Etymology And Magic Yaskas Nirukta Flatos Cratylus And Riddle Of Semanticetymologies

Author(s): Johannes Bronkhorst
Publisher: Johannes Bronkhorst
Catalog link: https://jainqq.org/explore/269441/1

Page #1

ETYMOLOGY AND MAGIC: YASKA'S NIRUKTA, PLATO'S CRATYLUS, AND THE RIDDLE OF SEMANTIC ETYMOLOGIES*
JOHANNES BRONKHORST
Summary
Semantic etymologies are to be distinguished from historical etymologies. A historical etymology presents the origin or early history of a word. Semantic etymologies do something completely different. They connect one word with one or more others which are believed to elucidate its meaning. Semantic etymologies are practically universal in pre-modern cultures, and there are treatises in some cultures - such as Yaska's Nirukta in ancient India, Plato's Cratylus in ancient Greece - that specifically deal with them. This article addresses the question how modern scholarship should try to understand semantic etymologizing. It is argued that, being a universal phenomenon, semantic etymologizing is in need of a universal explanation. Drawing inspiration from certain pre-modern philosophies, it is proposed to study this phenomenon in the light of another category of phenomena that is often called "magical".
1.
Yaska's Nirukia and Plato's Cratylus are two texts which belong to entirely different cultures (India and Greece respectively) and which deal essentially with one and the same problem. Both try to bring order to a phenomenon which is extremely wide-spread - not only in ancient India and early Greece, but in probably all pre-modern cultures -the phenomenon of what I will call "semantic etymologizing".
A semantic etymology is to be distinguished from a historical etymology. A historical etymology presents the origin or early history of a word; it tells us, for example, that a word in a modern language is derived from another word belonging to an earlier language, or to an earlier stage of the same language. The English word militant, for example, is derived from Latin militans through the intermediary
"I thank the editors of Numen for useful criticism.
Koninklijke Brill NV, Leiden (2001)
NUMEN, Vol. 48

Page #2

Etymology and Magic
149
148
Johannes Bronkhorst of French militant. And the Hindi pronoun maim 'T' is derived from Sanskrit mayd through Prakrit mae (Oberlies, 1998: 17). Semantic etymologies do something different. They connect one word with one or more others which are believed to elucidate its meaning. The god Rudra, for example, has that name according to the Vedic text called Satapatha Brahmana (6.1.3.10), because he cried (rud-) in one story that is told about him. Semantic etymologies tell us nothing about the history of a word, but something about its meaning.
Semantic etymologies have largely gone out of fashion these days. Most sensible people have serious doubts about the possibility of finding the meaning of just any word by comparing it with other, more or less similar words. We tolerate such semantic etymologizing from children, who indulge in it quite freely, as Jean Piaget (1925) and others after him have shown. We are less tolerant with respect to adults who do so; the person who analyzes the word contentment as conceming being content with men, or with tea (content-men-t), is categorized as schizophrenic by modern investigators, perhaps rightly
the deity created to alleviate this pain is called Ninkasi; similarly, the goddess bom to alleviate the pain of the rib, the Sumerian word for which ti, is named Ninti, etc." (Kramer, 1969:37 n. 13).
An ancient Egyptian text carved inside two pyramids dating from the 24th century is full of plays on words such as: "O AtumKheprer,... thou didst arise (weben) as the ben-bird of the benstone in the Ben-House in Heliopolis." (Wilson, 1969: 3). Morenz (1957) refers to many 'word-plays' (Wortspiele) in Egypi' and onserves: "Für die alt-orientalischen Hochkulturen darf bemerkt werden, dass im Akkadischen (amatu), im Hebräischen (dabar) und auch im Agyptischen (md.r) derselbe Ausdruck 'Wort' und 'Sache bezeich net" (p. 24). Sauneron (1957: 133 f.) adds further examples and points out that plays on words' were considered to give an explanation of the world
In the Hebrew Bible etymologies are common, especially in connection with names: Adam is linked with adama earth (Gen. 2.7): woman, isha, is derived from man, ish (Gen. 2.23); Cain from qaniti have gotten' (Gen. 4.1); etc. (Bohl, 1991: 163 f.).
Kirk (1974: 57f.) emphasizes the use of etymologies in Greek myths and states (p. 58): "The poets of the Homeric tradition were already intrigued by the resemblance of the name 'Odysseus to the verb odussomai "I am angry.... Pytho, the old name for Delphi, is derived [in the Hyment to Apollo, probably late in the seventh century B.C.E.] from the serpent destroyed there by Apollo and allowed to rot, puthein. ... Heraclitus found it significant that one word for a bow resembled the word for 'life (biós and bios), and Aeschylus related the name of Helen to the idea that she took the ships' (hele naus), that of Apollo to apollunai, 'destroy, and that of Zeus to zen, 'live'." Similar efforts at etymologizing characterize later Greek antiquity
An example from medieval Europe is provided by the secret spiritual organization of the Fedeli d'Amore, whose representatives were
80,
And yet semantic etymologies are wide-spread in all pre-modem cultures. Here are a few examples from some cultures different from ancient and classical India :
In the Sumerian myth of Enki and Ninhursag the former is cured when Ninhursag causes deities to be born corresponding to Enki's sick members: "The correspondence between the sick member and the healing deity rests on the ... etymologizing of the ancient scribes: the Sumerian word for the sick organ contains at least one syllable in common with the name of the deity. Thus e.g. one of the organs that pained Enki was the mouth', the Sumerian word for which is ka, and
So Werner and Kaplan (1963: 259), citing a patient of Maria Lorenz (1961: 604).
Several studies collect and discuss semantic etymologies from Vedic and postVedic Indian literaturc, c.g. Balbir, 1991: Bhavadar, 1969: Dange. 1989; Deeg. 1995: Diksi. 1989. Gonda. 1955: Kahrs. 1998: Kantawala. 1967 1963: Norman. 1980. Schneider, 1954: Shastri 1997, Simson, 1988: Singh, 1952: Singh. 1994. Tsuji. 1977: Verma. 1991.
See further Sander Hansen, 1946. pp191
For a study of the eynologies in Himmersee Rank. 1951: For all
implete list of clymogies in Parch. S
Kraus, 1987: 311. I

Page #3

150
Johannes Bronkhorst
active in France, Italy, and Belgium from the 12th century onward. They used a hidden language in order to keep their mystery of love secret. Love for them is a soteriological means, and accordingly the word amor 'love' is interpreted as a-mor 'without death':
A senefie en sa partie
Sans, et mor senefic mort
Or l'assemblons, s'aurons sans mort.
Caesarius of Heisterbach (ca. 1170-ca. 1240) gives an explanation of the word mors 'death' in his Dialogue on Miracles:
Through the transgression of the first created, death entered into the world. Hence death (mors) received its name from 'biting' (morsus). As soon as mun bit tomordir) the apple of the forbidden tree, he incurred death and subjected himself as well as his whole posterity to its necessity. Death is also said to have come from 'bitterness" (amaritudine), because, as it is said, no pain in this life is more bitter than the separation of body and soul. Elsewhere he explains the word puer "boy": "Puer ("boy") signifies purus ("pure")".?
The Chinese language, with its many homonyms, is particularly suited to connect unrelated things that have the same name; the link with what we call semantic etymologizing seems obvious. Indeed, "Han commentators applied a form of correlative thought in their philological studies, frequently explaining the meaning of obscure characters by sound analogy on the assumption that a phonetic correspondence indicated a semantic relation". "Sometimes highly complex circular shou emblems [symbols of long life or immortality] had incorporated into their design a swastika (pronounced wan), to express by a pun the concept of wan shou, meaning 'ten thousand years of long life." Similarly: "The endless knot [was] interpreted... as symbolizing Buddha's intestines (ch'ang). ...[S]ince its name, ch'ang, made a pun on the word for long, the whole figure... symbolized [to the
5 See Eliade, 1986: 112.
Cited in Zaleski, 1988: 50. Cited in Zaleski, 1988: 52. *Henderson. 1984: 19-20.
Etymology and Magic
151
later Chinese] a long life.... etc. Emperor Wang Mang "had the 'screen-walls' fu-ssu of the parks of the Wei and the Yen tombs pulled down, so that the people should not "think again' fu-ssu (of the Han Dynasty)" (Ts'eng Chu-shen, 1949: 126). An example closer to our time is found in the weekly journal Newsweek of July 6, 1987, p. 18: "Hong Kong's new British governor, Sir David Wilson, bowed to local tradition by changing his Cantonese name, Ngai Tak-ngai, shortly before assuming office last April. Its characters were homophones for the phrase 'so hypocritical it's dangerous'; his new moniker, Wai Yikshun, means 'guardianship' and 'trust', conjuring up more soothing images to colony residents..."
An example from ethnographic records is the following: Among the inhabitants of the Trobriand islands the word vatuvi occurs in a magical formula." This word has no grammatical form; it is neither noun nor verb. Malinowski (1935: II: 249, cf. p. 260-61) observes: "the real etymological identity of this word will define it as connected with vitawo, or the prefix vitu-, and the word vituvatu, 'to institute', 'to set up', 'to direct', 'to show'. [It has] also... fortuitous, but magically significant associations with vatu, 'coral boulder', 'coral reef', and the more or less real word va-tuvi, 'to foment', 'to make heal","l
The word 'etymology' itself has an etymology which presents its meaning as 'discourse that makes known the true meaning of a word', from Greek etumos 'true' and logos 'word'. In other words, if we had to decide which of the two, historical etymologies or semantic etymologies, should most appropriately be called etymologies, there can be no doubt that the historical linguist would have to search for
another term.
9 Cammann, 1962: 98, 99-100. I thank Ms. Michèle Boin for this and the following reference.
10 Malinowski (1935: I: 96, 11: 257) describes it as the most important formula in all Omarakana garden magic.
Regarding the last association, va-tuvi, Malinowski observes (p. 260-61): "As
a matter of fact, one or two natives... gave me this explanation of the word when commenting upon the spell." It is not clear whether any native made the association with varu explicit.

Page #4

152
2.
Johannes Bronkhorst
The omnipresence of semantic etymologies, illustrated above, raises important questions: what do these etymologies mean? what are they supposed to explain and how? why do people invent them?
Yaska's Nirukta and Plato's Cratylus deal with these and related questions. I will present the positions of these two texts, beginning with the Nirukta. After that I will briefly discuss the issue how we, in the 21st century, should deal with these same questions.
First the Nirukta. This text is considered a 'limb of the Veda' (vedänga), one of the auxiliary sciences needed to interpret the Veda. It can approximately be dated on the basis of the following reflections. There is reason to believe that Yäska knew Panini's grammar and must therefore be dated later than that famous grammarian (Thieme, 1935: *23*-*24 (530-31); Bronkhorst, 1984: 8 f.). The Nirukta is known to Patanjali's Mahabhāṣya, and is therefore older than that text.12 The Mahabhāṣya was composed in or soon after the middle of the second century before the common era (Cardona, 1976: 266), and Panini appears to belong to the middle of the fourth century before the common era, or to the decennia immediately following it (Hinüber, 1989: 34-35; Falk, 1993: 304). Yäska must fit in between, so that we may date him approximately 250 B.C.E., which is well after most Vedic texts, including the prose portions called Brahmanas, had been composed.
The Nirukta tries to make sense of, and bring order into, the semantic etymologizing that is common in the Vedic Brahmanas. We will see that in doing so it secularizes and rationalizes this practice. In order to appreciate this procedure, we have to first look at semantic etymologizing as we find it in the Brähmanas.
In etymologizing as we find it in the Vedic Brahmanas the following features can be observed:
12 Cp. e.g. Limaye, 1974: 9, 14, 15, 93.
153
Etymology and Magic
1) Etymologies in the Vedic age were more than mere intellectual amusement. Knowing them was believed to be important: it will secure those who know them various advantages.
2) There is a close connection between etymologies and myths. Etymological explanations' refer, almost without exception, to myths,13
3) Etymologies often deal with a hidden dimension of linguistic reality: they reveal hidden layers of language.
4) The number of etymologies for each word is not confined to just one. 14
Each of these features could be illustrated with the help of numerous examples. To keeps things simple, I will give just one or two examples for each.
ad 1) Some passages are quite explicit about the importance and advantage of knowing certain etymologies. The following one is from the Taittiriya Brahmana (3.11.8.7-8; tr. Witzel, 1979: 13):16
Prajapati (the creator god) did not know how to give the sacrificial fee (dakṣina). He put it in his right hand (dakyinah). He took it, speaking the ritual formula (mantra): For fitness (daksu) I take you, the sacrificial fee (dakṣina). Therefore he became fit (adakṣata). The one who knowing thus receives the sacrificial fee (dakṣina), becomes fit (dakçate).
This passage clearly indicates that the etymological link which supposedly exists between the sacrificial fee (dakṣind), the right hand
13 One is here reminded of Max Müller's "etymological method" of studying myths. See van den Bosch, 1993: 188 f. In a lecture delivered at the University of Oslo on February 9, 1996, Eivind Kahrs has suggested that Müller may have been influenced by the chapters on divinities of Yäska's Niruktu. Something like Müller's method has still been used by some authors in the present century; sec Kraus, 1987: 17.
Cf. Deeg, 1995: 397 1. The same is true for ancient Greece: see Lalles. 1991; 137 f. Ovid gives at times two etymologies for one word: Ar focus a flammis et quod fouet omnia dictus ("le foyer tire son nom des flammes et du fait qu'il réchauffe tout". Desbordes, 1991: 155), Various rabbinical etymologies of one and the same word are simultaneously presented; Böhl, 1991: 162.
15 Cp. Deeg, 1995: 411 1.
16 Cf. Gonda, 1991: 177.

Page #5

154
Johannes Bronkhorst
Erymology and Magic
155
(daksina), and fitness (dakşa), must be known. This knowledge guar antees that he who knows, having received the sacrificial fee, becomes fit. Etymology is here more than a mere intellectual help to find the meaning of a word, more precisely, it is no such thing at all. Etymolo. gies reveal links between words, and therefore between the objects they denote, links which it is advantageous to know.
ad 2) The passage just cited illustrates that etymological explanation usually involves reference to myths. Occasionally it appears that etymologies have given rise to myths." This may be illustrated by Satapatha Brāhmana 3.6.1.8-9: "Now, the gods and the Asuras, both of them sprung from Prajāpati, were contending. Then all the plants went away from the gods, but the barley plants alone went not from them. The gods then prevailed: by means of these [barley-grains they attracted to themselves all the plants of their enemies; and because they attracted (ayuvata, from yu) therewith, therefore they are called yava 'barley* (tr. Eggeling)
ad 3) A hidden layer of language is revealed in Satapatha Brahmana 6.1.1.2.18
such as Indra. Sometimes however the gods are said to use the correct forms, only human beings using incorrect forms. So at Satapatha Brahmana 1.1.4.4: "He takes the skin of a black antelope by means of the formula) thou art protection, (a bestower of] bliss (šarman). for carman is its name among men, but Karman is that used among the gods" (tr. Gonda, 1988: 248).
ad 4) Different etymologies of one and the same word (often a name) are frequently met with, sometimes even in one and the same text. A passage from the Satapatha Brāhmana (10.6.5.5) links the name Aditi to the root ad. But another passage from the same Brahmana has an altogether different explanation (7.4.2.7): "Aditi is the earth, for this earth gives (dadate) everything here." Besides the etymology of Indra discussed above (from indh), the Taittiriya Brahmana (2.2.10.4) offers an altogether different one: "No one withstood this power idam indriyam) in him. That is why he is called 'Indra'. Two different etymologies of one word in one and the same passage occur at Satapatha Brahmana 11.1.6.7: "The gods were created on entering the sky, and this is the godhead of the gods (deva) that they were created on entering the sky (div). Having created them, there was, as it were, daylight from him (.e. Prajāpati), and this also is the godhead of the gods that, after creating them, there was, as it were, daylight (divā) for him" (tr. Eggeling).
Summing up: The etymologies in the Brahmanas were believed to bring to light connections between objects that are normally hidden. Similarities between words can reveal those connections. More often than not these connections link the objects concemed with the mythological realm, i.e., with a reality which is not directly accessible to our senses. The fact that multiple etymologies for a single word are frequently met with, suggests that the connections established with their help constitute a network rather than a one to one correspondence. The practical advantage of these etymologies is that they allow man to obtain knowledge about these connections with the hidden reality. This knowledge the texts emphasize it repeatedly - is of great importance: it can convey a number of advantages to him who knows.
Indra, by his power indriya), kindled those other vital airs from the midst; and inasmuch as he kindled (indh), he is the kindler indha): the kindler indeed. -
him they call "Indra' cryptically, for the gods love the cryptic. We learn that Indra's 'real name is Indha, not Indra. We learn from this and many other passages that things and persons (including gods) have a real name which corresponds to their essence. This real name is sometimes hidden. "for the gods love the cryptic". It seems clear that both the gods themselves and human beings use 'incorrect' forms
17C. Devasthali. 1965: 131. Also in ancient Egypt certain myths appear to be based on etymologies, see Malaise, 1983.
I See in this connection Charles Malumoud's article "Les dieux n'ont pas d'ombre: remarques sur la langue secrete des dieux dans l'Inde ancienne": Malamoud 1989: 241-252, originally published in 1984 (Traverses 30-31. p. 86-94).
Deeg (1995: 406 n. 302) points out that these secretive etymologies are not contined to Vedic expressions, as had been claimed by Bhavasar in an unpublished doctoral dissertation (Poona, 1969).

Page #6

156
Johannes Bronkhorst
Etymology and Magic
157
This short résumé shows the extent to which the etymologies of the Brahmanas fit in with other aspects of the religion that expresses itself through these texts. It is not only through etymologies that the Brahmanas establish links with the hidden realm of mythology. And it is not only the knowledge of etymological links that is stated to convey numerous advantages. Nor is the idea of a network of connections only noticeable where the Brāhmanas present etymologies. In short, all the characteristic features that reveal themselves in our study of the etymologies are also found in other aspects of the religion of the Brahmanas.
Similar links are established, not on the basis of verbal similarities, but on the basis of other similarities as well. Michael Witzel (1979: 11 ff.), following Karl Hoffmann (1975-76: II: 524 f.), speaks of 'noems' and 'noematic categories' to refer to traits that objects may have in common, and which are behind the "identifications" which are so typical for these texts. In the pravargya ritual, for example, the glowing red pot is identified with the sun (Witzel, 1979: 2), and the common features between these two objects are easy to guess. In a more recent publication Witzel puts it like this (1996: 169): "The matter may be summarized as follows: any two objects, ideas, entities can be linked with each other by establishing connections of smaller or greater similarity (bandhu, nidana) between them. Then they are not only regarded as linked but as essentially identical - at least within the framework of the ritual. Whatever is done to one object or entity affects the other. ... The identifications or homologies can cover a single aspect of the two or three entities involved (even the number
of syllables of the word signifying both entities) or they can cover a larger number of such links. To discover them is the aim of much of the discussion in the Brahmana style text."
The following characterization of Vedic thought is found in Brian K. Smith's Reflections on Resemblance, Ritual, and Religion (1989: 80-81): "Connection thus bring together the immanent and the transcendent in such a way that the inaccessible is made accessible by the play of resemblances, and the manifest is fulfilled by participation in the transcendent. The universal elements, emitted from the Cosmic One, attain their full, actualized reality only when linked one to another and to their point of origin. Such a composition based on connection rejoins the Cosmic One into a unity of parts in which the simple and the limited - while remaining simple and limited - participates in the whole, the unlimited. Each particular 'name and form' can realize its true nature only by finding its place in this chain of resemblance
- or, rather, by being placed in bondage with all of its counterparts under the umbrella of the prototype. Universal resemblance keeps sep. arate while it unifies, its specific economy regulated by avoiding the extremes of identity and individuality."
Both the notion of "identification" and that of "resemblance" in this connection have been criticized by Albrecht Wezler (1996), who draws attention to the fact that some "identifications" do not connect with the mythical realm, i.e., with the transcendent. Sometimes also an explicit justification is given which has nothing to do with resemblance, but all the more with other factors, such as a causal link, or something else. While Wezler's cautionary observations may no doubt contribute to a fuller understanding of the texts concerned, there is no reason as yet to doubt that many of these "identifications." like many of the etymologies, establish a link with the transcendent. Indeed, the fact that resemblance (phonetic resemblance) plays such an important role in the etymologies, is an argument - if such was needed in support of the importance of resemblance in at least a considerable part of the "identifications."
* Cp. Brian K. Smith's (1989: 47) following observation, which sets the tone for his hook Reflections on Resemblance. Rimal, and Religion: "I would suggest here. and will be at pains to prove thoughout the remainder of this study, that there is a philosophical center around which all Vedic thought resolves. That center will call resemblance. This concept, I believe, underlies Vedic religious and philosophical discourse in its entirely.... Universal resemblance, whereby entities, things, forces, activities, cosmic planes - indeed, all the components of the universe as a wholc - have essential affinities to related others, helps us to reform our understanding of the Vedic preoccupation with making and finding connections."

Page #7

158
Johannes Bronkhorst
Erymology and Magic
159
The Nirukia - it was stated before - tries to make sense of, and bring order into these etymologies. How does it do so? Here we have to keep in mind that Yaska, being most probably a Vedic Brahmin. could not reject the validity of these etymologies. Their validity was, for him, beyond doubt. His question was rather: how have they been arrived at and, how does one establish new ones? The fact that he asked these questions and looked for general rules underlying the etymologies of the Brahmanas no doubt betrays Yaska's intellectual distance from the etymologies he tried to understand. It is certainly no coincidence that most of the four features which characterize Vedic etymologizing and which were enumerated above, no longer characterize Yaska's procedure. His etymologies do not typically establish a link with the mythological realm, nor do they as a rule reveal hidden layers of language. They are still secret, but no longer for the concrete reasons given for this in the Brahmanas. The one feature that remains valid for Yaska's etymologies is that there can be several etymologies for one word.
It was pointed out above that the enumerated features of Vedic etymologizing fitted very well into Vedic religion. Etymologizing was just one more way of establishing the links that according to Vedic religious understanding link different objects belonging to this and the other world. Since in Yaska's etymologies the religious dimension (connection with the mythological realm, access to a hidden layer of language) has virtually disappeared, we may conclude that Yaska's religious views were no longer those of the Vedic period. This, however, raises the question what good etymologies would be for him.
A number of rules are formulated in the second chapter of the Nirukta to help the student find etymologies on his own. The most important among these rules is no doubt the one that etymologizing should first of all, be guided by the meaning of the word concerned; phonetic considerations play a less important role: "One should examne la word being intent upon [ils] meaning, with the help of some similarity in function (with other words). When not even such a simi
larity is present one should explain on the basis of similarity in a syllable or in a single sound" (Niruka 2.1). In the case of unknown words, therefore, one looks at the context in which they occur (usually a Vedic hymn), so as to get a first impression as to their meaning. Subsequently one looks for other words (they have to be verbal forms, according to the Nirukta) which are more or less similar to the word under study. Semantic considerations, however, come first. So a verbal form which is less similar but closer to the expected meaning is to be preferred to a more similar verbal form which does not support the desired meaning, And words which are known to have several meanings, have also several etymologies. An example is the word go: "The word go is a name for earth' because it goes (gata) far and because living beings go (gac. chanti) on it. Or it is a name of something which moves (gāti). o (in go) is a nominal suffix. Moreover, the word go] is the name of an animal (viz. cow) for this same reason. ... Also a bow-string is called go ... because it sets arrows in motion (gamayati)" (Nirukta 2.5). And if one does not find verbal forms that resemble the word to be explained, one should not be discouraged.
The Nirukta gives no explanation as to why etymologies' should be valid at all. The explanation which has been offered by most modern interpreters of the text, viz., that the etymologies' tell us something about the history of the words concerned, is demonstrably incorrect. It appears that the author of the Nirukta did not look upon language as something developing in the course of time. This is not surprising in the Indian context; we know that the Veda, and therefore also the language in which it is composed, came to be looked upon as eemal, i.e, without beginning. It is not impossible that this view existed already in the days of the Nirukta, i.e., several centuries before the beginning of the Christian era (see above). Classical India, unlike for example classical Greece, could believe in the existence of one 'real' language, all other languages being, at best, imperfect reflections of it.
12 This interpretation of Sanskrit gater w follows Eivind Kahrs, 1984: $12: cp. Kahrx. 1998: 115, 132-133.
2. This is shown in Bronkhorst, 1981: Kahrs, 1983, 1984.
For a full discussion, see Deeg. 1995:78f.

Page #8

160
Johannes Bronkhorst
Etymology and Magic
161
One way to account for the validity of semantic etymologies based on the similarity between words (for those who accept this validity) would be to claim that there are ultimate meaning bearers, such as individual sounds or small groups of them, each with its own specific meaning, Plato's Cratylus does indeed explore this possibility, as we will see below. However, the Nirukia does not adopt this position. Interestingly, a number of early Buddhist texts, while referring to Brahmanic learning, mention the term aksaraprabheda (Pali akkharappabheda)," which 0. Franke (1913: 87 n. 6) translates "Unterscheidung der Silben", the Pali commentators specify that the reference is to two forms of linguistic analysis, one of them being etymologizing. It seems, therefore, that the idea that individual sounds or syllables have meanings of their own, and that this presumed fact explains semantic etymologies, was not unknown in ancient India, even though the Nirukta does not mention it.
The grammarian Patañjali (2nd century B.C.E.), the most authoritative commentator on Panini's grammar, considers and subsequently rejects the proposal that individual sounds have meanings. Among the reasons he adduces-following his predecessor Kätyāyana, whose statements (varttika) he comments is the following: In grammatical derivations there can be transposition, loss, addition and modification of sounds. If sounds had meanings, these meanings, too, would undergo transposition, loss, addition and modification. Such is not how ever the case. This argumentation is of particular interest, for Yaska's Nirukta (2.1-2) had presented almost exactly the same reasons in order to show that in etymologizing one is free to transpose, remove, add or modify sounds. This proves that in the opinion of their practi
tioners neither etymologizing nor grammar could possibly arrive at meanings of individual sounds."
This does not mean that the idea of "real" meanings attaching to individual sounds was abandoned hy all in ancient India. A different attitude towards language, and towards sacred utterances in particular, manifests itself in the religious literature of India. This different attitude is interested in the deeper - some would say: mystical
- meaning of these utterances. Already the Vedic texts sometimes ascribe significances to parts of words that have nothing to do with their ordinary meanings. For example, the three syllables of the word pu-ru-sa person, self' are stated to correspond to a threefold division of the self: to be placed in the world of the sacrificer, in the world of the immortal(?) and in the heavenly world respectively (Altarea Brahmana 3.46 (15.2)). The three syllables of hy-da-yam 'heart' are explained as follows: "hr is one syllable. Both his own people and others bring (hr) offerings unto him who knows this. da is one syllable. Both his own people and others give (da) unto him who knows this, yam is one syllable. To the heavenly world goes (eti (pl. yanti]) he who knows this." The 36.000 syllables of 1000 brhati hymns correspond to as many days of a hundred years, according to the Aitareya Aranyaka (2.2.4). The seventeen syllables of the utterances o fravaya, astu srausat, yaja, ye yajāmahe and vausar are the seventeenfold god Prajapati (Satapatha Brahmana 12.3.3.3).
More detailed are two passages from the Chandogva Upanisad. The word satyam "truth' is said to consist of three syllables sa-ti-vam: saft) is the immortal, ti the mortal, with vam the two are restrained (root yam-) (Chandogya Upanisad 8.3.5). The three syllables of d-gi
* It does occasionally present 'deep' forms which "hide behind the surface forms; eg Nir. II: re nigantava er santo nigamanan nighantava contain aupamanyavah/ "According to Aupamanyava, these lists of words] are called righantus, being really nigantes because they are quoted (nigmandr)."
25 See Bronkhorst. 1989: 129 f. 26 Mahabhapya, ed. F. Kielhom, Bombay 1880-1885, vol. 1 p. 30-32.
27 An exception must of course he made for such verbalnoots and other grammatical elements as consist of just one sound. 2% Brhadaravaka Upanised 53: tr. Hume.
i is the dual of las Keith (1909: 207) pointed out. The analysis sull-ti-am is also found in Brhadirayaka Upanised 55.1. Aire Angaku 2.1.5. For another explanation of unum ( sol + vam), see Brhadarumvuke Upwisud 2.3. kowski Upinised 1.6. Taittiriya Upanisod 2.6: further Kudelska. 1995,

Page #9

162
Johannes Bronkhorst
tha 'chanting of the Samaveda' mean respectively 'breath' because by it one stands up (uttisthati) 'speech' (gir), and "food" - in which all this is established (sthita)-(Chandogya Upanisad 1.3.6). The second of these two analyses tries to keep contact with the 'real' meanings of the syllables concerned, the first one does not even try to do so.
The Brhadaranyaka Upanisad contains a story (5.2) that is interesting in the present context. The gods, men and the demons dwelt with father Prajapati as students of sacred knowledge. Asking for instruction. Prajapati uttered the same syllable da to each of them. The gods understood this as damyata 'restrain yourself", the men as datta 'give', while the demons understood this same syllable da as dayadhvam 'be compassionate'. The divine voice which is thunder repeats the same: da da da, which is: damyata, datta, dayadhvam. Therefore one should practice restraint (dama), liberality (dana) and compassion (daya). Unfortunately the passage does not explain what is the point of this story, and perhaps one should not attach too much significance to it. It may however be legitimate to surmise that it attributes three different meanings to the single syllable da, meanings which normally express themselves through the intermediary of the words damyata (or dama). datta (or dana) and dayadhvam (or dayā).30
Concern with single syllables may also be visible in the explanation of the word sarkara 'pebble' with the help of the syllable sam 'welfare: Sarkara is called thus, while welfare (sam) befell us."
An early indication that individual speech sounds were looked upon as possessing powers may be found, according to Thieme (1985), in the last verse of the Maitrāyaṇīya Samhita, which is also the first verse of the Saunakiya recension of the Atharvaveda. This verse reads: 32
Cp. Houben, 1997: 70.
31 Taittiriya Brahmana 1.1.3.7: sam vai no 'bhad in/ tue charkaranam surkarat
vam/.
Maitrayantva Samhita 4.12.1 (ed. von Schroeder p. 179 1. 14 f.)~ Atharvaveda (Saunakiya) 1.1 Atharvaveda (Paippalada) 1.6: vé trisaptan parivanti viśvā rūpāni bibhrataly/ vācāspátir bálā téṣām tantüvà dá dadharu mel/ tr. Thieme. Doubts
Etymology and Magic
- let the
"The thrice seven that go around, wearing all the shapes Lord of Speech put their powers into my body's [parts] today." Thieme argues that 'the thrice seven' are the sounds of language and shows how they can, and may have been, looked upon as constituting a list of 21 elements. He then concludes (p. 565 (938)): "The basic sound units of the sacred language, amounting to the sacred number 'thrice seven', are the basic sacred elements of the sacred language. Being sacred, they are loaded with magic powers. Rehearsing them the brahmacarin will not only obtain the technical ability of correctly repeating and retaining what his teacher recites to him, he will, also, appropriate those magic powers: 'May the Lord of Speech put their powers into my body's [parts] (or: in my body) today (i.e., at the beginning of my vedic studies)'."
163
These examples from Vedic literature point the way to a much more wide-spread concern with the deeper significance of small groups of sounds, and even individual sounds. It manifests itself in the speculative analyses of the sacred syllable om and elsewhere, and reaches its apogee in certain Tantric texts, which attribute a specific metaphysical significance to every sound of the Sanskrit language.34 These Tantric
regarding Thieme's interpretation of this verse have been raised by Deshpande (1997:
33 f.).
33 Similar examples can be found in more recent literature. The Devi Bhagavata Purana (9.1.6-7: cited and translated in Jacobsen, 1999: 26-27) explains the word. prukti in two ways, the second one dividing the word into the three syllables prakr-ti: "The pra-word means the most excellent sattva guna, kṛ means the middle rajus guna, and ti denotes the famas gunnar. She whose own nature is triguna, is endowed with powers. She is superior in creating, therefore she is called prakṛti." (gume suve prakrste ca pra-fabdo vartate śrutah/ madhyame rajasi krs ca ti-fabdas tamasi smrtal/l trigintasvaräpä ya să ca faktisamanvital pradhana sṛṣṭikarane prakrtis tena kathvatel/).
34
See Padoux, 1990. These Tantric developments are not without precursors in Vedic literature. See, for example, the following passage from the Puncavisa Brahmana (Tandya Maha Brahmana) (20.14.2) and Jaiminiya Brahmaṇa (2.244; close to, but not identical with it): "Prajapati alone was here. Vac alone was his own; Vac was second to him. He reflected, 'Let me send forth this Vāc. She will spread forth, pervading all this. He sent forth Vac. She spread forth, pervading all this. She

Page #10

164
Johannes Bronkhorst
Etymology and Magic
165
people went on etymologizing. No satisfactory theoretical justification was however worked out, even though the idea that constituent syllables or sounds somehow possess meanings that account for the meaning of the whole word survived in various forms.
speculations present, in a way, the Indian counterpart of Plato's primary names and the primary sounds' of the Stoics, to be discussed below. Yet there is a major difference. These Tantric speculations base themselves primarily on so-called bija-mantras, utterances which are usually devoid of ordinary meaning. The metaphysical meanings assigned to the individual sounds are not, therefore, meant to contribute to the meanings of ordinary words that contain them. No longer restrained by the shackles of ordinary language use, the Tantric authors could establish the meanings of all the sounds of the Sanskrit lan
guage.15
It will be clear that these Tantric speculations are far removed from the etymologies which form the subject matter of this study. We will not, therefore, study them in any detail. Be it however noted that these Tantric speculations have parallels in the Jewish Kabala and similar developments within Islam. Yet, though removed from etymologies, these speculations cannot be completely separated from them. They are, in a way, the ultimate outcome of the process of analysis which found its inspiration in those etymologies.
Where does this leave us with regard to the question as to how Indian thinkers explained semantic etymologies? During the Vedic period the validity of these etymologies was not questioned since they were based on the more general principle, not confined to language, that similar things are connected or even identical - with each other. During the then following period this justification fell away, but
The situation in ancient Greece is rather different from India, in that the Greeks and their successors did not look upon their language as the only the language. This complicated matters considerably, and it is not impossible that this fact is partly responsible for the relatively suspicious way in which the problem was often approached in the Western tradition. Yet there is a respectable list of thinkers who occupied themselves with it."
Plato's Cratylus is the first full investigation of etymologies that has survived. In this dialogue Socrates is engaged in a discussion with two other characters, Cratylus and Hermogenes. It is possible, but not certain, that Cratylus in real life represented an etymologist'; it seems however certain that the etymological point of view did have real supporters. Plato's dialogue, i.e. the person of Socrates in it, initially Seems to support it, but in the process of working it out in detail changes position.
The basic question discussed in the dialogue is whether "everything has a right name of its own, which comes by nature" (383a). Arguing that this is the case, Socrates is led to conclude that the initial law givers knew "how to embody in the sounds and syllables that name which is fitted by nature for each object" (389b). Astyanax 'Lord of the city for example, being the name of the son of Hector the ruler of Troy, is appropriate (392d-e) or, as he says earlier (385bf.), true. This example takes us right into the analysis of words in view of determining their
extended upwards as a continuous stream of water. (Uttering the sound), he split of a third of it-that became the earth... Uttering the sound kohe split off a second] third-that became the midregions... Uttering the sound ho he cast the last third upwards that became the heaven." (tr. Holdrege, 1994: 44). The content provides no clue as to why exactly these sound have the effect described
See Padoux, 1991): 235 if: Ruepr. 1959: 108
For the Jewish Kabala, see G. Scholem. 1983: 55.99 (Le nom de Dice ou la théorie du langage dans la Kahale, mystique du langage"), for Sufism, see Schimmel. 1975: 411 ff. "Letter symbolism in Suh literature"). Staal (1979: 7)brielly refers to the parallelism between Kahala and the Tantric speculations under consideration
Note that right at the beginning of Plato's dialogue of that same. Cratylus is presented as holding the view that there is a kind of inherent correctness in names which is the same for all men, both Greeks and barbarians" (3832)
See Kraus, 1987, Lallot, 1991 Rijluarsdam (1978: 65 .) discusses the use of the word 'name' ( MRI).

Page #11

166
Jolicomes Bronkhorst
Etymology and Magic
167
appropriateness. This analysis does not however have to fit the word too closely. It doesn't matter if a letter is added or subtracted (393d). **Varieties in the syllables is admissible, so that names which are the same appear different to the uninitiated, just as the physicians' drugs, when prepared with various colours and perfumes, seem different to us, though they are the same, but to the physician, who considers only their medicinal value, they seem the same, and he is not confused by the additions. So perhaps the man who knows about names considers their value and is not confused if some letter is added, transposed, or subtracted, or even if the force of the name is expressed in entirely different letters. So, for instance, in the names we were just discussing. Astyanax and Hector, none of the letters is the same, except i but nevertheless they have the same meaning." (394a-b). "We often put in ar take out letters....and we change the accent" (399a). Originally the true nature of words was more easily recognizable, but people attach "more importance to euphony than to truth" (404d), they beautify names (408b). they add sounds merely for the sake of euphony (412e). to make the words prettier (417e), they "care nothing for the truth, but only for the shape of their mouths" (414d). "The original words have before now been completely buried by those who wished to dress them up for they have added and subtracted letters for the sake of euphony and have distorted the words in every way for omamentation or merely in the lapse of time" (414c); "they keep adding to the original words until finally no human being can understand what in the world the word means" (414d). In brief, words get twisted in all sorts of ways" (421d). The ancient language, on the other hand, shows clearly the real sense of words (418b). It is also clear that when anyone knows the nature of the name - and its nature is that of the thing
- he will know the thing also, since it is like the name" (435d-e). It is striking to see that the factors specified in the Nirukta as to be taken into consideration in etymologising (transposition, loss, addition and modification of sounds) are found here, too.
The Cratylus contains a great number of practical examples of etymologies'. For our present purposes it is not necessary to discuss these in detail. It is more interesting to see how Plato, through Socrates, faces the problem of the multitude of languages. Greek is for him not the only language, nor indeed the only correct, or even the best language. The creators of other languages are clearly assumed to have taken equally great care to make names in those other languages 'fit' their respective objects. The ideal names are expressed differently in different languages: "Then, my dear friend, must not the lawgiver also know how to embody in the sounds and syllables that name which is fitted by nature for each object? Must he not make and give all his names with his eye fixed upon the absolute or ideal name, if he is to be an authoritative giver of names? And if different lawgivers do not embody it in the same syllables, we must not forget this ideal name on that account for different smiths do not embody the form in the same iron, though making the same instrument for the same purpose, but so long as they reproduce the same ideal, though it be in different iron, still the instrument is as it should be, whether it be made here or in foreign lands... On this basis, then, you will judge the lawgiver, whether he be here or in a foreign land, so long as he gives to each thing the proper form of the name, in whatsoever syllables, to be no worse lawgiver, whether here or anywhere else" (389d-390a). The result of this is that the etymological method can be applied to other languages too: "f foreign names are examined, the meaning of each of them is equally evident" (400b-c). However, if we should try to demonstrate the fitness of [foreign] words in accordance with the Greek language, and not in accordance with the language from which they are derived, you know we should get into trouble" (409e). Socrates admits however that where he finds it hard to understand a word and its etymology he applies the contrivance of claiming it to be of foreign origin (416a).
The discussion of this device induces Socrates to take his investigation even further than heretofore, and it is this continued investigation which provides him with one of the arguments because of which he
4! They have been collected and systematized in Gaiser, 1974: 54-57.

Page #12

168
Johannes Bronkhorst
Ermology and Magic
169
tums in the end against etymologies'. His train of thought runs as follows: "If a person asks about the words by means of which names are formed, and again about those by means of which those words were formed, and keeps on doing this indefinitely, he who answers his questions will at last give up... Now at what point will he be right in giving up and stopping? Will it not be when he reaches the names which are the elements of the other names and words? For these, if they are the elements, can no longer rightly appear to be composed of other names" (421d-422a). This gives rise to a question: "How can the earliest names, which are not as yet based upon any others, make clear to us the nature of things, so far as that is possible, which they must do if they are to be names at all?" (422d-e). The answer proposed by Socrates is that the name-maker grasps with his letters and syllables the reality of the things named and imitates their essential nature" (424a-b). Socrates admits that it will seem ridiculous that things are made manifest through imitation in letters and syllables" (425d); yet there is no alternative, unless we were to believe that the gods gave the earliest names, or that we got the earliest names from some foreign folk and the foreigners are more ancient than we are, or resort to some other evasive tactic (425d-e). Socrates therefore proceeds to assign meanings to individual letters, it would take us too far to give a detailed account of his results, but the principle is simple: the phonetic nature of a sound corresponds to the object it denotes, the active sound rho, for example, expresses activity. By combining these individual letters, the lawgiver makes by letters and syllables a name for each and every thing, and from these names he compounds all the rest by imitation (427c).
Having reached this far, Socrates discovers an insufficiency in the view propounded, which he uses as one of his arguments against it! "If the name is like the thing, the letters of which the primary names are to be formed must be by their very nature like the things" (434a). But not infrequently a word contains sounds which have no right to be there, such as the sound lambda, which expresses softness, in the word sklerótes "hardness' (434d). One might of course argue that this is an added sound which does not really belong in this word, but
this raises the question how it got there. The answer can only be 'by custom' or 'by convention, but this takes us back to the position which was intended to be refuted in the first place, viz., that the relationship between words and their objects is determined by convention. Socrates concludes: "I myself prefer the theory that names are, so far as is possible, like the things named; but really this attractive force of likeness is, as Hermogenes says, a poor thing, and we are compelled to employ in addition this commonplace expedient, convention to establish the correctness of names" (435c).
One of the things to be noted in this dialogue is the desire to identify the ultimate elements of language and their meanings. Indeed, Socrates tums against the position of Cratylus precisely because his attempt to connect the primary names with the things denoted does not succeed.
In contrasting the Cratylus with the Nirukra and with Indian etymologising in general, several important differences deserve our altention. It has already been pointed out that the Greeks did not look upon their language as the only true language. To this must be added that Plato speaks about words as having been created by a or several lawgivers with the dialectician as his supervisor": 390d), which is in total contrast with the Indian conception of things. Indeed, the grammarian Patanjali (introduced above) made the famous, and opposite. observation that no one who is in need of words would go to a grammarian the way someone in need of a pot goes to a potter to have one made. The idea of words being made by anyone, human or superhuman, was totally unacceptable in India. For Plato, on the other hand, it is fundamental. What is more, the original name-givers were no ordinary persons (401b), and the suggestion is made that he who gave the first names to things (here the singular is used) is more than human (438c). The name-givers are sometimes called demiourgos (43le), and it is not impossible that Plato looked upon the original name-giver as close to, or identical with, the Demiurge, the maker of this world mentioned in some other dialogues (esp. Timaeus). This link is particularly interesting in that it connects etymologising with cosmology, a connec
Milli.ed. Kieler, Bumbu XII 1885, w
Ip 7-8

Page #13

170
Johannes Bronkhorst
Ervmology and Magic
171
tion that came to play a role in the thought of later Platonic thinkers as we will see below.
I will not pursue the further development of semantic etymologizing in classical Europe. The Stoics actively practiced it, as did the Alexandrian tradition of grammarians. The situation in European antiquity came to be somewhat complicated by the fact that some ideas about the development of one language out of another one came to be ac. cepted. Latin, in particular, was often thought of as having derived from Greek. Myths, such as the one about the origin of the founders of Rome, supported this belief (Strobach, 1997: 85). When, therefore, an author like Plutarch derives Latin words from Greek words, he may intend his etymologies to be understood as histories of the words con cerned. Whether all his etymologies are to be understood in this manner remains unclear (Strobach, 1997: 55 ff.). There can however be no doubt that texts like Plato's Cratylus deal with semantic etymologising, not with historical etymologies.
I think we are entitled to conclude from the above observations that early Indian and classical European thinkers were aware of some of the difficulties surrounding semantic etymologies. The author of the Nirukta tries to formulate the rules that permit us to obtain valid semantic etymologies, but does not feel free to doubt their validity The author of the Cratylus, on the other hand, arrives at the conclusion that the validity of semantic etymologies almost inescapably implies that individual sounds have each their own meaning. Since he finds this difficult to accept, he raises doubts as to the validity of such etymologies.
Two directions are open to us. On the one hand, we can try to understand semantic etymologies against the background of their respective cultures, in other words, Indian semantic etymologies are to be explained in terms of Indian culture, Greek semantic etymologies in terms of Greek culture, and so on. Alternatively, we can see in semantic etymologies a universal phenomenon, not inherently linked to any particular culture, and therefore look for a universal explanation. The examples adduced in the preceding pages should have left no doubt that semantic etymologizing is indeed a universal phenomenon, which is in need of a universal explanation. However, before tuming to it we will have to pay attention to the one serious attempt that has been made to understand Indian semantic etymologies as an Indian phenomenon.
Eivind Kahrs' recent study called Indian Semantic Analysis: The nirvacana' tradition (1998) presents a new and interesting hypothesis concerning semantic etymologies in classical India, not primarily Vedic etymologies. For Kahrs, etymologies are part of the Indian universe of discourse. That is to say, etymologies are part of the methods used in Indian culture to interpret its texts. The study of etymologies in India is therefore the investigation of a systematically applied means of interpretation (p. 9). By finding here patterns that are repeated over and over again, one may detect certain basic features of classical Indian traditions. "For the present investigation --- Kahrs states on p. 11 - "it is precisely the constant factors and the indigenous interpretations of them at various points in time which are of interest
Kahrs study concentrates, as far as etymologizing is concerned, on the Nirukta, and on some Saiva texts from medieval Kashmir. These texts use etymologizing as a conscious device. Briefly put, Kahrs claims (p. 174) that "it is possible to argue that ultimately all nirvacanas (= semantic etymologies, JB) are to be understood according to a substitutional model".
Chapter 4 ("The universe of Yaska") deals in great detail with the various ways in which nirvacanas are presented in the Nirukra. The most important ones of these, Kahrs argues on the basis of copious material, use essentially a genitive case ending. The most perfect way of presentation is of the type megho mehatiti satah, which literally
For us, modern researchers, the validity of semantic etymologies is no longer an issue: semantic etymologies are not generally valid. Neither Yaska's method nor Plato's speculations as to the meanings of individual sounds are acceptable to us. We are nevertheless confronted with the problem that many people apparently did accept these etymologies as valid, and our problem is to make sense of that.

Page #14

172
Johannes Bronkhorst
Etymology and Magic
173
means: "megha ('cloud) is of something really existing such that one can say (of it): mehati 'it rains'." The genitive ending finds expression in the word 'of' of the translation.
The next question is what is the exact meaning of this use of the genitive? What does it mean to say that the word megha is of something of which one can say it rains' (mehari)? The most straightforward interpretation might seem to be that the word megha belongs to something, viz. a cloud, of which one can say that it rains. This would seem to make perfect sense. Yet Kahrs does not seriously consider this possibility. He rather translates phrases like this in the following manner: "megha is in the meaning of that which really exists so that one says (of it): 'it rains" (p. 162; my emphasis). Yaska, according to Kahrs, employs a genitive to indicate a substitution procedure as well as to indicate that which is signified by a word and thus ultimately its synonym. The substitutional model, mentioned earlier, is thus based on a particular interpretation of the genitive. Is this a regular interpretation of the genitive in Sanskrit?
Kahrs claims it is. This kind of use of the genitive ending is called sthānasasthi in Sanskrit grammatical literature. Kahrs believes that it is firmly rooted in ordinary Sanskrit (p. 234). He comes to the conclusion
that the usage of the sthānasasthi is a well established feature of ordinary language" (ibid.). "It is evident" -he states on the same page
- "that you could get the usage of the sthānasashr from the Sanskrit language itself". And again, one page earlier (p. 233): "Such a usage of the genitive is in accordance with established Sanskrit usage."
This all sounds rather favourable to Kahrs's main thesis. All that remains to be done, one would think, is cite some passages from classical or Vedic literature that show that such a usage of the genitive is indeed well established in Sanskrit. No attempt is however made to prove the point, so often repeated, that the substitutional use of the genitive is well established Sanskrit usage. The reason is easy to guess. The genitive of substitution may not be all that well established in Sanskrit
An exception has to be made for grammatical literature. Pāņini's grammar, in particular, uses the genitive in this way. Kahrs rightly
points out that "to interpret a genitive as a substitutional genitive is nothing remarkable in Sanskrit grammatical literature" (p. 236). Then. however, he continues: "Nor is there anything... which restricts such a usage of the genitive to grammatical texts." This is far from obvious. The substitutional genitive in grammar is a technical device which, like most other technical devices of grammar, is most unlikely to be valid anywhere else. Panini's grammar uses a number of technical terms and devices, which are properly introduced, and which claim no validity outside this grammar. The special use of the genitive is just one of such devices. It is introduced in sūtra 1.1.49 (sasthisthänevoga). which means in Böhtlingk's translation: "Der Genitiv in einem Sûtra bezeichnet dasjenige, an dessen Stelle Etwas treten soll." The use of the genitive in Panini's grammar is therefore special, and should not one would think, be used to elucidate the use of the genitive in other works.
Kahrs does not agree. He discusses in detail the portion of Patanjali's commentary (the Mahabhäsya) on this sūtra of Panini, as well as the subcommentaries thereon, and finds there an argument which, in his opinion, "would not work were not the use of the sthānasasthi firmly rooted in the usage of ordinary Sanskrit" (p. 234). This, and only this, makes him conclude that the usage of the sthānasasthi is an established feature of ordinary language". This conclusion, which is vital to Kahrs's thesis, depends therefore, not on an attestable feature of the Sanskrit language, but on the correct interpretation of a scholastic argument, and on nothing else.
Which is the argument which leads Kahrs to his conclusion? It is essentially this. Patañjali points out that the genitive case ending can have "one hundred meanings, or as many as there are," and that therefore sūtra 1.1.49 restricts, for Panini's grammar, the meaning of the genitive ending to the single meaning "that in the place of which something will be substituted" (p. 197). Kahrs is of the opinion that the sutra can only restrict the meaning in this manner, if the meaning
that in the place of which something will be substituted" does already belong to the genitive ending. Normally the genitive ending expresses

Page #15

174
Johames Bronkhorst
Erymology and Magic
175
this meaning as well as all the other ones, but as a result of sūtra 1.1.49 only this meaning remains.
Since this part of Patanjali's comments - which follow a vårttika by Kätyāyana -- sets the tone for the then following discussion, it is important to understand it correctly. According to Kahrs these comments presuppose that the genitive covers, of its own, the substitutional meaning. From a purely logical point of view he is right. Logically speaking, the substitutional meaning must be one of the hundred meanings,"or as many as there are that Patanjali assigns to the genitive case ending. But this is not the same as concluding that the substitutional genitive is a well established feature of Sanskrit in general. The general mean ing of the genitive case ending is relation in general, more precisely everything that remains after specifying the meanings of the other case endings (P. 2.3.50; sasthi sese). Kahrs describes it as follows (p. 237): "A/städhyayi/ 2.3.50 sasthr sese teaches that a genitive case ending is introduced to denote the rest' (sea). According to the Kasikāvrtti this rest is any relation, sambandha, which is not a karaka-relation and different from the meaning of the nominal stem. In other words, a genitive case suffix is introduced to denote any relation sustained between entities, that is to say, any non-verbal relation in general, such as father-son, master-servant, part-whole, etc. The fact that two entities are mutually related by their appearance in a given context is expressed by the genitive case. But the particular type of relation is not specified." This, of course, includes an enormous lot. It includes, for example, the sense uncle of". and in certain exceptional circumstances the Sanskrit equivalent of "John is of Mary" or "John belongs to Mary" will have to be interpreted as meaning "John is Mary's uncle". This is not however the same as saying that the meaning "uncle of" for the genitive case ending is a well established feature of the Sanskrit language. The situation is not different, as far as I can see, for the substitutional meaning of the genitive case ending. There is just no evidence that this meaning is a regular feature of Sanskrit. This does not change the fact that this meaning, like the meaning "uncle of", is somehow included in the hundred or more cases covered by the genitive.
It is interesting, but also somewhat puzzling, that Kahrs seems to agree with this conclusion. He cites (p. 238 ff.) the example devadattasva vajñadattah "Devadatta's Yajñadatta", where the idea is that Yajñadatta is the son of Devadatta. He compares this with the grammatical rule (2.4.52) aster bhūh "ofas, bhu". Then he remarks (p. 238): "Just what the relation is ... can not be known from the statement aster bhah alone, anymore than one knows from the statement devadattasya yaradantah just what relation obtains between Devadatta and Yajñadatta. But if yaradattah is replaced by an obvious relational term such as putrah'son'. the relation in question is immediately understood: Devadatta's son. Now, in the case of aster bhūh, where bhu itself is no obvious relational term, the relation marked by the genitive is determined by the expression sthine which defines the genitive (in Pāņini's grammar). [The rule aster bhah.) then, teaches that bhū occurs in the sthana of as in certain contexts..." Kahrs even refers to Nägeśa's position, according to which sthana is not a meaning of the genitive case; it rather conditions the relation which is the meaning of the genitive case (pp. 241 f.). He concludes on p. 248: "I think it proper to accept what Kaiyata and Nageša say, which also Annambhata says: the sthāna is called a sambandha 'relation metaphorically, because it is a necessary condition for the relation in question. Note that I have accepted what most Pāņiniyas say, namely that the sthana is not itself a sambandha and thus not something directly conveyed by the genitive ending..." (my emphasis).
And yet, in the Epilogue (p. 268-269) the old position is back into place: "... there is nothing remarkable in interpreting a genitive as a substitutional genitive in the context of relations between linguistic elements. ... Nor is there anything in the discussions of Asstadhyayi) 1.1.49 sasthi sthāneyoga and the nirdiyamana-paribhasa which prohibits such an interpretation within or without the boundaries of vyakarana. ... the substitutional use of the genetive is part of the Sanskrit language so that any genitive in a suitable context could be interpreted in such a way" (my emphasis).
It should be clear that it is possible to have serious doubts with regard to the thesis that ultimately all Indian semantic etymologies are

Page #16

176
Johannes Bronkhorst
Etymology and Magic
177
to be understood according to a substitutional model. Kahrs himself has such doubts. Towards the end of his book he makes the following remarks (p. 278): "In the light of the technical framework of the Nirukta it is of course also possible to adopt a different model for the interpretation of nirvacunas. On the view that the -leh and -eh forms are ablatives one would simply face statements that single out the verbal element which underlies the grammatical formation and identify the action or event considered the reason for a particular name. This is a possibility also if they are considered genitive forms, and it is no longer crucial to determine which case ending we are dealing with." He then adds that perhaps this model and the substitutional model are both simultaneously valid. Perhaps, but one cannot avoid the conclusion that the problem of the Indian semantic etymologies is not definitely solved with the help of the substitution model
about the Azande: "The homoeopathic element is so evident in many magical rites and in much of the materia medica that there is no need to give examples. It is recognized by the Azande themselves. They say, 'We use such-and-such a plant because it is like such-and-such a thing, naming the object towards which the rite is directed." Similar objects are here brought into connection, just as similar words, or parts of words, are connected in traditional etymologies. It may be interesting to see what explanations practitioners and believers offer for the presumed efficacy of magical acts.
It appears that often the people concerned do not think much about this question. Hallpike (1972: 284; cited in Hallpike 1979: 157-58). e.g., maintains that the Konso never refer to any kind of force or supernatural power to explain their belief in the efficacy of their symbolism. Ohnuki-Tierney (1981: 44-45), similarly, observes that among the Ainu materia medica alone, without the involvement of a spiritual being or a ritual, is considered sufficient to effect a cure: "The analogy between the physical characteristics of the beings used in materia medica and the illness... generates the power of what Frazer once called 'sympathetic magic". Among the Nuer-EvansPritchard observes (1956: 104) - magical substances have an efficacy in themselves and do not derive their power from Spirit.
Yet there are cases documented where people do speculate about the reasons why their magic works, or should work. These reasons can be some specific power, or the involvement of spirits. The Azande. Evans Pritchard (1976: 177) observed, "do not think very much about the matter", yet they see that the action of medicines is unlike the action of empirical techniques and that there is something mysterious about it that has to be accounted for. In the case of vengeance-magic, for example, they say that the mbisimo ngua, 'the soul of the medicine'.
It appears, then, that semantic etymologizing is a universal (human) phenomenon which is in need of a universal explanation. In order to make progress in this direction, we may observe that there is a shared element between semantic etymologies and so-called magical acts which are of almost universal occurrence in human societies." Both may be looked upon as expressions of the analogical mode of thought, to borrow a term from the anthropologist SJ. Tambiah. Both in etymologies and in magical acts (or what are often referred to as such) similarity frequently plays a determining role. It must suffice here to cite Evans-Pritchard's (1976: 177) following remark
Some authors (eg, Brian K. Smith, 1989: 36 1.) ohject against the use of the word magic, claiming that this term indicates utter foreignness and difference of the activities concerned that it distinguishes them from proper religion, that it emphasizes their problematic nature, etc. None of this is here intended. No claim is here made that there is such a thing as magic, or that the term has been, or can be meaningfully used. The word has however often been used in academic literature in connection with activities that show some kind of similarity with the etymologies we are studying, The term is here merely used for convenience, without any claim as to the unity or coherence of the activities covered by it
The attitude of the Azunde towards witchcraft is not dissimilar: "But even to the Azande there is something peculiar about the action of witchcraft. They know that it cxists and works evil, but they have to guess at the manner in which it works.... They only know what the others know that the soul of Witchcraft goes by night and dewurs the soul of its victim." (Evans-Pritchard, 1976:31

Page #17

Johannes Bronkhorst
178
• Erymology and Magic
179
has gone out to seek its victim." These rationalizations themselves lead to other activities: "The virtue of a medicine is sometimes spoken of as its soul, and is believed to rise in steam and smoke when it is being cooked. Therefore people place their faces in the steam so that the magical virtue may enter into them. Likewise, Azande say that when they cook vengeance-medicines the soul of the medicine goes up in the smoke from the fire and from on high surveys the neighbourhood for the witch it goes forth to seek." (Evans-Pritchard, 1976: 200). Malinowski (1922: 423) observed that among the Kula the spirits "are not agencies which get to work directly in magic). In the Trobriand demonology, the magician does not command the spirits to go and set to work. The work is done by the agency of the spell, assisted by the accompanying ritual, and performed by the proper magician. The spirits stand in the same relation, as the performer does, to the magical force, which alone is active. They can help him to wield it properly, but they can never become his instruments." On p. 427 Malinowski states that "magic... is a specitic power, ... an inherent property of certain words, uttered with the performance of certain actions... The words and acts have this power in their own right, and their action is direct and not mediated by any other agency." Here the efficacy of magic is explained with the help of certain forces rather than spirits. "Much of the sympathetic magic" (said to underlie) the 'Black Art' of the Malays seems to work by control of spirits", according to Endicott (1970: 174). This is true to the extent that the manipulation of wax figures has been said to serve the purpose of giving the spirits an example of what is expected of them (Wilkinson, 1906: 73). The medicines of the Zulu-the sympathetic associations of which with the desired cllects is often plain (Berglund, 1976: 352 f.) - are believed to contain candla power" (id.. p. 256), which is, however. not traced to any particular source Ip. 257).
These examples show that the effectiveness of magical acts is not always taken for granted, and not therefore in all cases beyond the need of some form of explanation in the eyes of those who carry them out. Yet the explanations offered by the performers in these cases seem completely inapplicable to etymologies. This does not
mean that there are no similarities. We have seen that the Vedic etymologies refer virtually without exception to a mythical reality, and that sometimes a mythrappears to have been created under the influence of the etymology concerned. In both cases a hidden reality is postulated in order to explain the effectiveness of magical acts and of the validity of etymologies respectively.
At this point it will be interesting to consider the theories of Neoplatonism. This is what, in the words of R.T. Wallis (1972: 70), Plotinus (204-270 C.E.) thought about paranormal phenomena: "In these (i.e.. paranormal phenomena) Plotinus, like virtually all his contemporaries, except the most determined atheists and materialists — by the third century A.D. a very rare species--firmly believed. Where he differed from many of them was in attempting to accommodate such phenom ena to a rational, orderly view of the world. The basis of his explanation is the Stoic doctrine of cosmic sympathy', the view that, since the world is a living organism, whatever happens in one part of it must pro duce a sympathetic reaction in every other part. It is by studying and applying the relevant forces that magicians produce their effects. One might cite here Plotinus' Enneads IV.4.40:"But how do magic spells work? By sympathy and by the fact that there is a natural concord of things that are alike and opposition of things that are different, ..." (tr. Armstrong, 1984: 26).
Neoplatonists after Plotinus frequently use the term 'theurgy'." Wallis (1972: 107) observes: "The methods of theurgy were essentially those of ritual magic, its aim the incamation of a divine force either in a material object, such as a statue, or in a human being, the result being a state of prophetic trance. Its justification, most clearly expounded in Proclus' little essay on the Hieratic Art, is the magical Principle of Correspondence', the idea, first that each part of the universe mirrors every other part, and secondly, and more important, that the whole material world is the mirror of invisible divine powers; hence, in virtue of the network of forces linking image to archetype, manipulation of
-Theurgy took as its authoritative basis the Chaldean Oracles which date from the mid-second century C.E.: see Johnston, 1997

Page #18

180
Johannes Bronkhorst
the appropriate material objects brings the theurgists into contact with the deities they represent."
For our present purposes it is interesting to note that some Neoplatonists, among them Proclus and others, extended these ideas beyond magic to the field of language. There is a similarity between words and objects which is of the same type as the similarity which exists between a god and his statue: "Wie die Konsekrationskunst durch gewisse Symbole und geheime Zeichen die Standbilder den Göttern ähnlich macht .... so bildet auch die Gesetzgebungskunst... die Wörter als Standbilder der Dinge, indem sie bald durch solche bald durch andere Laute die Natur der Dinge abbildet." And again: "Just as the demiurgic intellect brings into existence in matter the appearances of the very first Forms it contains in itself, produces temporal images of eternal beings, divisible images of indivisible beings, and from beings which are really beings produces images which have the consistency of shadow, in the same way. I think, our scientific knowledge also, which takes as its model the productive activity of the Intellect, makes by means of discourse similitudes of all the other realities and particularly of the gods themselves:... Since then it produces the names in that way, our scientific knowledge presents them in this ultimate degree as images of divine beings; in fact it produces each name as a statue of the gods, and just as theurgy invoked the generous goodness of the gods with a view to the illumination of statues artificially constructed, so also intellective knowledge related to divine beings, by composition and divisions of articulated sounds, reveals the hidden being of the gods."47
45 Note that a Christian author like Pseudo-Dionysius does not hesitate to describe the eucharist as theurgy. Here "[t]he bread and wine are representations of divine power in the same way that divine names are statues" (Janowitz, 1991: 370). For divine names as 'statues, see below.
46 Procli diadochi in Platonis Cratylum commentaria, ed. G. Pasquali. Leipzig 1908, p. 191. 12 ff. Tr. Hirschle, 1979: 12.
47 Platonic Theology (ed. H.D. Saffrey and L.G. Westerink, Paris 1968 f.) bk. 1, chap. 29. pp. 123-124. Tr. Janowitz, 1991: 368-369. See also Shaw, 1995: 179 ff. ("Naming the Gods").
Etymology and Magic
Since, then, words imitate their objects, one can arrive at a knowledge of objects through words, especially through etymologizing, i.e.. through the constituent syllables of the words (Hirschle, 1979: 20),** Be it noted that Proclus distinguishes three kinds of words: divine. daemonic, and human. Divine words are closest to their objects, they are 'coexistent with them, daemonic words less so, and human words have only limited similarity with their objects. Obviously human words are least capable of consequent etymological analysis. The situation is quite different with the secret names of gods, whose efficacy is the result of specially efficacious combinations of sounds." Hirschle (1979: 27-28) draws in this connection attention to certain secret names of gods found in Greek magical papyri from Egypt, which belong to no known language: "Es sind scheinbar bedeutungslose Namen, nichts anderes als bizarre Lautkombinationen, die bis zu 100 und mehr Buchstaben umfassen können". The parallelism with the meaningless bijamantras of Tantrism is striking. Both in India and in the Hellenistic world, it appears, the search for the elementary constituents of language went hand in hand with the postulation of higher levels of language, in which ordinary meanings are no longer present.
This is of course not the place for a general exposition of Neoplatonic philosophy as a whole, of which the above ideas about magic and etymologizing are part. For our present purposes it suffices to retain the following observation: Neoplatonism explained both the effectiveness of magical rites and the revealing potential of etymologies with the help of one mechanism, that of cosmic sympathy. Cosmic sympathy creates a network that links similar objects, and similar words, to
181
8 This interest in non-historical etymologies is all the more striking in view of the fact that someone like Varro, many centuries before Proclus, seems to have made what he considered were historical etymologies (Pfaffel. 1980; cf. Barwick, 1957: 66 f.; Desbordes, 1991: 150). Regarding Plotinus' views on etymologizing, see Heiser. 1991: 20: "Plotinus himself has no comment to make on Plato's project in the Cratylus. and his occasional use of a Platonic etymology is not enough to indicate his view of the matter."
According to lamblichus, the seven vowels were connatural with the seven planetary gods; Shaw, 1995: 185 (

Page #19

182
Johannes Bronkhorst
Etymology and Magic
183
excite in the mind. And this origin I believe occurs not merely in the primal language, but also in languages that have grown up later in part from the primal language and in part from the new usage of men dispersed over the globe. (Aarsleff, 1969: 88). 52
each other, and to higher forces with which they share features. Cosmic sympathy is possible because the world is a living organism.
Neoplatonic ideas played a major role in the natural magic that exerted much influence in Renaissance Europe (Walker, 1958; Yates, 1964: cf. Hadot, 1982). This magic - as Thomas M. Greene points out (1997: 262) -- requires the assumption that there is an inherent correspondence if not an identity between sign and its object, that there is a natural language and a natural semiotics. It comes as no surprise that sixteenth century critics of magic emphasize the conventional nature of language, which is the exact opposite of the position maintained by the upholders of magic (ibid. p. 255 f.). The same period has an interest in etymology. For Petrus Ramus etymology "means that one looked in it to discover, not the original meanings of words, but the intrinsic properties of letters, syllables, and, finally, whole words" (Foucault, 1966: 35). The situation is similar "in the Neoplatonic and cabalistic exegesis of texts and analyses of language, such as one finds in Pico or Fabio Paolini. You have a significant while, a lexi... which can be analysed into still significant parts, words (or propositions: then you go a stage further and try to find elements of the significance of the whole in single letters.... where in fact they do not exist." (Walker. 1958: 118). The interest in etymologies initially centered on Hebrew, the first language, but then extended to other languages as well (Maillard, 1991; Dubois, 1970: 80 f.). Even Leibniz, though critical of the idea of an original 'Ursprache', as the mystic Böhme called it came to accept a modified form of the Platonic doctrine of the nature of language: 'For languages have a certain natural origin, from the agreement of sounds with the dispositions of the mind (or 'affects"), which the appearances of things
The importance of similarities in "magical" acts was already noted by anthropologists in the last century. We shall here consider some of the ways in which these anthropologists and their successors have tried to make sense of this observation, with the ultimate aim of discovering to what extent their theories can help us to explain etymologies.
Sir Edward Tylor (1865: 124: 1891: 115 f.), followed by Sir James Frazer, ascribed the frequent presence of similarities in acts of magic to a confusion between thought associations and objective connections, to the mistake of taking "ideal connections for real connections". Frazer (1922: 14), in particular, distinguished two principles of thought on which magic is based; he called them the Law of Similarity and the Law of Contact or Contagion. These two give rise to Ho moeopathic or Imitative Magic and Contagious Magic respectively. 54 Frazer explained these principles as "misapplications of the association of ideas" (p. 15). In spite of this, these two principles constituted, for Frazer, a faith, as is clear from the following citation (p. 63-64): "Wherever sympathetic magic occurs in its pure unadulterated form,
Closely similar ideas are found in ancient China (Needham, 1956: 253 f.). Needham points at the connection with Frazer's law of similarity' (p. 280) and with the correlations accepted in Renaissance Europe (p. 296 f.). It is not necessary to believe that these ideas in different parts of the world must be explained by mutual influence, as Needham (p. 297.) tends to think. Rather, it appears that the idea that Similar things act on similar things is a rather obvious rationalization of the presumed effectiveness of certain magical and related facts
53 For a recent survey, see Cunningham, 1999
5* The discussion of magic was introduced in the second edition of The Golden Beagh, which came out in 1900: see Ackerman, 1987: 166 1. Ackerman, referring to EE Evans-Pritchard, calls this analysis of magic "Frazer's single most important contribution to the anthropology of religion".
p. Vickers, 1984: 95: "The occult tradition does not recognize the distinction hetween words and things and between literal and metaphorical language Words are treated as if they are equivalent to things and can be substituted for them." The term pra c htge' has been used Gerl. 1982) to describe the changed
languapehu.cn Locado Bruni 137-144) an Franco Patrici(15291547

Page #20

184
Johannes Bronkhorst
it assumes that in nature one event follows another necessarily and invariably without the intervention of any spiritual or personal agency. Thus its fundamental conception is identical with that of modern science; underlying the whole system is a faith, implicit but real and firm, in the order and uniformity of nature." The Law of Similarity he described as the principle "that like produces like, or that an effect resembles its cause", the Law of Contact or Contagion as the principle "that things which have once been in contact with each other continue to act on each other at a distance after the physical contact has been severed" (p. 14). However, "the primitive magician... never reflects on the abstract principles involved in his actions. With him... logic is implicit, not explicit": "he tacitly assumes that the Laws of Similarity and Contact are of universal application and are not limited to human actions" (p. 15).
Tylor and Frazer have frequently been criticized by more recent anthropologists." "Perhaps one of the most devastating criticisms levelled against Tylor (that is equally appropriate to Frazer) is his never posing the question why primitives would mistake ideal connections for real ones in one domain when they do not do so in their other activities. As Evans-Pritchard puts it- and in this he and Malinowski stand together: The error here was in not recognizing that the associations are social and not psychological stereotypes, and that they occur therefore only when evoked in specific ritual situations, which are also of limited duration..."56
Frazer's 'laws', too, have been severely criticized. Beattie, for example, observed (1964: 206): "Nobody in their senses could possibly believe that all things that share some common quality, and all things that have once been in contact, are continually affecting one another;
55 Cp. Douglas, 1978. Ackerman (1998: 129) observes: "By the late 1960s the reputation[...] of Frazer [was] about as low as [it] could be. Whenever an anthropol ogist interested in the history of the idea of "primitive' religion bothered to consider Frazer, he was regarded as wholly lacking in redeeming intellectual value, the very model of how not to do anthropology or think about religion".
56 Tambiah, 1990: 51, with a quotation from Evans-Pritchard, 1933: 29.
Etymology and Magic
185
in a world so conceived almost everything would all the time he affecting almost everything else, and all would be chaos. Tambiah (1968: 37) remarked, similarly: "... Frazer's principles... lead to absurd inferences about the logic of magic."
Here it is to be recalled that Frazer did not himself subscribe to this conception of the world. Quite on the contrary, he attributed it to those who practise and believe in magic. Frazer's critics are no doubt right in thinking that most magicians and their followers do not entertain such a conception, but saying that "nobody in their senses could possibly believe" in it certainly goes too far. We have seen that a respectable school of philosophy, Neoplatonism, adhered to ideas very similar to those formulated by Frazer, and was capable of inspiring thinkers many centuries later. Indeed, recent research suggests that Frazer may have formulated his theories under the direct or indirect influence of Renaissance thought." In fact, our preceding exposition has shown that also those who tried to give a rational explanation, and justification, of the use of etymologies arrived at views not dissimilar to the ones which Frazer ascribed to his "primitives". While many anthropologists have, no doubt rightly, criticized Frazer for underestimating the amount of common sense in the people he describes, no one seems to have raised the equally valid criticism that he overestimated their desire, or tendency, to create rational systems of thought.
It is of course possible to maintain that Frazer's classification, as classification, leaves to be desired. This is John Skorupski's position, who in his Symbol and Theory (1976) proposes instead the following modified classification: symbolic identification and contagious trans
57 See further Skorupski, 1976: 138 f.
58 Similar ideas existed in China, too: see Henderson, 1984: ch. 1 ("Correlative thought in early China").
Cp. Hanegraaff, 1998: 266: "Frazer's sympathetic magic can be divided into homeopathic (imitative, mimetic) and contagious magic: a distinction which may well have been taken straight from Tylor's Researches who, in turn, could have found it in the great compendium of Renaissance magic: Agrippa's De occulta philosophia" and ibid. n. 47: "Tylor repeatedly quotes Agrippa in Primitive Culture"

Page #21

186
Johannes Bronkhorst
fer. The importance of identification in the first of these two is emphasized: "The symbol in some sense is, or participates in, the reality it represents." (Skorupski, 1976: 144). It is not difficult to see that this modified classification, and in particular the symbolic identification which Skorupski recognizes in a part of so-called magical acts, are as useful as Frazer's Law of Similarity, if not more so, to make sense of etymologies. David Freedberg, speaking about images in particular. maintains that many theories, from Frazer's laws of similarity and contagion to more contemporary notions of sympathy, identification. symbolic linkage, association of ideas, evocative resonance of symbols. or what have you, assume the disjunction between the symbol and the symbolized between representation and reality. But this is precisely what is not given at the level of our emotional and cognitive response to images. Hence he says (Freedberg, 1989: 436): "we will only come to understand response if we acknowledge more fully the ways in which the disjunction... lapses when we stand in the presence of images." Once again it is possible without difficulty to transfer this to the understanding of semantic etymologies.
A more serious criticism would be to doubt the belief of the actors in the efficacy of magical acts. Gilbert Lewis puts it as follows (1994: 568): "Take, for example, sorcery as an example of magical belief. If we assume a man's true and literal belief in his sorcery, then either violence or the sorcery will seem to be ways to harm his enemy. The sorcery might substitute for the violence. But if we slacken the certainty of his belief. impute less of the literal to his statement, then his sorcery action may become that much more of an act which stands for something violent he would like to do but which he does not wholly dare, and perhaps not really desire, to carry out. It is a substitute, but a partial one. And it becomes in part symbolic to the man himself." This position, which seems to be representative of the majority of present
As presented in Sharf, 1999; 85, Sharf's own concern is primarily with Buddhist relics, which are "for all intents and purposes, formless" and do not represent or signify anything. For a presentation of the differences between Freedberg's and Skorupski's positions, see Freedberg, 1989: 274 T
Etymology and Magic
day thinkers in this domain, implies for homeopathic magic that its actors do not really think that similar things are related to each other. A parallelism with semantic etymologies is hard to maintain in this
case.
187
Lewis expresses himself carefully in the above passage ("it becomes in part symbolic to the man himself") and further softens down his position in the next paragraph. He does not totally reject the idea that perhaps some practitioners of magic sometimes believe that there is after all some kind of connection between the substitute and the object it represents. For an analysis of the situation, we must consider S.J. Tambiah's Magic, science, religion, and the scope of rationality (1990). After discussing Tylor and Frazer, Tambiah turns to Malinowski's views on magic, and observes (p. 73):
Malinowski had two specific insights into the internal structure and constitution of Trobriand rites. The first was that they exploited simultaneously both words and acts, both speech and the manipulation of objects and substances, thereby posing the problem of the logic of use of multiple media in ritual for his successors to ponder over. Secondly his so-called 'ethnographic theory of the magical word' proposed some illuminating insights which foreshadowed and anticipated in England Austin's "linguistic philosophical notions of performative force carried by speech acts, that is, how speech acts created both illocutionary and perlocutionary effects by virtue of being conventional acts; and in this Country [= U.S.A., JB, Kenneth Burke's discussion of the 'rhetoric of motives".
Yet Tambiah is not completely happy with Malinowski's position. Observing that "it would seem that we cannot yet completely exorcize the ghosts of Tylor and Frazer", he concedes that magic has a dual structure (p. 82-83):
On the one hand, it seems to imitate the logic of technical/technological action that seeks to transform nature or the world of natural things and manifestations. On the other hand, its structure is also transparently rhetorical and performative (in that it consists of acts to create effects on human actors according to accepted social conventions). Tylor and Frazer fastened exclusively on the first equation
Lewis, 1994: 568: "People must differ individually in how they view the truth of what they assert in common with others in their community... Emotion and feeling as well as reason enter into the link between assertion and conviction..."

Page #22

188
Johannes Bronkhorst
Etwnology and Magic
189
and said it was had science; Malinawski appreciated the force of the second cquation and said that magic was constituted of speech acts in a performative and persuasive mode, and that therefore they were pragmatically reasonable.... The now puzzling duality of magic will disappear only when we succeed in embeddling magic in a more ample theory of human life in which the path of ritual action is seen as an indispensable mode for man anywhere and everywhere of relating to and participating in the life of the world
Tambiah's recognition of the other side of magic would seem to be a major step forward with regard to his own earlier studies (e.g., Tambiah 1968, 1973) and those of others. Tambiah develops this idea further by distinguishing two 'orientations to the world', which he calls participation and causality. He then explains (p. 108): "Although 'causation and participation may seem different or contrastive orientations to the world, the analyst must maintain that both are projected on the experiential and symbolizing capacities of the same sensory modalities of man — the modalities of touch, taste, hearing, seeing.... If participation emphasizes sensory and affective communication and the language of emotions, causality stresses the rationality of instrumental action and the language of cognition. But these are ideal type exag
gerations, and neither can exclude the devices of the other." In other words, at least some of the practitioners of magic do not totally exclude the idea that magic has a causal effect on the intended person or object.
The idea of multiple orientations to the world is plausible, and Tambiah presents a number of arguments which support it. This does not change the fact that his position remains in some respects very close to the one of Tylor and Frazer. These scholars had claimed that magic made the mistake of taking ideal connections for real connections. Tambiah rather speaks of an orientation or ordering of reality, in which people believe in the causal efficacy of communicative acts. In both cases there is a mistake regarding objective reality. And in both cases this mistake finds its origin in the subjective realm.4 It appears, then, that Tambiah's latest attempt, in spite of its obvious
62 See, e.g., Kilani, 1989: 126: "La magie est un langage symbolique, un mode de communication sociale.... L'efficacité d'un acic magique consiste dans le fait de dire des choses sur l'individuc un groupe d'individus qui sont en train d'accomplir une action donnée. La magic a une efficacité sociale, elle peut dans certains cas se transformer en moyen de mobilisation sociale. Similarly Kilani, 1983: passim. Waardenburg, on the other hand, recognizes the ohjective connections that are supposed to underlie magical acts; See Waardenburg, 1986: 196: "Wesentlich bei diesen Völkern ist eine Grundanschauung von Zusammenhangen, die es zwischen den Dingen gibt." (The Dutch version of this book (1990: 203) speaks of "verbanden en samenhangen... die wij in het Westen niet kennen" ("relations and connections which we in the West do not know')) H.E. Brekle, speaking from the point of view of 'popular linguistics', observes (1990: 42): "Ce qui est essentiel pour toutes sortes d'activités magiques à l'aide de moyens langagiers, c'est la foi ou la croyance dans les effets produits par la seule énonciation de certains mots ou de certaines formules. Cela implique que ces croyants' (exécuteurs et victimes') prennent pour assure qu'il existe des rapports nécessaires et causaux, voire des rapports d'identité, entre le can d'une chose et la chose elle-même, ou l'énonciation d'une formule et l'état de choses évoqué par cette formule."
In spite of claims to the contrary, Freedberg would seem to admit the same in the following passage (Freedberg, 1989: 276; the use of 'elide' and 'elision in this passage does not appear to have any of the meanings enumerated in the Webster's and the New Shorter Oxford English dictionaries: the index states under 'Elision of image and prototype': "See also Fusion of image and prototype" (p. 524): "When we see the resembling image, we elide it with the living prototype it represents... This tendency to clision does not happen by some kind of magical process. It is part of cognition and it lies at the root of the helict in the efficy of magical images. Aware of the supervening tendency to abstract and dillerentiate, makers of clumatory or magical images encourage the clision, and set out to preempt the move to differentiation."
This, of course, opens the way to psychological explanations, such as the one ollered by C.R. Halipike (1979: +29), which draws on the work of Page: "Ethnographic literature is replete with examples of the way in which primitives treat mental and bodily conditions and processes, properties and qualities of physical objects, and physical processes, as well as conditions of society such as ill luck, sin, and general ill health, as entities which can be transmitted from man to nature, from one natural object to another, and from natural objects to man, in an enormous variety of ways. The true significance of this cognitive phenomenon is not so much that it is a case of Frazer's homeopathic magic, of like producing like. as of the pre-operatory propensity to isolate particular phenomena and treat them as bounded entities which can he klached from their physical context with solute properties and an inner dynamism of their own. The reification of p ens in particular is a notable example of this proclivity of mind..For an attempt to provide an evolutionary explanation

Page #23

190
Johannes Bronkhorst
Etymology and Magic
191
Strong sides, has indeed not been able to exorcise the ghosts of Tylor and Frazer
In a way Tambiah's latest position is not very different from the one presented by Skorupski already in 1976, in a work mentioned earlier. Skorupski, to discusses the relationship between magic and perfor mative acts, he speaks in this connection of operative actions. While emphasizing the identificationist view of symbolic magic, he concedes that magic can be incorporated within operative theory if it is inter preted as a way of triggering a consequential action, that is as a way of signalling what is being done and thereby doing it" (Skorupski, 1976: 153: cp. Cunningham, 1999: 84). It is further interesting to note that L. Wittgenstein, who often appears to hold an expressive, antiinstrumental view of magic, elsewhere provides the most persuasive arguments against his own anti-instrumental objections, so that what we have is more a matter of genuine ambivalence than simple incon
sistency 66
mean that all things are related to or identical with everything that they resemble, even though this belief has occasionally been maintained (e.g. in Renaissance magic), as we have seen. The apparently widespread conviction that similarity can indicate relatedness or identity between things allows us to make sense of semantic etymologies. These etymologies are meant to reveal the connections that exist between the words concerned, and consequently between the things denoted by them.
We must be careful not to attribute explicit convictions of this kind to all those who use semantic etymologies. This would be as mistaken as attributing similar ideas to those who practice homeopathic (or imitative) magic; this was Frazer's mistake, for which he has repeatedly been chided, as we have seen. Most users of semantic etymologies will not have any systematic world view that explains their assumed validity. In this respect they contrast with those users of magic who often "explain the presumed efficacy of magical acts with the help of notions of spirits or something of the kind. An exception is constituted by certain Neoplatonic thinkers who, as we have seen, elaborated a view of the world in which similarities, also between words, played a central role
How about the other explanation of magical acts that we have discussed? Is the notion of speech acts in a performative and persuasive mode' able to account for semantic etymologies? This is unlikely. Se mantic etymologies are not performative acts and have no persuasive validity, as far as I can see; they certainly don't in early and classical Indian literature. Their aim appears to be to bring to light existing connections or identities (i.e., connections or identities that are presumed to exist), not to bring about new connections or to persuade others.
This leads us to the following conclusion. Semantic etymologies share with many acts of so-called sympathetic magic the underlying belief that similar things can be related to, or even identical with, each other. This belief is not normally systematized (with some rare exceptions), and indeed it is rarely formulated. It is for this reason perhaps better to speak of it as an intuition rather than as a consciously held belief. There is no claim that all similar things are related to each
Let us now return to our main subject matter, semantic etymologies Does the above discussion about magical acts help us to understand these etymologies better? It does if we assume - with Tambiah, with Skorupski and. yes, with Frazer - that at least in certain cases magical acts are believed to transform nature or the world of natural things and manifestations" (Tambiah). In other words, things are accepted -in specific circumstances perhaps - to be related to or identical with certain other things which they resemble. This does not have to
of the two laws of sympathetic magic as observed in the psychology of disgust, see Pinker. 1997: 378 ff.
68 One may also wonder - as did Sharpe a quarter century ago (1975: 94) - whether perhaps the time is now approaching when fashionable impatience with Frazer will give place to a sober estimate of his contribution to comparative religion in its anthropological aspect. He may then prove to have been greater rather than smaller than we thought."
Ciorti, 1998: 155-182 ("Wittgenstein on making homeopathic magic clear"); quoted sentence on p. 156.

Page #24

192
Johannes Bronkhorst
Etymology and Magic
193
other (again with some rare exceptions), and it may be difficult to discem in each particular case why certain things rather than others are assumed to be thus connected. This intuition does not only account for the almost universal occurrence of sympathetic magic, but also for the equally quasi-universal use of semantic etymologies.
JOHANNES BRONKHORST Section de langues et civilisations orientales Université de Lausanne BFSH 2 CH-1015 Lausanne
REFERENCES
Aarsleff, Hans 1969 "The study and use of etymology in Leibniz." (Studia Leibnitiana Sup
plementa, vol. II: Erkenntnislehre. Logik. Sprechphilosophie. Editionsberichte. Wiesbaden: Franz Steiner. Pp. 173-89.) Reprinted: From Locket
Saussure. London: Athlone. 1982. Pp. 84-100. Ackerman, Robert
1987 J.G. Frazer: His life and work. Cambridge University Press, 1990 Ackerman, Robert 1998 "J.G. Frazer and the Cambridge Ritualists and the scientific study of
religion." Religion in the Making: The Emergence of the Science of Religior. Arie L. Molendijk & Peter Pels, Eds. Leiden etc.: Brill. (Studies in the
History of Religions (Numen Book Series), 80.) Pp. 129-158 Armstrong. A.H. 1984 Plotinus, with an English translation, IV: Enneads IV. 1-9. Harvard Univer
sity Press. (Loeb Classical Library. Balbir, Nalini 1991 "Le discours etymologique dans l'heterodoxie indienne" Discount en
mologique. Actes du Colloque international organisé à l'occasion du contenaire de la naissance de Walther von Wartburg Edités par Jean-Pierre Chambon et Georges Lüdi, avec la collaboration de Hans-Martin Gauger. Frank Lestringant, Georges Pinault (Max Niemeyer Verlag. Tübingen 1991). Pp. 121-134
Barwick, Karl 1957 Probleme der storischen Sprachlehre und Rhetorik. Berlin: Akademie
Verlag. (Abhandlungen der Sachsischen Akademie der Wissenschaften zu
Leipzig, Philologisch-historische Klasse, Band 49. Heft 3.) Beattie, John 1964 Other Cultures. Aims. Methods and Achievements in Social Anthropology
London: Routledge & Kegan Paul, 1967. Berglund. Axcl-Ivar 1976 Zal Thughr-Pultenes and Symbism. London: C. Hurst. (Studia Mission
alia Upsulinsin No XXIII Bhavasar
1969 Ermology in the Brāla s, Poona (unpublished thesis, unseen) Böhl, Felix 1991 "Die etymologische Namensdeutung in der rubbinischen Literatur." Dix
cours étymologique. Actes du Colloque international organisé à l'occasion du centenaire de la naissance de Walther von Wartburg. Edités par JeanPierre Chambon et Georges Lüdi, avec la collaboration de Hans-Martin Gauger, Frank Lestringant. Georges Pinault (Max Niemeyer Verlag. Tuhan
gen 1991). Pp. 161-172. Brekle. Herhert E. 1990 "La linguistique populaire." In: Histoire des idees linguistiques Tome I: LA
Maissance des métalungages en Orient et en Occidenr. Sylvain Auroux, Ed.
Liège-Bruxelles: Pierre Mardage. Pp. 39-44. Bronkhorst, Johannes 1981 "Nirukla and Astadhyayi: their shared presuppositions." Indo-Iranian
Journal 23, 1-14. Bronkhorst, Johannes
1984 "Nirukta, Undi Sotra, and Agdhyāyi." Indo-Iranian Jouw 27.1-15. Bronkhorst, Johannes
1989 "Veda." Annals of the Bhandarkar Oriental Research Institute 70, 125-135. Canimann. Schuyler 1962 Substance and Symbol in Chinese Tegles. Chinese Belt Toggles frein the
C.F. Bieber Collection. Philadelphia: University of Pennsylvania Press Carna. George
1976 Pomiar. A suite w research. The Hague - Paris: Meutun. Cioffi. Frank 1998 Wirtgenstein on Freud and Frucer. Cambridge cte: Cambridge University
Press

Page #25

194
Cunningham, Graham
1999 Religion and Magic. Approaches and theories Edinburgh: Edinburgh University Press.
Dange. Sindhu S.
Johannes Bronkhorst
1989 Puranic Etymologies and Flexible Forms. Aligarh: Viveka Publications. Deeg, Max
1995 Die altindische Etymologie nach dem Verständnis Yäska's und seiner Vorgänger: Eine Untersuchung über ihre Praktiken, ihre literarische Verbreitung und ihr Verhältnis zur dichterischen Gestaltung und Sprachmagie. Dettelbach: Josef H. Röll. 1995. (Würzburger Studien zur Sprache und Kul Jur. 2).
Desbordes, Françoise
1991 "La pratique étymologique des poètes latins à l'époque d'Auguste." Discours étymologique. Actes du Colloque international organisé à l'occasion du centenaire de la naissance de Walther von Wartburg. Edités par JeanPierre Chambon et Georges Lüdi. avec la collaboration de Hans-Martin Gauger. Frank Lestringant, Georges Pinault (Max Niemeyer Verlag. Tühingen 1991). Pp. 149-160.
Deshpande, Madhav M.
1997 Saunakiya Caturädhyāyikā. A Prärisäkhya of the Saunakiya Atharvaveda. with the commentaries Caturädhyayibhāṣya, Bhargava-Bhaskara-Vṛtti and Pancasandhi: critically edited, translated & annotated. Cambridge, Massachusetts and London, England: Harvard University Press (HOS, 52.).
Devasthali. G.V.
1965 Religion and Mythology of the Brahmanas, with particular reference to the Satapatha-Brahmana. Poona: University of Poona. (Bhau Vishnu Ashtekar Vedic Research Series. 1.)
Dikṣā. Saroja
1989 Aitareya evam Taittiriya Brahmanom ke Nirvacana. Delhi: Nag Publishers. Douglas, Mary
1978 "Judgments on James Frazer." Daedalus (Journal of the American Academy of Arts and Sciences), 107(4), 151-164.
Dubois, Claude-Gilbert
1970 Mythe et language au seizième siècle. Bordeaux: Ducros
Eggeling. Julius (tr.)
1882-1900 The Satupatha-Brahmana, according to the text of the Madhyandina school. 5 parts. Motilal Banarsidass, 1978-79.
Etymology and Magic
Eliade, Mircea
1986 Histoire des croyances et des idées religieuses. 3. De Mahomet à l'âge des Réformes. Paris: Payot
Endicott, Kirk Michael
1970 An Analysis of Malay Magic. Oxford University Press. 1981. Evans-Pritchard, E.E.
195
1933 "The intellectualist (English) interpretation of magic." Bulletin, Faculty of Arts, voll. 1, part 2 (Cairo, Egypt: Farouk University) pp. 282-311. (not seen) Evans-Pritchard, E.E.
1956 Nuer Religion. Oxford: Oxford University Press. Evans-Pritchard, E.E.
1976 Witchcraft, Oracles and Magic among the Azande. Abridged with an introduction by Eva Gillies. Oxford: Clarendon Press, 1983.
Falk. Harry
1993 Schrift im alten Indien. Ein Forschungsbericht mit Anmerkungen. Tübingen: Gunter Narr. (ScriptOralia. 56.)
Foucault. Michel
1966 The Order of Things. A translation of Les Mots et les Choses. New York: Vintage Books. 1973.
Franke, Otto
1913 Dighanikaya. Das Buch der langen Texte des buddhistischen Kanons. Göttingen: Vandenhoeck & Ruprecht.
Frazer, James George
1922 The Golden Bough. A Study in Magic and Religion. Abridged Edition. London and Basingstoke: Macmillan. 1974.
Freedberg, David
1989 The Power of Images. Studies in the history and theory of response. Chicago and London: University of Chicago Press.
Gaiser, Konrad
1974 Name und Sache in Platons Kratylos. Heidelberg: Carl Winter. (Abhandlungen der Heidelberger Akademie der Wissenschaften, Philosophischhistorische Klasse, Jahrgang 1974, 3. Abhandlung.)
Gerl, Hanna-Barbara
1982 "Humanistische und geometrische Sprachphilosophie: Ein Paradigmenwechsel von Leonardo Bruni zu Francesco Patrizi." Zeitschrift für Philosophische Forschung 36, 189-207.
Gonda, Jan
1955 "The etymologies in the ancient Indian Brähmanas." (Lingua 5, 61-86.) Reprint: Selected Studies. Volume II. Leiden: E.J. Brill. 1975. Pp. 32-57.

Page #26

196
Johannes Bronkhorst
Etymology and Magic
197
Gonda, Jan 1988 Mantra Interpretation in the Sarapatha-Brahmana. Leiden: EJ Brill
(Orientalia Rheno-Traiectina, 32.) Gonda, Jan 1991 The Functions and Significemce of Gold in the Veda. Leiden: E.J. Brill.
(Orientalia Rheno-Traiectina. 37.) Greenc, Thomas M 1997 "Language, signs and magic. Envisioning Mugic. A Princeton seminar und
symposium, Peter Schäfer and Hans G. Kippenberg. Eds. Leiden etc.: Brill.
(Studies in the History of Religions (Namen Book Scries). 75.) Pp. 255-272 Hadox, Pierre 1982 "L'amour magicien'. Aux origines de la notion de 'magia naturalis':
Platon, Plotin, Marsile Ficin." Revue Philosophique de la France et de
l'Étranger 107,283-92 Hallpike, C.R.
1972 The Konso of Ethiopia. Oxford: Clarendon Press. Hallpike, CR
1979 The Foundations of Primitive Thoughr. Oxford: Clarendon Press. Hanegruall, Wouter. 1998 "The emergence of the academic study of magic the occult philosophy in
Tylor and Frazer." Religion in the Making: The Emergence of the Science Religion. Arie L. Molendijk & Peter Pels, Eds. Leiden etc.: Brill. (Studies
in the History of Religions (Numen Book Series), 80.) Pp. 253-275. Heiser, John H. 1991 Logos und Language in the Philosophy of Plotinu. Lewiston / Queenston
/ Lampeter: Edwin Mellen. (Studies in the History of Philosophy. 15.) Henderson, John B. 1984 The Development and Decline of Chinese Cosmology. New York: Columbia
University Press. Hinüber, Oskar von 1989 Der Beginn der Schrift wid frilie Schriftlichkeit in Indien. Stuttgart:
Franz Steiner Verlag Wiesbaden. (Abhandlungen der Akademic der Wise Senschaften und der Literatur, Mainz, Geistes- und Savialwissenschaftliche
Klasse, Jahrgang 1989, Nr. 11.) Hirschle, Maurus 1979 Spruchphilophie wnd Numenmagie in Newpletemismus. Meisenheim am
Glan: Anton Hain (Beiträge zur klassischen Philologie, Heft 96.)
Holdrege, Barbara A 1994 "Veda in the Brahmanas: Cosmogonic paradigms and the delimitation
of canon." Authorin: Anxiety and Canon. Essays in Vedic interpretation, Laurie L. Patton, Ed. Albany: State University of New York Press. Pp. 35
66. Houben, Jan E.M. 1997 "The Sanskrit tradition." The Emr e Semantics in for Linguistic
Traditions: Hebrew: Sanskrit. Greek, Arabic. By Wout van Bekkum, Jan Houben, Ineke Sluiter and Kees Versteegh. Amsterdam/Philadelphia: John Benjamins. (Amsterdam Studies in the Theory and History of Linguistic
Science, 82.) Pp. 51.145. Home, Robert Ernest (tr.) 1931 The Thirteen Principal Upanishads, translated from the Sanskrit. Second
edition, revised. Oxford University Press, 1975. Jacobsen, Knut A. 1999 Prakri in Samkhya-Yoga. Material principle, religious experience, ethical
implications, New York etc.: Peter Lang. (Asian Thought and Culture. 30.) Janowitz. Naomi 1991 "Theories of divine names in Origen and Pseudo-Dionysius." History of
Religionis 30(4), 359-372. Johnston, Sarah lles 1997 "Rising to the occasion: theurgic ascent in its cultural milieu. Emisioning
Margic. A Princeton seminar and swiprasinet. Peter Schäfer and Hans G. Kippenberg, Eds. Leiden etc.: Brill. (Studies in the History of Religions
(Numen Book Series). 75.) Pp. 165-194 Kahrs, Eivind
1983 Yaska's use of kasedi." Indo-Tremium Journal 25.231-37 Kahrs, Eivind 1984 "Yaska's Nirukta: the quest for a new interpretation." Indologicu Taurinen
sia 12 (Proceedings of the Scandinavian Conference-Seminar of Indological
Studies, Stockholm, June 1st-5th, 1982). 139-154. Kahrs, Eivind 1998 Indian Semantic Analysis: the nincung tradition. Cambridge etc.: Cam
bridge University Press. (University of Cambridge Oriental Publications,
55.) Kantawala.SG 1967 "Puranic etymologics." Kaviraj Abhinanda Gravithur. Ed. Būhūrāmu
Saksenā et al. Lucknow. Akhila Bharatiya Samskrta Parisad. Pp. 278-280

Page #27

198
Johannes Bronkhorst
Erymology and Magic
199
Kantawala, S.G 1971 Puranic etymologies on the cable Purana). Ilgal
S
. Bale Derrick Sanskrit. Chersity of Delhi, sol.2.11 . August
1973. pp. 13 L. (xscent Kaniawala.SG
1993 "Parām etymologies some remarks Purina 3521 171-176. ma i Keith, Arthur Berniedale (cd. & r.)
19019 The Allure Anapukur. New Delhi: Master Publishers, 1981 Kilani, Mondher 1983 Les cules de cargamelandslens. Mythe et romlik en hulle
Lausanne: Editions d'en bas. Kilani, Mondher
1989 hardiction à 'amp ogle. Lausanne: Payot Kromer. S.N.
1969 "Sumerian myths and epic tales. Pritchard, 1969: 37-59. Kraus, Manfred 19N7 Nuwe w Suche: Ein Pwlew im friherichischen Denken. Amsterdam.
BR. Grüner (Studien zur antiken Philosophie, 14.) Kudelska, Marta 1995 "Etymology of the word 'satyam in the light of the concept of Brahman."
In: International Conference on Sanskrit and Related Studies, September 23-26, 1993 (Proceedings). Cracow Enigma Press (Cracow Indological
Studies, 1), 1995. Pp. 179-188. Lallo, Jean 1991 "L'étymologie en Grèce ancienne d'Homère aux grammairiens alexan
drins." Dieselmologique. Actes du Colloque international organisé à l'excasion du centenaire de la naissance de Walter von Wartburg. Edités par Jean-Pierre Chambon et Georges Lidi, avec la collaboration de HansMartin Gauger. Frank Lestringant, Georges Pinault (Max Niemeyer Verlag
Tubingen 1991). Pp. 135-148 Lewis, Gilbert 1994 "Magic, religion and the rationality of beliel." Companio Encopedia of
Anthropology. Tim Ingold, Ed. London and New York: Routledge. Pp. 563
540 Limaye. V.P. 1974 Critical Studies on the Mahabhasvu. Hoshiarpur: Vishveshvaranand Vedic
Research Institute
Lorenz, Maria
1961 "Problem pred hy schizophrenic language." Archives of General Psychi •
4.603.10. Maillard, Jean-François 1991 "L'autre vérité: le discoursemithologique chez les kabbalistes chrétiens de
La Renaissance Diwwwr o gique Actes du Colloque international organisé à l'occasion du centenaire de la naissance de Walther von Wart. burg Edités par Jean-Pierre Chambon et Georges Ludi, avec la collaboration de Hans-Martin Gauger, Frank Lestringunt, Georges Pinault (Max Niemeyer
Verlag. Tubingen 1991). Pp. 185-199. Malaise. Michel 1983 "Calembours et mythes dans l'Egypte ancienne." Le mythe Son langage et
Son message. H. Limet & J. Ries, Eds. Louvain-la-Neuve: Centre d'Histoire
des Religions. Pp. 97-112 Malamoud, Charles 1989 Cuire le monde. Rire et pensée dans l'Inde ancienne. Paris: Éditions la
Découverte Malinowski, Bronislaw
1922 Argonauts of the Western Pacific New York: EP Dutton. 1961. Malinowski, Bronislaw 1935 Coral Gardens and Their Magie. Volume 1: SoilTilling and Agricultural
Rites in the Trobriand Islands, Volume II: The Language of Magic and Gardening. London: George Allen and Unwin. Second Edition 1966. Indiana University Press, Indiana University Studies in the History and Theory of
Linguistics.) Mehendale, MA 1963 "Upanisadic etymologies." Munshi Indological Felicitation Volume. Bom
hay: Bharatiya Vidya Bhavan. Jayantkrishna H. Dave et al., Eds. (Bharatiya
Vidya 20 & 21, 1960 & 1961). Pp. 40-44 Needham, Joseph 1956 Science and Civilisation in China. Il History of Scientific Thought.
Cambridge University Press. Norman, K.R. 1980 Fourtymologies from the Sabhiya-sutta." Festschri for Walpola Rahula.
Somaratna Balasooriya et al., Eds. London: Gordon Frazer. Pp. 173-184. Reprint: Collected Papers, vol. II. Oxford: Pali Text Society. 1991. Pp. 148161. (references to the reprint)

Page #28

200
Johannes Bronkhorst
Etymology and Magic
Oberlies, Thomas 1998 Historische Grammarik des Hindi Die Genese seines morphologischen
Systems an dem Mittelwul Aliadischen Reinbek: Inge Werler. (Indol
ogische Lehrmaterialien, 3. Ohnuki-Tierney, Emiko 1981 illness and Healing among the Sukhlin Ainu. A symbolic interpretationi.
Cambridge etc: Cambridge University Press. Padoux, Andre 1990 Vac The concept of the word in selected Hindu Tantras. State University
of New York Press Pfaffel, Wilhelm 1980 "Prinzipien der varronischen Etymologie." Lautgeschichte und Etymolo
gie. Akten der VI. Fachtagung der Indogermanischen Gesellschaft, Wien. 24.-29. September 1978. Herausgegeben von Manfred Mayrhofer, Martin
Peters, Oskar E. Pfeiffer. Wiesbaden: Ludwig Reichert. Pp. 353-66. Piaget, J. 1925 "Le réalisme nominal chez l'enfant." Rowe Philosophique de la France et
de Enger 44. 189-234. Pinker, Steven
1997 How the Mind Wwks. Allen Lane, The Penguin Press, 1998, Plato: Cratylus. 1) With an English translation by H.N. Fowler. The Loeb Classical Li
brary 167 (Plato IV). Cambridge, Massachusetts. Harvard University Press: London: William Heinemann. 1977. (This translation has been followed in
the text.) 2) A commentary on the text is given in Rijlaarsdam, 1978 Pritchard, James B. (ed.) 1969 Ancient Near Eastern Texts relating to the Old Testament. Third Edition
with Supplement. Princeton, New Jersey: Princeton University Press Rank, Louis Philippe 1951 Enmologiseering en Verweile Verschijnselen bij Homeru. Dissertation
Utrecht Rijlaarsdam. Jetske C. 1978 Plato über die Sprache. Ein Kommentar con Karyles. Utrecht: Bohn.
Scheltema and Holkema. Ruege, David Seyfort 1959 Contributions à l'histoire de la philosophiae linguistique indienne. Paris: E.
de Boccard. (Publications de l'Institut de Civilisation Indienne, Fasc. 7.)
Sander-Hansen, C.E. 1946 "Die phonetischen Wortspicle des ültesten Ägyptischen." Actu Orientali
20), 1-27. Schimmel, Annemarie
1975 Mystical Dimensions of Islam. University of North Carolina Press Schneider. Ulrich 1954 "Acht Etymologien aus dem Agara-Sutta." Asiatica. Festschrift Friedrich
Weller. Zum 65. Geburtstag gewidmel von seinen Freunden. Kollegen und Schülern. Johannes Schubert and Ulrich Schneider, Eds. Leipzig: Otto
Harrassowitz, 1954. Pp. 575-583 Scholem, Gershom 1983 Le nom et les symboles de Dieu dans la mystique juve. Paris: Éditions du
Cerf. Sharf, Robert H. 1999 "On the allure of Buddhist relics." Representations (University of Califor
nia) 66, 75-99, Sharpe. Eric.
1975 Comparative Religion. A history. London: Duckworth. Shastri, Satya Val 1997 "Etymologies in the Yogavasisha." Lex el Lilllerie Studies in Honour of
Oscar Botto. Siegfried Lienhard and Irma Piovano, Eds. Alessandria (1):
Edizioni dell'Orso. Pp. 469-473. Shaw. Gregory 1995 Theurgy and the Soul. The Neoplatonism of lamblichus. University Park.
Pennsylvania: Pennsylvania State University Press. Simson, Georg von 1988 "Etymologie als Mittel ideologischer Auseinandersetzung: Bemerkungen
zum Ageunnusuttu des Dighanikaya." Saudia Indogermania Slova. Festgahe für Wemer Thomas zum 63 Geburtstag München: Oke Sagner
1988. Pp. 97.98 Singh, Fatal 1952 The Vedir Erww.logy. A critical evaluation of the science of etymology as
found in Vedic literature. Kola: The Sanskriti-Sudan. Singh. Maan
1994 The Upanisdie Ermologies. Delhi: Nirmal Publications. Skorupski, John 1976 Symbol and Theory. A philosophical study of theories of religion in social
anthropology. Cambridge University Press.

Page #29
202 Johannes Bronkhorst Erymology and Magic 2013 203 , Anika Smith, Brian K. 1989 Reflections on Resemblance, Ritual and Religion. New York -- Oxford: Oxford University Press. Staal, Frits 1979 "Oriental ideas on the origin of language." Journal of the American Oriental Society 99, 1-14. Strobach, Anika 1997 Plutarch und die Spruchen. Ein Beitrag zur Fremdsprachenproblematik in der Antike. Stuttgart: Franz Steiner. (Palingenesia, 64.) Tambiah, Stanley Jeyaraja 1968 "The magical power of words." = Tambiah, 1985: 17-59. (Originally published in Man 3 (2), 175-208.) Tambiah, Stanley Jeyaraja 1973 "Form and meaning of magical acts." = Tambiah, 1985: 60-86. (Originally published in Modes of Thought, ed. Robin Horton and Ruth Finnegan (London: Faber and Faber, 1973), pp. 199-229.) Tambiah, Stanley Jeyaraja 1985 Culture, Thought, and Social Action. An anthropological perspective. Harvard University Press. Tambiah, Stanley Jeyaraja 1990 Magic, Science, Religion, and the Scope of Rationality. (The Lewis Henry Morgan Lectures 1984.) Cambridge University Press. Thieme, Paul 1935 "Zur Datierung des Panini." Zeitschrift der Deutschen Morgenlandischen Gesellschaft 89, *21* - *24* = Kleine Schriften (1971), pp. 528-531. Thieme, Paul 1985 "The first verse of the Trisaptiyam (AV, S 1.1 ~ AV, P 1.6) and the beginnings of Sanskrit linguistics." Journal of the American Oriental Society 105(3), 559-565. = Kleine Schriften II (Franz Steiner, Stuttgart 1995), pp. 932-938. Tsuji, N. 1977 "Etymologia upanishadica." Melanges d'etudes vediques (Tokyo), pp. 19 39. (not seen) Tylor, Edward Burnet 1865 Researches into the Early History of Mankind and the Development of Civilization. London: John Murray. Tylor, Edward Burnet 1891 Primitive Culture. Third cdition, revised. Vol. I. (The first edition appeared in 1871.) London: John Murray. van den Bosch, Lourens P. . 1993 "Friedrich Max Muller: een Victoriaans geleerde Nederlands Theologisch Tijdschrift 47, 107-118 & 186-200. Verma, Nargis 1991 The Etymologies in the surupatha Brahmana. Delhi: Nag Publishers. Vickers, Brian 1984 "Analogy versus identity: the rejection of occult symbolism, 1580-1680." Occult and Scientific Mentalities in the Renaissance. Ed. Brian Vickers. Cambridge etc.: Cambridge University Press. Pp. 95-163. Waardenburg, Jacques 1986 Religionen und Religion. Systematische Einfuhrung in die Religionswis senschaft. Berlin - New York: Walter de Gruyter. (Sammlung Goschen 2228.) Waardenburg, Jacques 1990 Religie onder de loep. Systematische inleiding in de godsdienstweten schap. Baarn: Ambo. Walker, D.P. 1958 Spiritual and Demonic Magic. From Ficino to Campanella. London: The Warburg Institute, University of London. Wallis, R.T. 1972 Neoplatonism. London: Duckworth. Werner, Heinz, and Kaplan, Bernard 1963 Symbol Formation. An organismic-developmental upproach to language and the expression of thought. New York-London-Sydney: John Wiley & Sons. Wilkinson, R.J. 1906 Malay Beliefs. London: Luzac & Co. Leiden: E.J. Brill. (The Peninsular Malays, I.) Wilson, John A. 1969 "Egyptian myths, tales, and mortuary texts." = Pritchard, 1969: 3-36. Witzel, M. 1979 On Magical Thought in the Veda. Leiden: Universitaire Pers. Yates, Frances A. 1964 Giordano Bruno and the Hermetic Tradition. London: Routledge and Kegan Paul; Chicago: The University of Chicago Press, 1982. Zaleski, Carol 1988 Otherworld Journeys. Accounts of near-death experience in medieval and modern times. Oxford University Press. .