “question (pañha) about the ram (meṇḍa)” with the connotation “horned question.” Alternatively, perhaps due to the nature of these questions or how they are presented by the king, meṇḍakapañha is also rendered as a ‘dilemma’ by modern translators (see below). As such, questions have been raised as to why the word meṇḍaka relates to the dilemma; is there any etymological link to it or is it thus called through inference. In order to answer these questions, here, we will start off by exploring how logic and literature define these words; meṇḍa, meṇḍaka, and dilemma, and in what way the “question about the ram” or “horned question” links to the dilemma by alluding to a tale in the Jātaka (Stories of the Buddha’s Former Births) which features cooperation between a ram and a dog giving rise to a puzzling question. We will then take a look at how the Milindapañha and its commentarial texts describe the nature and the characteristic of a meṇḍaka question. We will also demonstrate, through a few examples, the manner in which meṇḍaka questions are framed and put forth by the king, which seems to utilize the principles of logic in dilemmatic forms of argument, as well as the method deployed by Nāgasena in answering them. Lastly, we will examine one of the Pāli commentarial texts to the Milindapañha, composed in Burma in 1940 by a scholar monk, in which he expresses his view on the etymology of the word meṇḍaka. An English translation of this section of the monk’s composition is included in the “Appendix” of this article.

Direct Meanings of Meṇḍa

Childers (1875) defines meṇḍa as ‘ram’ or ‘groom.’ The Pali-English Dictionary (PED) renders meṇḍa as ‘ram’ cross-referencing the Dhātumañjusā (Dhtm) with the root word of meṇḍ (meḍ) carrying the meaning of ‘koṭilla’ which is crookedness.⁴ The PED has the following entry for meṇḍaka, a derivative of meṇḍa, as either “made of ram(s) horn,” or “belonging to a ram,” as in meṇḍaka-pañha; “question about the ram.” Edgerton (1953) put in an entry to the Buddhist Hybrid Sanskrit Dictionary for meṇḍa-viṣṇikā as “a ram’s horn.” Monier Williams (1899) defines Sanskrit meṇḍha, meṇḍhra or meḍhra as a ram, while meṇḍa is cross-referenced to meṭha, which is also a ram or an elephant keeper. The Dictionary of Pāli Proper Names (DPPN) has meṇḍaka as the name of a very rich householder of the city of Bhaddiya in Aṅga,⁵ the grandfather of Visākhā. He was one of the five treasurers of King Bimbisāra. He was called ‘ram’ because behind his house, some golden rams as big as elephants, horses, or bulls, pranced up and down hoofing the earth, smiting each other back to back (Dhp-a III 364; Paṭis-a II–III 679).

In Pāli literature, for example, the Jātaka, meṇḍa is defined as one who has twisted horns (Ja VI 354).⁶ A twelfth-century CE grammatical work, the Saddanīti, composed in Pagan, Burma, defines meṇḍa as crooked horns of an eḷaka (a goat or a ram) (Sadd II 345).⁷ According to the commentary to the Theragāthā (Verses of the Elders), an arahant monk is called Meṇḍa-sira because his head resembled that of a ram (Th-a I 182). Fǎxiǎn (法顯), a fourth-century CE Chinese monk who travelled to India, translated meṇḍa (from the name meṇḍaka) as a ram (羊), and similarly, a fifth-century CE Indian translator, Dharmakṣema, also translated meṇḍa into Chinese as a (black) ram (羖羝).⁸ From here, we could see that the word ‘meṇḍa’ is associated with a ram or a ram with twisted horns.

Modern Translations of Meṇḍa

The editor of the Pāli Milindapañho (Mil), Trenckner, states: “meṇḍaka-pañha, a puzzling question, no doubt alludes to the story about a ram which forms part of the Ummagga-Jātaka and is thus entitled” (Mil 422–423). Horner, the translator of the Milindapañho, adds: “meṇḍaka-pañha, question belonging to the ram or questions made of ram’s horns, thus dilemmas” (Horner, 1963, p. 3 n.4).

To understand how far the story related to the ram can be described as puzzling or even a dilemma, let us take a look at the brief account of what the Pāli Ummagga-Jātaka (No. 546) entails as follow:

On another day, the king (King Vedeha of Mithilā) was strolling on his long walkway after breakfast when he saw through a window in the gateway a goat (eḷaka) and a dog (sunakha) exchanging friendly greetings with one another… he thought, “I have never seen this happen before. These two enemies are living together in harmony. I will take this question and ask the wise men, and the one who cannot find the answer I will drive from the kingdom” (Appleton & Shaw transl., 2015).

Hence, the king saw an unusual event “a goat and a dog [are observed together]” that needed an explanation, because it should logically be “either a goat or a dog [is observed at one place]” (for example, only a dog, only a goat, two dogs or two goats, but not a dog and a goat together), since both cannot be together, but they are. So, we do not have a disjunctive statement here, although it was expected. The next day the king poses his question to four sages and Bodhisatta Mahosadha⁹ by uttering this verse:

There are two enemies who never in this world go within seven paces of each other,

Who live now in accord, the greatest of friends, what is the reason?

The king gives them one day to ponder over the reason. On the following day, the king asks first the four sages who provided him with partial answers. The king then turns to the Bodhisatta who replies:

Great King, who else, from the Avīci hell to the highest heaven, would know the answer?

And he proceeds to answer the king’s question with this couplet of verses:

The ram (meṇḍa),¹⁰ with eight pads on his four feet, and eight hooves,

Unobserved, brings meat for one, and he (dog) brings grass for him.

In his own palace the Videhan lord saw, it seems,

The exchange of food between one another,

Between the one who goes “baa-baa” and the one who goes “woof-woof” (pp. 232–237).

The other four sages could only provide partial answers to the question, but the king assumes they might have known the answer. Even then, it is not without the help of the Bodhisatta in private. The king is delighted with their answers and bestows upon each of them a carriage full of she-mules and a prosperous village. The Bodhisatta knows the solution to the puzzle, because he sees for himself the cooperation of the two animals from the same walkway as the king notices the event unfold, and upon further investigation, finds out the reason behind it.¹¹ As such, the question related to the ram in this tale is a puzzle that involves what appears to be a union of contradictions, as demonstrated in most meṇḍaka questions. And it is only difficult to those who do not have an insight into it, and thus they operate on the premises of this syllogism in the Aristotelian way:

Enemies do not live in harmony (major premise).

Goats (or rams) and dogs are enemies (minor premise).

Therefore, a goat (or a ram) and a dog do not live in harmony (conclusion).

Thus, an unusual event is observed by the king such as “a goat (ram) and a dog [are observed together]” and from this our hypotheses that “either a goat (ram) or a dog [can be observed at one place]” seems to be false. In propositional logic:

It is true that a goat (ram) and a dog.

Therefore, it is false that either a goat (ram) or a dog.

This reasoning is logically correct. Symbolically:

It is true that A and B (symbolically: “A & B”).

Therefore, it is false that either A or B (symbolically: “¬ (A ⊕ B)”).

(A & B) ⇒ ¬ (A ⊕ B).

Here, it is evident that we do not have a dilemma “A ⊕ B”, but its refutation through the observation of “A & B”. Then a puzzling question should mean something which is refuted as an imaginary dilemma.

In the second edition of the devanāgarī Pāli Milindapañho, the editor M. S. Bhat, listed the compound meṇḍaka-pañhaṃ in the glossary as: “literally questions belonging to the ram, that is to say, shaped like a ram’s horns, hence dilemmas” (1972, p. 390), although the story about the goat (ram) and the dog is not a dilemma in the narrow sense, but its refutation. This entry is aligned with Trenckner’s and Horner’s, but it does not go further to describe why ram’s horns can be translated as a dilemma. Perhaps, we could get a clue from the way the king describes his question as ‘ubhato-koṭika,’ which Horner translated as “a point at both ends.” And she added that; “it is, in fact, a dilemma, a question having a ‘horn’ both ways” (Mil 95; Horner, 1963, p. 133 and n.3).

Meṇḍaka as a Logical Term

In the Milindapañha the term ‘meṇḍakapañha’ (“horned question”) means a question “Does either A or B take place at once” (“A ⊕ B”) in which it is hard to define what is true from either A or B.¹² Therefore, it is translated as a dilemma by Trenckner and Horner. The Oxford Dictionary defines it as a situation in which a difficult choice has to be made between two or more alternatives, especially ones that are equally undesirable. In logic, a dilemma is “an argumentative device in which <…> the premises of the syllogisms so combined are formulated disjunctively, and devised in a way designed to trap the opponent by forcing him to accept one or the other of the disjuncts” (Copi et al., 2014, pp. 288–289).¹³ In an argument, a dilemma could be framed into a fallacious dichotomy to undermine an opponent.

One of the earliest dilemmas well known in Ancient Greece is called “being horned” (κερατίνης): “Have you ceased to be horned? Yes or no?” According to the principle of the excluded-middle (tertium non datur), each statement is either true or false. Therefore, the statement “You have ceased to be horned” is either true or false, too. But both possible answers (“He has ceased to be horned” and “He has not ceased to be horned”) seem to presuppose that the respondent was surely horned in past. The first formulation of this paradox belongs to Eubulides, the Megaric philosopher: “If you did not lose a thing, then you have it. You never lost horns; therefore, you are horned” (εἴ τι οὐκ ἀπέβαλες, τοῦτο ἔχεις κέρατα δὲ οὐκ ἀπέβαλες κέρατα ἄρα ἔχεις) (Diogenes Laertius, Lives of Eminent Philosophers VII, 7).

The Stoics solved this paradox by introducing negation in relation to the entire proposition by denying the existence of the subject (Schulthess, 1996). So, we cannot take the statement “A ⊕ B”, as it is false. Instead of that we have the statement “¬ (A ⊕ B)”. Therefore, the dilemma “being horned” is imaginary for the Stoics. Furthermore, each dilemma is imaginary and sophistic for them¹⁴. Later, the name ‘horned’ (κερατίνης) became a household name for any dilemma which is always false. For instance, this term along with mentioning the Stoics occurs several times in Lucian’s works: Hermotimus or Concerning the Sects (Ἑρμότιμος ἢ Περὶ Αἱρέσεων / Hermotimus) 81; The Carousal Symposium or The Lapiths (Συμπόσιον ἢ Λαπίθαι / Symposium) 23. As a result, each logical technique of denying the dilemma became to be named ‘horned’ (κερατίνης, κερατίτης) or “horned reasoning” (κερατίτης λόγος) and “horned syllogism” (κερατίνης συλλογισμός). In Scholastic logic it was called “horned syllogism” (syllogismus cornutus), too.

In the Cyclopædia: or, An Universal Dictionary of Arts and Sciences written by Ephraim Chambers (1750), the first edition in 1728, the term ‘dilemma’ (δίλημμα, “double proposition”) is defined as follows:

A Dilemma is an argument consisting of two contrary parts, or sides, either of which catches the adversary. And hence it is also called syllogismus cornutus, a horned syllogism, its horns being so disposed, that if you avoid the one, you run upon the other.

Immanuel Kant (1992) defines the dilemma thus:

A dilemma is actually a ratiocinium hypotheticum [“conditional rational inference”], whose consequentia [‘consequent’] is disjunctive. Let one proceed in this case according to the rules of the hypothetical as well as those of the disjunctive mode of inference. Atqui falsum est prius, ergo etiam posterius [“But the former is false, therefore also the latter”]. One never has a dilemma as a modus ponens, but always as a modus tollens. By means of a dilemma we do not seek to prove a proposition properly, but instead only to convince someone of its falsehood. The dilemma is also called a cornutus [‘horned’], because by means of it one refutes another and also at the same time cuts off all paths[;] he may take whichever he will, still he is caught (Kant, 1992, p. 231).

A dilemma is an inference of reason that is composed of a hypothetical and a disjunctive inference, where the major is a hypothetical proposition, whose consequent, however, is a disjunctive judgment[.] E.g., If this world is not the best, then God was not able to create a better one or he did not want to. Now I infer a remotione consequentis ad remotionem antecedents [“from the removal of the consequent to the removal of the antecedents”]. I reflect on the consequens per omnia membra [“consequent through all the members”]. I say: He could have created a better world, for God is omnipotent[;] he wanted to create a better world, for he is good[;] consequently it is false that this world is not the best. A dilemma is thus a hypothetical inference, for the major propositio [“major proposition”] contains a hypothesis, but the consequens of this proposition is disjunctive and has to be negated per omnia membra [“through all the members”]. The ancients made much of this dilemma and called it a horned inference, syllogismus cornutus. They commonly said, then, if you maintain this, then you must also maintain this and that. But these propositions are false, consequently what you maintain is also false[;] all the ways you can go are false, consequently you have no way to go (Kant, 1992, p. 406).

Let us assume that the sentence “Either A or B” (“A ∨ B”) is a dilemma¹⁵. Then, according to Kant who continues the Stoic reasoning, in fact we deal with the implication “If there is a hypothesis H, then either A or B” (“H ⇒ (A ∨ B)”). Then we deny the dilemma: “It is false that either A or B” (“¬ (A ∨ B)”). From this it follows by modus tollens that H is also false (“¬ H”). Symbolically:

H ⇒ (A ∨ B) – major premise;

¬ (A ∨ B) – minor premise;

therefore, ¬ H – conclusion.

Kant’s example of this “horned reasoning:”

If “This world is not the best” (H), then “God was not able to create a better one” (A) or “He did not want to” (B). But “He could have created a better world” (“¬ A”) and “He wanted to create a better world” (“¬ B”). Consequently, “It is false that this world is not the best” (“¬ H”).

In this reasoning, we deny the disjunction “A or B” (“A ∨ B”) by setting the conjunction of the negations “not-A and not-B” (“¬ A & ¬ B”). It is correct from the point of view of propositional logic, because “¬ (A ∨ B)” is equivalent to “¬ A & ¬ B”. At the end, from “¬ (A ∨ B)” we obtain “¬ H”.

There is also another scheme of syllogismus cornutus that was not mentioned by Kant:

(A ∨ B) ⇒ H – major premise;

¬ H – minor premise;

therefore, ¬ (A ∨ B) – conclusion.

We apply here modus tollens again. It is worth noting that from “¬ (A ∨ B)” we entail “¬ (A ⊕ B)”.

In the Milindapañha the term ‘meṇḍaka’ also means a logical technique for refuting each dilemma and this techique is the same as in the Stoic or Scholastic logic. The point is that for the author(s) of this treatise each dilemma is imaginary to the same extent as it was for the Stoics. So, the Greek analogue for the technical term ‘meṇḍaka’ is represented by κερατίνης or κερατίτης with the same direct meaning and the same connotation as a logical term that if one faces a dilemma where a choice must be made between two options and no matter which option is chosen, the chooser would be impaled by one of the ram’s horns, but we can refute this problem logically. In this way, we may assume that the term ‘meṇḍaka’ can be a direct translation of the Greek term ‘horned’ (κερατίνης) with some other modifications of it, such as “horned reasoning” (κερατίτης λόγος) or “horned syllogism” (κερατίνης συλλογισμός), made into Pāli.

Logical Scheme of a Meṇḍaka Question

Let us look at two examples given in the Meṇḍaka-pañha section of the text to reconstruct an appropriate logical scheme of meṇḍaka to compare it with the Stoic syllogismus cornutus. In the first example, King Milinda asks Nāgasena which of these is correct: “If the Buddha accepts homage (A), then the Buddha is not completely liberated (na parinibbuto) (“¬ B”).¹⁶ He is still tied by the world, is included and similar to the world of living beings. Therefore, an act of merit rendered to Him is empty and does not bear fruit [for the doer] (H);”¹⁷ or “If He is completely liberated (B), untied by the world, dismissed from all forms of existence, then homage for Him does not arise. One who is completely liberated does not accept anything (“¬ A”). Therefore an act of merit to the Buddha, who is not accepting, is empty and does not bear fruit [for the doer] (H).”¹⁸

Thus, if Nāgasena chooses the first option, then it means the Buddha is not fully liberated from the cycle of rebirth (“¬ B”), which is contrary to the textual tradition and, consequently, paying homage to Him is empty and does not have fruit (H). On the other hand, if he selects the second option, then why is there a need to pay homage to the Buddha, who is not accepting (“¬ A”)? Thus, the practice of paying homage to the Buddha by thousands, if not hundreds of thousands of people, is again rendered meaningless (H). In other words, the king’s argument is as follows: “Either the Buddha is not fully liberated (“¬ B”) or He does not accept homage (“¬ A”)”, but “If He is not fully liberated (“¬ B”), then paying homage to Him is empty (H)” and “If He does not accept homage (“¬ A”), then paying homage to Him is again empty (H).” This puzzle (pañha) is called ‘double-horned’ (ubhatokoṭika) by the king:

“If either not-B or not-A, then H.”

(¬ B ∨ ¬ A) ⇒ H.

This statement is logically equivalent to the following one:

“If not-B, then H and if not-A, then H.”

(¬ B ⇒ H) & (¬ A ⇒ H).

It is the pañha, the first step in logical conclusion as it is being constructed in the Meṇḍaka-pañha. Also, it is the major premise of the syllogismus cornutus.

As we can see, the design of a meṇḍaka question is presented in a dichotomy fashion, as “a point at both ends” (ubhatokoṭika) from which we entail that paying homage to the Buddha is meaningless in all cases: “Either the Buddha is not liberated (“¬ B”) or He does not accept paying homage (“¬ A”).” However, this implication “(¬ B ∨ ¬ A) ⇒ H” is built on a shared assumption that “homage is paid only to those who can receive it.” Nāgasena’s answer is that this sentence is not entirely valid; because even though the Buddha is not around to accept homage, paying homage to Him could be part of the conditions to support the practice of the teaching the Buddha left behind. Therefore, for those who do that, the homage paid to the Blessed One, even though He is not around to accept it, is not empty but bears fruits for them. The construction of this syllogism by the king is similar to the earlier example in the Jātaka, where a dog and a ram seem to be a dilemma, but they can live together in harmony, so does paying homage and fully liberating that can be always together.

It is the second step of reasoning, called ‘inference’ (anumāna). Also, it is the minor premise of the syllogismus cornutus:

“B and not-H.”

B & ¬ H.

From this we have “not-H” (“¬ H”). Then this premise “¬ H” is illustrated by an example. It is the third step in reasoning, called ‘illustration’ (opamma): “Fire is an unconscious thing and it can burn or not, but it does not accept fuel. Nevertheless, fire is not meaningless and always possible.”

From “¬ H” and its illustration (opamma) Nāgasena infers the negation of the dilemma by modus tollens:

“It is false that either not-B or not-A.”

¬ (¬ B ∨ ¬ A).

Notice that from “¬ (¬B ∨ ¬ A)” we entail “¬ (¬ B ⊕ ¬ A)”, too. It is the fourth step in reasoning, its finish.

To sum up, we see the following conclusion (called meṇḍaka) in short:

• (1) Pañha (major premise): (¬ B ∨ ¬ A) ⇒ H. It can also be formulated as follows: (¬ B ⊕ ¬ A) ⇒ H.

• (2) Anumāna (minor premise): ¬ H.

• (3) Opamma (illustration): there exists something like ¬ H.

• (4) Conclusion by modus tollens from (1) and (2): ¬ (¬ B ∨ ¬ A). It can also be formulated as follows: ¬ (¬ B ⊕ ¬ A).

The difference from the Stoic syllogismus cornutus consists only in the requirement to have an illustration (opamma)—the third step in inference.

The second example of a meṇḍaka question is when the king asks Nāgasena why the Buddha refused to answer certain questions, for example, those questions posed by Elder Māluṅkyāputta.¹⁹ The king argues that it is either the Buddha is ignorant (A) to the answer or He is unwilling to share his wisdom (B)²⁰ (since it is claimed that the Buddha is omniscient²¹) (Si 199, Mil 144).²² In the form of two implications: “If the Buddha is not omniscient (A), then He does not reply to the question (H)” and “If He is unwilling to share his wisdom (B), then He does not reply to the question (H).” Hence, the king’s puzzle is constructed as below:

“If either A or B, then H.”

(A ∨ B) ⇒ H.

It is logically equivalent to the following sentence:

“If A, then H and if B, then H.”

(A ⇒ H) & (B ⇒ H).

In this dialogue, Nāgasena asserts that there are questions which can be put aside (ṭhapanīyo), such as “Is the universe eternal?” (asassato lokoti) or “Is the soul different from the body?” (aññaṃ jīvaṃ aññaṃ sarīranti) or “Does the Blessed One exist after death?” (hoti tathāgato paraṃ maraṇāti). The point is that “there is no reason or cause to answer it” (na tassa dīpanāya hetu vā kāraṇaṃ vā atthi). That is why they should be put aside, since they are not questions in the narrow sense—now, they are called rhetorical questions. Thus, the Buddha answers all not-rhetorical questions: “¬ H”. From this it follows by modus tollens that “¬ (A ∨ B)”.

The steps of this reasoning are as follows:

• (1) Pañha (major premise): (A ∨ B) ⇒ H. It can also be formulated as follows: (A ⊕ B) ⇒ H.

• (2) Anumāna (minor premise): ¬ H.

• (3) Conclusion by modus tollens from (1) and (2): ¬ (A ∨ B). It can also be formulated as follows: ¬ (A ⊕ B).

As we see, Nāgasena did not illustrate this conclusion through the procedure of opamma. So, this inference is completely identical to the Stoic syllogismus cornutus, according to its scheme, although we should remember that the opamma is mainly a necessary step in reasoning of Nāgasena. Hence, the meṇḍaka in comparison to syllogismus cornutus usually assumes an additional step to exemplify the minor premise.

All the applications of meṇḍaka from this treatise are to show that each dilemma is only imaginary and, therefore, can be logically refuted. In Ancient Greece, the same idea was shared only by the Stoics. Other philosophers accepted the existence of dilemmas and one of them was Aristotle who first formulated the dilemma of equally hungry and thirsty humans: “Or else of the extremely hungry and thirsty man, but both equally, and at an equal distance from food and drink, will be compelled to remain where he is” (καὶ τοῦ πεινῶντος καὶ διψῶντος σφόδρα μέν, ὁμοίως δέ, καὶ τῶν ἐδωδίμων καὶ ποτῶν ἴσον ἀπέχοντος· καὶ γὰρ τοῦτον ἠρεμεῖν ἀναγκαῖον; On the Heavens / Περὶ οὐρανοῦ / De Caelo 295b). Then this dilemma became to be called Buridan’s ass in the Scholastic logic: “An ass cannot choose either food or drink.” For the Stoics, such dilemmas do not exist. They can be always denied by modus tollens—to the same extent as Nāgasena did it. The matter is that this teacher of Milinda supported the Stoic idea of logical determinism (Schumann, 2021, 2023). Everything has its cause and inconsistencies are impossible. If I cannot choose one statement from the exclusive disjunction, then this exclusive disjunction is false and must be rejected.

In the two examples of meṇḍaka considered above, the king’s questions containing a dilemma sound first reasonable, but this dilemma was refuted then by applying the logical scheme of syllogismus cornutus. The two alternatives the king presented in each argument are unfavourable. In response to these questions, Nāgasena does not flinch and calmly “going between the horns” to dismiss the shared premise that is used to negate the dilemma.²³

It is worth noting that these conclusions by Nāgasena are absolutely correct from the point of view of modern propositional logic and correspond to some schemes which were first proposed by the Stoics. In the opening verses of the section of meṇḍakapañha, the king is being described as one who is experienced in debate; an expert in reasoning (vetaṇḍī); extra intelligent, and has the ability to discern (Mil 90).²⁴ The word vetaṇḍī could mean a sophist, an eristic, or one who is an opponent of a particular school. Rhys Davids translated the word vetaṇḍī as a sophist, or one who is skilled in sophistry (vitaṇḍa) (1890, p. 137).²⁵ However, a commentarial text to the Milindapañha has it as representing the opponents of the Theravāda school.²⁶ The king himself indicates that his arguments may have come from holders of other tenets (paravādin) (see below). Nevertheless, it is also noteworthy that different recensions of the Milindapañha manuscripts in Siam (Thailand) are not in agreement about using this term to describe King Milinda. For example, some Khom-script manuscripts have omitted the word vetaṇḍī entirely,²⁷ perhaps taking it as a derogatory term to refer to the king. Others have the word as daṇḍī or daṇḍiko (one who is armed), or bhaddantī (a revered person) instead.²⁸ Moreover, the oldest manuscript of the Milindapañha, a fifteenth century Lanna-script manuscript preserved in Lampang, Thailand, has it as ‘taṃ kiñci’ (whatever it is).²⁹ It appears that the sophist is equally misunderstood in this text itself. At the moment, it is unclear which term is used in the original composition of the Milindapañha, as the Pāli text is also an adapted version.

Though the arguments presented by the king seem to explicitly make reference to the principles of logic, or some scholars refer to as nīti (Lenagala Siriniwasa, 2017, p. 891 and p. 896), it is uncertain whether in this context it is referring to the Nyāya philosophy. In Mil, it is said that the king mastered several skills including nīti (Mil 3). Trenckner and Horner take the word nīti here as referring to the Nyāya philosophy (Horner, 1963, p. 5 and n.6; Andersen, 1908, p. 107). However, Schumann argues that the logical source of the Milindapañha may have preceded the Nyāya philosophy known by the Nyāyasūtra, indicating that the systematic logic of the text may have been written under the influence of Gandhāran Buddhists (Schumann, 2019). This might be the case, as according to Baums, the Milindapañha may have been composed with the intention to convert an audience beyond the Indians or Greeks, perhaps the cosmopolitan inhabitants of Gandhāra. The text is Indianized enough in its literary form, as well as Hellenized enough to be persuasive in a Greek style of argumentation (Baums, 2018, p. 42).³⁰ Furthermore, Bhikkhu Bodhi notes that: “The dialogues do not bear at all the character of debates: there is no sparring for victory among rival disputants. Nor do they exhibit the character of the Socratic dialogue, that is, a methodical quest for truth through a graduated course of shared inquiry. Rather,… their binding principle being only the need to resolve conundrums in points of Buddhist teachings” (Bodhi in Mendis, 1993, p. 4). In the Siamese printed edition, Si, as well as the Siamese Khom-script manuscript of the Milindapañha Saṅkhepa (Abridged version of the Milindapañha), dated to the Third Reign (1824–1851), defined nīti (spelt nitti and niti respectively) as the art of kingship.³¹ As such, taking the above into consideration, the word nīti mentioned in Mil is now left open for interpretation. It could also mean another branch of study, like policy or moral philosophy.³² The word nīti listed here in the text, in our opinion, is just to show that the king is both an able ruler as well as a keen scholar who has mastered the various arts and sciences of his time, and it is not there to particularly referring to any section of text which utilises a specific set of philosophical principles. Nevertheless, we also cannot rule out that nīti might be indicating the existence of a form of logic tradition which was practiced in Northeast India in the early centuries CE.

It is noteworthy that not all the questions or arguments presented by the king in the section of the Meṇḍakapañha in Mil, especially the second half, are that of “a point at both ends.” They are simply grounds for puzzlement and perplexity. For example, questions like why is there a need to go forth (to become a monk) if a householder, who gets to enjoy family life, accompanied by worldly material things to please his senses, practices rightly, he too would be successful in the method of Dharma and what is skilled (Mil 242–244); what is the dissimilarity between a householder and a monk who is poor in moral habit (Mil 257–258)?; what is there that does not exist in the world (Mil 267–268)?; and, do all bodhisattas give away their wife and children and do they give them away with their own consent (Mil 274–284)?³³ These questions, despite not being double-pronged, in Mil, are put under the broad umbrella of the Meṇḍakapañha. However, different printed editions, such as the chaṭṭhasaṅgīti (Six Council) edition, classified these non-double-pronged questions into a different broader category of the Anumānapañha (Question [Solved by] Inference) instead.³⁴

As we have seen through the examples above, the meṇḍaka questions, as observed by Bhikkhu Bodhi, are clearly designed to resolve confusing and difficult issues concerning the Buddhist system, either to alleviate anxious concern within the Buddhist community or to refute the challenges from opponents of Buddhist schools who were conversant with Buddhist texts. These notions are echoed in the king’s expression after presenting his arguments to Nāgasena. Below are some of his assertions:

This question is a point at both ends. This [question] is not in the domain of those whose minds are yet to attain [liberation],³⁵ but only in the domain of those who are great. Tear asunder this net of views! Set it to one side!³⁶ This question is [now] put to you. Give insight to the future sons of the Conqueror³⁷ and refute those who hold other tenets.³⁸

This question is also a point at both ends. It is subtle, hard to explain, confusing and profound. It is [now] put to you. This [question] is not possible to be answered by one who has little knowledge except by the one who possesses insight like you do.³⁹

This question is also a point at both ends. It is knottier than a knot, it has more twists than a twist, it is more tangled than a tangle. It is [now] put to you. Here, you must have the desire to refute those who hold other tenets.⁴⁰

The Text and Its Commentaries:

As described by the king above, the nature and the features of a meṇḍaka question are also reflected in the opening verses (gāthā) of the section of the Meṇḍakapañha:

(King Milinda) saw the meṇḍaka questions, which are hard to penetrate and vulnerable to refutations.

There are utterances which are figurative in nature, their meanings are implied, abridged and derived from their own unique conditions.

For those who do not know the meaning, the utterances of the Conqueror are the meṇḍakas, in the future, there will be arguments.⁴¹

One of the commentaries to the text, the Milinda-ṭīkā (Mil-ṭ), believed to be composed in Bingaraṭṭḥa (Chiang Mai) around the second half of the fifteenth century (Mil-ṭ xiii–xiv), further explains the first sentence of the opening verse above as:

“(The king) saw the meṇḍaka questions” meaning: he saw that the meṇḍaka questions are deep through his eye of knowledge. Or it should be said that of Senaka and the rest in the case similar to the question regarding the ram (meṇḍakapañha), which is deceptive and multiple figurative in nature. Or it is also proper to say that he is incapable of discerning⁴² the dichotomic expressions of the question which is similar to two rams battling each other.⁴³

Senaka and the rest mentioned above, no doubt, are referring the four sages, namely, Senaka, Pukkuso, Kāvinda and Devinda, in the Ummagga-Jātaka (Ja VI 330.4–5), where King Vedeha poses the ram and the dog question to them as discussed earlier. Now, it appears that the connection between the “question about the ram” or “horned question” (meṇḍakapañha) in Mil and the Ummagga Jātaka may have come from the commentarial tradition. Other modern translators from Southeast Asia define a meṇḍaka question more or less in line with the text which is puzzling and difficult to answer. U Pu of Burma translated the word meṇḍaka as “that which are twisted like the horns of a goat” (U Pu, 2006, p. 236). Some authors in Thailand narrate it as if two rams’ horns are intertwined when they go on head to head in a fight, the question is of that level of difficulty (Vadhanachaiya, 1993, p. 105). In another Thai translation, the term is being described as a puzzle, a riddle, a secret, and a mystery with two hidden points like the horns of a ram (Mahāmakuṭarājavidyālaya, 2008, p. 145). In a (modern) Chinese translation of the Pāli Tipiṭaka, meṇḍaka is rendered as a question that is difficult to answer (CBETA 漢文大藏經).

Despite all these knotty questions, Nāgasena handles them eloquently, his answers are illustrated by similes and, at times, backed by the unimpeachable word of the Buddha (buddhavacana). In every case, the king accepts and approves Nāgasena’s answers. The usual ending statement in a dialogue of the Meṇḍakapañha section is: “Well done, Venerable Nāgasena, in this way, I accept [your answer] as such.”⁴⁴ Nevertheless, this stock phrase does not always appear at the end of every dialogue too. Below are some further examples of closing statements uttered by the king:

O Venerable Nāgasena, the deep question has been well addressed and made clear, what has been hidden manifested, the knot has been destroyed, the thicket has been made clear, those who hold other tenets are defeated, wrong views have been shattered, those who have been brought up in other tenets are [now] without splendour. You have come to be an excellent and noble [teacher] of a group of followers.⁴⁵

O Venerable Nāgasena, the secret of the Conqueror which has been long hidden away, is now and today, made known and clear in the world.⁴⁶

A Possible Etymological Explanation?

Based on what has been discussed above, there is a strong logical connection between meṇḍaka and κερατίνης or syllogismus cornutus of the Stoics. However, there is also another perspective as to how the word meṇḍaka might have derived coming from a later commentary to the text. The commentary, the milindapañhā-aṭṭhakathā (Mil-a), was composed in Pāli by Thaton Mingun Zetawan Sayadaw of Burma in 1940. We do not know whether this account is based on the author’s personal view or a view held generally by the Burmese tradition. The account below is based on our understanding and interpretation of the commentary. For the translation of the commentary, please refer to the “Appendix”.

In essence, the Sayadaw is of the opinion that the word meṇḍa comes from the word maṇḍa as in maṇḍakappa. Kappa is an aeon or a cycle of the world’s evolution and dissolution.⁴⁷Maṇḍakappa means in this particular world cycle, two Fully Enlightened Buddhas will appear.⁴⁸ As such, that brings in a notion of two into the word maṇḍa. Then the letter ‘e’ in meṇḍa which seems to have been elided from the word maṇḍa, derives from the root √i; to go.⁴⁹ Thus, the ‘e’ here signifies going around (pavatta) as in the sense of “moving” (gatyattha). With that, the word maṇḍa carries the connotation of going around between two meanings (dvīsu atthesu pavatto). Moreover, the word maṇḍa also indicates the presence of a challenger who wants to “rob and steal the splendour” of his opponent.⁵⁰ Furthermore, the word maṇḍa also denotes hard to penetrate (the meaning) and vulnerable to refutations (dunniveṭhe saniggahe) (Cf. Mil 90). Whereas, the ‘-ka’ (as in meṇḍa-ka) is to be understood as a suffix in the meaning of collection as stated in the Kaccāyana Sutta (Kacc. sutta No. 354 and 379). Therefore, in this way, the word meṇḍaka is to be understood—a collection of various components that made up of this word as discussed above. And because the king’s question goes in the combination of two meanings (dichotomic expression), therefore, “that is a question regarding meṇḍaka.” In this way, the Sayadaw indicates, the word meṇḍaka should be understood.

In this commentary, the meaning of the word maṇḍa has been given a different interpretation from its usual meaning of “the essence or finest part of anything” (Cf. PED). Unfortunately, at this moment, some of the passages or quotations in the commentary (given as āgataṭṭhāne) are untraced. As such, it is not clear whether they are the monk’s personal ideas or if he is referring to other texts. There is also a possibility that the explanation for meṇḍaka in this commentary comes about after the monk examines the details and the circumstances pertaining to its usages and then constructs an etymological narrative to fit it in.

Conclusion

In summary, the link between the word meṇḍaka and κερατίνης (‘horned’) or syllogismus cornutus (“horned syllogism”) of the Stoics is mainly based on inference—by way of the phrasing of the king’s argument in a dilemmatic form of expression built on syllogisms. According to the text, a ram or meṇḍaka question, to unenlightened ordinary people, is knotty, figurative in nature, and difficult to solve. The connotation that “a ram question” or “horned question” is both puzzling and tough might have derived from a tale that connects a ram to a riddle and in a way meṇḍa to a difficult puzzle. Metaphorically, a ram could be linked to a dilemma through its two pointed crooked horns with either one of them ready to impale an opponent for choosing either one of the unfavourable choices. These conundrums in the Meṇḍakapañha may have come from the Buddhist community itself or challenges thrown by holders of other tenets. As such, the purpose of the text is to resolve these issues, either to alleviate concerns within the Buddhist community or to refute its opponents. In this way, the term ‘meṇḍaka’ could be a translation of ‘horned’ (κερατίνης) from Greek, since it means the same logical technique for refuting dilemmas which are always imaginary and false.

In a later commentarial text, the word meṇḍaka is said to be derived from a collection of meanings from the word maṇḍa, the root √i, and a suffix -ka. In this case, the word maṇḍa is given a different interpretation—the notion of going between two meanings, and subverting an opponent. With that, in Mil-a, a meṇḍaka question means a difficult question that comes from an opponent which runs between two (unfavourable) choices with the intention to rob the splendour of the Buddhist system.

In sum, a meṇḍaka question, besides carrying its literal meaning of “question about the ram,” could also mean, at least in the Milindapañha, “a dilemmatic expression put forth by a challenger with an intention to undermine an opponent, but which can be logically refuted.”

Appendix

Translation of the Commentary to the Word ‘meṇḍaka’ (Milindapañha-aṭṭhakathā)

‘Meṇḍaka’ is understood as a combination of two meanings. [And] how is such a meaning ought to be known? Here, just like in the passage “for the letter ‘e,’⁵¹ it may have been elided from the root i.”⁵² The letter ‘e’ being “going around” in the sense of ‘moving.’ On the other hand, just like the word ‘maṇḍa,’ in the passage “one Buddha in a world cycle (sārakappa); two Buddhas (maṇḍakappa); three Buddhas (orakappa), four Buddhas (sāramaṇḍakappa); and five Buddhas in a world cycle (bhaddakappa.),”⁵³ the word ‘maṇḍa’ oscillates between two meanings. The letter ‘ka’ occurs in the sense of collocation and a combination of them. In this respect, the grammatical rule (saddalakkhaṇa) for the letter ‘e’ does not need to be examined because it is self-evident. Now, as to the grammatical rule (for the derivation) of the word maṇḍa, [it is about]:

Rob his splendour! Plunder his beauty.

With that, he quarrels, like that he rises above.⁵⁴

In the passage “robbing the benefit of others, plundering the benefit of others, quarreling with others, doing bad things like others” is to be examined. Moreover, it is not only up to just this much. Beyond this, it is to be examined when it is said that “the meaning of maṇḍa is also hard to penetrate and vulnerable to refutations.”⁵⁵ In the passage “the sign of the letter ‘ka’ in the word kaṇhaka is in the meaning of collection,” “The grammatical rule of the letter ‘ka’ [in the word meṇḍaka] is taken from kaṇhaka.” It ought to be examined as the suffix (paccaya) -ka in kaṇhaka of the Kaccāyana Sutta.⁵⁶ And the meaning of collection is a method of collocation and combination. In this way too it is sought for the [word] meṇḍaka is involved, the meaning of the word that “goes in a combination of the meaning of the word maṇḍa,” and because the question is understood in a combination of two meanings, therefore, “that is a meṇḍaka question.” In this way, the meaning is to be known—a question regarding meṇḍaka.

Pāli Text

Except for a few cases, we follow here the printed edition edited by Deshpande (Mil-a 152–153). Punctuation or regularization of the spacing are changed silently.

Editorial Symbols and Abbreviations

]   lemma

/    daṇḍa

cf. confer/compare with

em.   emendation

meṇḍake ti dvinnam atthānaṃ missībhāvaṃ gate/ evaṃ attho kathaṃ jānitabbo/ ekāro⁵⁷ hi ettha idhātuvināsamentū ti āgataṭṭhāne ekāro viya gatyatthe pavatto/ maṇḍasaddo pana sārakappo/ maṇḍakappo/ orakappo⁵⁸/ sāramaṇḍakappo/ bhaddakappo⁵⁹ ti āgataṭṭhāne maṇḍasaddo viya dvīsu atthesu pavatto/ kakāro tesaṃ samodhānamissībhāve atthe pavatto/ tattha ekārassa saddalakkhaṇaṃ pākaṭattā gavesitabbameva natthi/ maṇḍasaddassa pana saddalakkhaṇam/

tassa corehi sobhaggam/ tassa kantiṃ vilumpati//

tena saddhiṃ vivadati/ tulyaṃ tenādhirohatī ti//

āgataṭṭhāne parassa atthacoro/ parassa atthavilumpako/ parena saddhim vivadako/ parena tulyapāpako ti gavesitabbo/ na kevalañca ettako eva/ parato vutte dunniveṭhe saniggahe pi maṇḍassa attho ti gavesitabbo/ kakārassa saddalakkhaṇaṃ samūhatthe kaṇhakā ti āgataṭṭhāne kaccāyanasuttante kaṇhapaccayo ti gavesitabbo/ samūhattho pana samodhānamissībhāvassa pariyāyo/ evaṃ gavesito ca pana so meṇḍako ayati maṇḍānaṃ atthānaṃ missībhāvaṃ gacchatī ti vacanattho/ yasmā ca yo pañho dvinnaṃ atthānaṃ missībhāvaṃ gacchati/ tasmā so pañho meṇḍako ti evaṃ attho jānitabbo/ tasmiṃ meṇḍake/