Vedic and Philological Studies

The Origins of Aryan Speech

In that pregnant period of European knowledge when physical Science, turned suddenly towards its full strength, was preparing to open for itself the new views, new paths and new instruments of discovery which have led to the astonishing results of the nineteenth century, an opportunity was offered to the European mind for a similar mastery of sciences other than physical. The Sanscrit language was discovered. It was at first imagined & expected that this discovery would lead to results as important as those which flowed from the discovery of Greek literature by Western Europe after the fall of Constantinople. But these expectations have remained unfulfilled. European knowledge has followed other paths and the seed of the nineteenth century has been Newton’s apple and not Sir William Jones’ Shakuntala or the first edition of the Vedas. The discovery of Sanscrit has, it is true, had a considerable effect on the so-called Sciences of Comparative Philology, Comparative Mythology, Science of Religion, ethnology and sociology; but these branches of knowledge are not sciences, they are systematised speculations. Their particular conclusions often change from generation to generation and none of them, not even the most certain, have the same cast of certainty as a scientific generalisation in the domain of physical inquiry. The law of gravitation is a permanent truth of science; the law that all myths start from the sun, the law of Solarisation, if I may so call it, is an ingenious error which survives at all only because it pleases the poetic imagination.

So great has been the failure that the possibility, even, of a Science of speech has been too readily scouted. But this is an excessive deduction, the reaction of disappointed expectation has exaggerated the meaning of the failure. To say that there can be no Science of speech is to say that the movements of the mind are not governed by intelligible processes, but rather by

Page 523

an incalculable caprice—a supposition that cannot be admitted. Intelligible processes there are for all natural action, and therefore also for mind & its instrument speech; the difficulty is to discover their precise lines of action. We need sufficient material and the right material; we need also the right way of approaching the material. Once that is discovered, the processes also will be discovered & the Science of speech founded.

In this volume I give the result of certain attempts I have made to approach the problem from a new standpoint. Sanscrit, I hold, is the key to the problem. In most other languages, we have a secondary or tertiary speech formation; we have to go beyond the actual form before us & reconstruct its parent tongue, to find again perhaps that the parent tongue has to be subjected to a similar reconstructive process. We have not sufficient materials for such a task; no instruments we possess can go deep enough. But Sanscrit, by a peculiar fidelity to its origins, presents us with a true primary form of speech, in which the vocabulary indeed is late,—a new structure of word flesh & tissue,—but the base of the structure is primitive, reveals the roots of its being and betrays the principles of its formation. The failure of the comparative philologists to make use of their opportunity has been signal; it has even obscured the issue by creating a false system; but it is no more fatal or final than the failure of Aristotle to discover the law of gravitation, although he must have seen many fruits falling from many boughs. Sanscrit still stands there offering to us its secret.

The fundamental mistake of the philologists is contained in their famous original formula, pitā, patēr, pater, Vater, father, and the hasty conclusions they have drawn from it which have prevented a deeper scrutiny of the roots of language. An identity of words between various languages can never in itself lead to any fundamental discovery. It does not even prove that the languages thus agreeing are of a single stock. In many of the most common domestic terms Tamil and Sanscrit agree, but they are still held to be of different families. All therefore that is proved is not the identity of these tongues but their contact—so close a contact of one with the rest that a number of the commonest

Page 524

ideas & relations came in all to be expressed by terms borrowed from one. Nothing more is proved; we have not advanced a single step towards a science of languages. Even the classification of tongues as Aryan, Dravidian, Semitic cannot be called scientific; it is empirical and depends upon identities which may not be fundamental. We must go deeper. European philology has started from word-identities and identities of final word meaning. I propose to start from root-identities and identities of original & derivative root-meaning and even from sound-identities and identities of fundamental and applicatory sound meaning. It is, I believe, possible in this way to establish the unity of the Aryan tongues and some at least of the laws governing the birth & development of Aryan speech.My inquiry does not carry me farther. I do not pretend as yet to inquire into the laws of speech itself, but only to establish from data, some facts of Aryan speech which may eventually help in solving the wider problem.

In another respect, also, the philologists seem to me to have misunderstood the conditions of their inquiry. They have been not rigid enough and yet too rigid. They have been too rigid in not allowing for the flexibility of mind movements. They have sought for the same invariable sequence which we observe in the physical world and admitted a law only where such sequence seemed to occur. The laws of physical formation follow a fixed line and their variations even are after a fixed fashion. But with the growth of life in matter there comes a growing element of freedom, of a more elusive principle & a more elastic variation; for this reason Science has found life more difficult to fathom & analyse than matter and her triumphs here have been far less notable than in the pure physical domain. Mind brings with [it] a still freer play, a still more elusive principle and flexible application. A general law always obtains, but the application, the particular processes vary more subtly and are more numerous.Science, not taking into account this law of increasing freedom, has in the domain of mind accomplished little or nothing.When we deal with the laws of speech, we must remember this flexibility of all mind-processes. We must ourselves keep a flexible mind to follow it and an open eye for all variations. It is for

Page 525

regularity in irregularity that we must always be on the watch, not for a fixed, a cast-iron regularity. On the other hand the few laws which Philology has admitted, have been, by a sort of false compensation for their original narrowness, used with too free and even lax a play of fancy. Often indeed instead of working as a law, the philological principle presents itself as an ingenious means for inventing word-identities.

I have disregarded as another error of imperfect inquiry the rigid philological divorce of the Dravidian & Aryan languages. Whether there be a separate Dravidian stock or no, it is to me a certainty that Tamil owes not only many of its most common terms, but whole families of words to the original Aryan speech. Its evidences cannot be neglected in such an inquiry as I have undertaken, for they are of the greatest importance. Indeed the theory worked out by me, took its rise originally not from any analysis of the Sanscrit word-system, but from an observation of the relations of Tamil in its non-concretised element to the Greek, Latin & Northern Indian languages. At the same time it is on an analysis of the Sanscrit word-system that I have chiefly relied. I have omitted from that system most of its Vedic elements. The meanings of Vedic words are often extremely disputable & it would be unsafe to rely whether on the significances fixed by the European scholars or on those fixed centuries ago by Sayana or even by Yaska. It is better, & quite sufficient for the immediate purpose, to rely upon the classical tongue with its undoubted & well-ascertained meanings.

These are the lines upon which I have conducted my enquiry. The full proof of the results arrived at depends upon a larger labour of minute classification both of root families and word families in all the greater Aryan tongues,—a labour which is already in process, but not yet complete. What I have written in this book, will, I hope, be judged sufficient for a secure foundation. If it does no more, it may possibly lead to a deeper & freer approach to the problem of the origin of speech, which, once undertaken in the right spirit and with an eye for the more subtle clues, cannot fail to lead to a discovery of the first importance to human thought & knowledge.

Page 526

Chapter II

In an ordinary language which has not preserved the evidence of its origins, we are compelled to start with the full-formed word as our first & earliest document. We then find words existing in very small, unconnected families, little individualistic groups which seem to have started life on their own account without any observable growth from a common stock with other words that have, physically, a race-resemblance to them.We can all see that doceo, doctrina, doctor, docilis, documen, doctus, docte are one family. They acknowledge their kinship openly. From this acknowledged kinship we can draw certain important conclusions; especially the law of development from a common root & certain fixed forms by the accretion of which to the root this development was effected. It is a beginning, but it does not carry us beyond the surface-strata of our subject of enquiry.

For when we look farther, we are confronted with a serious difficulty. We find a certain number of words which, in their formation, would seem to be connected like the family we have glanced at above though not so strictly connected:—doleo, I grieve, dolenter, painfully, dolor, grief are obviously so interconnected; dolus, fraud, dolosus, deceitful, dolose, deceitfully, are so interconnected; dolo, I hew, cut or break, dolabra, an axe, are so interconnected. But between these three families we have in Latin itself no proof of any connection. In languages which have so far worn away their original phonetic moulds that entirely unconnected words wear the same or a similar dress, this want of connection would not lead to any farther conclusion beyond our mere inability to establish a connection. But Latin is a language which has preserved its phonetic moulds to a considerable extent. If then these three little families are entirely unconnected, then any hope of establishing an effective Science of Speech-Origins or even a Science of Aryan Speech- Origins must be abandoned. For by the supposition to which we would then be inevitably led, there must have been three original roots, dol, to grieve, dol, to cut or split, and dol, to deceive, unconnected with each other in origin. How then did

Page 527

their significances come to be attached to them? By chance? by caprice? by arbitrary choice? by some obscure psychological law we cannot trace? We can no longer hope to decide.

The hypothesis I shall start from,—and every attempt to connect the superficially unconnected without which there can be no science, must start with hypotheses,—is that there can be no such want of connection, that dol, to grieve, dol, to split and dol, to deceive must have been & are one root and not three and the three different significances now attached to them, have been developed not by caprice, chance or arbitrary selection, but have a natural connection and were developed in intellect by an intelligible psychological movement behind intellect from an original common meaning or mind-impression created in the Aryan mind by the sound dol. For I hold it to be obvious that speech must have started from what we in India would call the guna of sound, some natural property of particular sounds to create under given conditions a particular kind of impression on the mind which, constantly associated with that sound, became the basis of a number of special intellectual significances, called by us the meaning of words, much more variable, much less fixed than the basic mind-significance. Afterwards the intellect playing consciously with the sound, by association, by analogy, by figure, by metaphor & simile, by transference, by a number of means, may carry the intellectual significance far outside the bounds of the original mental impression. Still, if we have some evidence, clues may be found & then the vagrant word may be traced back to its parent mind-impression. For this reason we have to catch a primitive language when it is young or else find one which even in its maturity is more faithful than others to its primitive mould and preserves on its face much of its ancient history. Such a language is Sanscrit; it is, in fact, almost the only language which at all answers to our need.

But a sound like dol is not & cannot be a pure, primary & isolated sound. It has congeners, at least in form, brothers, cousins, more distant relatives. Does this kinship in form involve an original kinship in mind-impression and therefore in history of significance? If the theory of guna is correct, there must be

Page 528

some such kinship. Turning from Latin to the more fruitful field, the more copious evidence of Sanscrit, we find this root dol in the form dal (a sounded like the English u in dull and represented both in Greek & Latin by either a, u or o) meaning also to split, burst, & then to bloom, open. We find dala, a fragment; a blade, petal or leaf; we find dalapa, a weapon, that which splits, just as we have dolabra, an axe, from dolo; dalmi, Indra’s thunderbolt, also the god Shiva; dalika, a piece of wood, that which is split. We find also dalbha, meaning fraud, dishonesty, sin, and we have this established that in Sanscrit also, the root dal meant to deceive as well as to split. We find also the reason why dal came to mean to deceive; for the word dala means not only the blade of a weapon, but the sheath of a weapon. In other words, dal must have borne the significance, to cover or to contain. We find from other Sanscrit instances that the idea of covering or hiding led naturally in the Aryan mind to the idea of fraud or deceit, as in chad, to cover, chadma, a disguise, pretext, fraud, dishonesty, trick. But how are the two significances, cover & split, connected? That they are connected, is established as a strong probability at least by the word, cha, cutting, dividing, a fragment or part, which in its feminine form chā means covering, concealing and the neuter cham, a house, that which covers. If they are connected, the idea of cutting must have led to that of cutting off, separating, screening and thence to the significance we find in chadman, covering, disguise, fraud. There is no distinct significance of pain attached to the root dal either in Sanscrit or Greek; but we do find that the word dalita in Sanscrit meant crushed, oppressed, trampled, and, more curiously & significantly, we find dālanam in the sense of toothache. It is easy to see how the idea of cutting, tearing, rending must have led easily to the sense first of a special kind of pain and then by detrition of force to that of pain generally. But we find more. We find not only dal, we find other roots kindred in sound, having something of the same history. For instance, dambh means to kill, destroy, strike down; but dambha, the noun proper to this verb, means deceit, fraud, trickery, sin, ostentation, pride (we see how starting from the idea of fraudulent

Page 529

intention or hypocrisy we come to the very different idea of ostentation without fraud or pride,—again by detrition of special force); we find dambha & dambholi meaning like dalmi Indra’s thunderbolt, and dambha means also, like dalmi, the god Shiva. It is impossible to avoid the conclusion that to the Aryan mind dambha & dalmi were words so closely akin that they carried easily the same impressions to the mind and the same significance to the intellect. But what is then common to these two roots? It is the sound da, which must, therefore, by my theory have had a guna or mind-impression which naturally adhered in common to the two roots dal and dambh.

It is the second step of my theory, therefore, that not only must the three dal families be one family, not only must one root-sound have had originally one root-meaning, but that all kindred root-sounds must also be of one family and have proceeded from the simple sound, consisting of consonant & vowel, which is common to all of them, and the guna or natural mind-impression belonging to that simple sound must have been the basis not only of the intellectual significances common to its progeny, but of those even which vary most from each other. Da is the simple root-sound,—the primary root; dal, dambh, dabh, daś, dah, dakṣ, daṁś, das, daṁs, dagh, daṅgh, daṇḍ, dad, dadh, dan, dam, day, roots which we find or can trace in Sanscrit, are its derivative, secondary or tertiary root-sounds. The simple sound contains in itself the seed significance which it imparts to its descendants, whether sons, grandsons or remote progeny.We have thus immensely widened our basis and approached much nearer to a scientific consideration of language.

Let us see whether the hypothesis finds any farther support in the facts of the Sanscrit language. We take the senses to split, burst open, cut, tear, crush, destroy, cheat, belonging to dal; we find the same senses or kindred senses such as hurting, as in dolor and doleo, not only in dambh, but in dabh, to injure, hurt, deceive, cheat and its Vedic derivatives dabdhi, hurt, & dabha, fraud,—if these be the right senses; & in dabhra, little, small, from the sense evidently of cutting, a slice, or small part.We find them in dam, to crush, afflict, & so to subdue, overpower, tame,

Page 530

conquer, restrain and its derivatives, dama, damaka, damathu, damana, damin, damya, dānta; the last containing a lengthening of the vowel, to which we shall have occasion to return.We also find in dama, damathu, damana, damya the kindred sense of punishment, & we find in the Vedic sense of dama, the significance house, as in Latin domus, Greek δóμoς,δῶμα (again we notice the lengthening of the vowel), from which at once we return to the idea of covering we had to infer in dal. All these are evidently kindred roots belonging to the labial variety of the da family, formed that is to say by accretion of the sounds p, ph, b, bh, m (labio-nasal) or any combination of which they are the base to the simple sound.

We turn to other subfamilies. We find in the guttural subfamily daks., to hurt or kill, dakṣāyya, a vulture (tearer of carrion); dagh, to kill, hurt; dāgha, burning; daṅgh, to abandon or leave, which I trace to the sense of cutting off, separating, casting away; an association of ideas we shall find again in Sanscrit.We find in the cerebral subfamily, daṇḍ, to punish, fine or chastise; daṇḍa, a cudgel, staff or sceptre,—afterwards any trunk, stalk or thing standing; fine, chastisement as in dama, damathu, damana; assault; subjection, control, restraint, as in dama; pride, as in dambha; a corner or angle, apparently from the sense of cutting off, separating & so containing, which mates it in its roots to dama, a house & dal, to feign or deceive. A number of derivatives from daṇḍ & daṇḍa repeat the same senses. We find also dāḍaka, a tooth or tusk and dāḍhā, a large tooth or tusk.We find in the dental subfamily danta, a tooth; also bower, arbour (to cover, screen, shelter); dān, to cut or divide, & its derivatives dānava, a Titan, dānu, a demon, also supposed to mean conquering or destroying, like damana; dadhi, a garment (to cover). We find in the liquid subfamily, along with dal, day, to hurt & dāya, loss, destruction, a part, share or gift. We find in the sibilant subfamily das, to destroy, bite, overpower (dam); to decay, waste, perish; to cast away (cf daṅgh, to abandon) & its derivatives, notably dasyu, an enemy; daśana, tooth & daṣṭa, bitten; daṁś, to bite, sting; daṁśa, bite, sting, cutting, tearing, tooth, pungency; a limb or joint; dandaśa, a tooth; daṁṣṭrā

Page 531

& other derivatives varying these senses; daśā, a division or period of time, afterwards a state or condition, age etc; but we find also daṁśana, daṁśa and daśana in the sense of armour; daṁśita, mailed or protected; dantura, covered, overspread,— which bring us back to the idea of covering. The idea of protection once ascertained & traced, we turn back and find it recur in dagh, daṅgh, in dānam, protection (as well as gift), day, to protect, have pity, with its derivatives. The sense of giving which we find in dāya & can trace to the idea of cutting up, distribution or casting away, abandoning, handing over—but it can be shown to result really from the former—we get in dā, to give, dānam, a gift & many other derivatives; dakṣiṇā, a gift; in dad & dadh, to give, dasma, a sacrificer and in day, to grant, divide, allot. We have, in addition, dah, to torment, grieve, burn & its derivatives in the same sense; dahara, small, fine, young (cf dabhra),—a mouse or rat (the gnawer); dahra, small, fine, thin, a fire. Lastly we have, proving the previous existence of an obsolete root da, the adjective da in the sense of giving, destroying, cutting off; the noun da, a gift or donation and the feminine dā in the sense of heat (dah) and of repentance (dolor). The evidence is almost of an oppressive conclusiveness. It is a family of words, which bear the same or kindred meanings and seem all to go back to the root meaning to divide, usually with some idea of completeness, force, or even violence.

There are, we must note, a certain number of significances which do not at once trace themselves to or apparently connect themselves with the original sense. Some of these can on a moment’s reflection be so traced & evidence of the particular association of meaning on which we have to proceed, can be found in other root families. Thus we have daśā in the sense of mind or understanding; dakṣ, to be competent or able; dakṣa & dakṣiṇa, expert, able, fit which connect themselves with the Latin doceo, to teach, the Greek δoχέω, I think, seem & δóξα, opinion, idea, judgment; but the sense here is evidently that of discrimination or analysis as in the Sanscrit roots ci & cit, cetas, mind, citta etc; and this brings us back to the idea of division and distribution. We may compare dambh, to arrange & so collect, and ci bearing

Page 532

the same senses as well as to heap up, increase which we find again in dakṣ., to grow, increase (cf dānam, addition). There are other common senses, mostly connected with the idea of moving or of shining, daṁś, to shine (also, to speak?); dakam, water (to flow); dakṣ., to go or move; dagh, to go, leap, flow, attain; dānu, a fluid or drop; dabh, to go, push, impel; dabhram, the ocean (to flow); das, to shine, dasma, beautiful (bright, shining). These dissociated meanings are very few in number & rare in occurrence. Such as they are, they occur in different parts of the family, guttural, labial, dental and sibilant, and their presence & distribution proves yet more powerfully the now apparent & established truth that all Sanscrit words having for their basis the sound da are of one family, go back to the simple sound da as their simple root of being and derive from it all their varying senses. We have to add this fact, important for the particular family & as we shall see for the whole theory but not affecting our general conclusion, that we must seek in the original mind impression of the sound da some force of guna which gives rise directly to the idea of dividing with force or completeness and also can enter into ideas of motion & shining.

But we have not yet finished with this sound da. For just as the derived sound dal had its congeners, sounds kindred to it in form, so has the simple sound da other simple sounds by its side which are kindred to it in form and ought therefore to be congeners. These sounds are dā, di, dī , du, dū, dṝ & dṝ. The vowel sounds e and o, ai and au are in Sanscrit merely modifications of i and u, so that these seven roots with the lost root da form the whole original family of simple sounds depending on and having for their common base & element, the consonant sound d. If these roots are found to be one original family, we have gained another step and come yet nearer to the foundations of speech. My third step in the hypothesis is to accept this supposition and to lay down the rule that all simple roots formed in sound by the accretion of a vowel to the consonant sound d are one family having the guna of that sound as their seed of meaning, just as they themselves are separately the seed of meaning to their own descendants. We get therefore

Page 533

a seed-sound in addition to the primitive root-sound and their descendants the secondary & tertiary root-sounds.

Let us see how this hypothesis fares when confronted with the facts of the Sanscrit language. We have seen in passing that da & dā are the same root, one the short form, the other the long form. Dā has the same sense as da, dāś, like daś, means to hurt, kill, it means also, to give. There are no senses of the dā roots which are not shared by or traceable to the da roots. We must therefore proceed to the other vowels as forming three & not six classes of roots; we may treat in the absence of any opposing facts dī as a lengthening of di, dū of du, dṝ of dṝ. From the da family I have omitted the words which have for their basis the syllables dar & dav; yet these words are of great interest. For we find dara, in the sense of tearing, rending and also, like dabhra and dahara, of little, small. This sense of tearing, breaking, hurting we get again in daraṇam, daraṇi, darita, dardara, darma, darva (injury, mischievous person, goblin, as in dānu), in dāra, a rent, hole, ploughed field extending itself to dārāḥ, a wife, dāraka (also a child, infant, young animal, sense belonging also to dahara), dāraṇam, dārikā, dāri, dārī, dārita, dārin, darbha, the sharp Kusha grass, dardura, a district, province, dāru (tearing, rending, also a piece of wood, wood or pine-tree), & dāruṇa, terrible, rough, cruel, frightful, sharp, severe, violent or agonising (of grief & pain), a word of great interest as it shows us how moral senses developed from the physical idea. We find too dara, a cave, daratha and darī (also a valley) in the same sense, from the idea of cleft or hole which we have already had in dāra & dārikā. Again we have the same word dara in the sense of a stream, daraṇi, an eddy, current or surf, daratha, fleeing, flight, scouring for forage, dārdura, water. Connected perhaps with the sense of flight but really expressing the oppressive troubling feeling of fear, we have dara, darad and darada, fear, daratha & darita, timid or frightened. We have daridrā, to be poor or needy, with its derivatives, connecting this family with the sense of suffering, oppression, distress, wretchedness, burning (cf dagdha, distressed, famished, dry, insipid, wretched, vile, accursed) we find in the da family.We have again dāru in the

Page 534

sense of liberal, a donor, kind (cf dakṣiṇa, also meaning kind). There are more curious identities. Darad means, among other senses, heart; now dahara and dahra also mean “the cavity of the heart or the heart itself”. Darad means also a mound, mountain or precipice; dardara, dardura, likewise means a mountain; but in the da family we have also da, a mountain, daśana, the peak of a mountain, danta, the peak, side or ridge of a mountain. The identification in sense of this dar basis in its stock with the da family is complete. Their only senses, not traceable to the common original meanings, which find no parallel in that family are those which spring from the idea of sound, dardura, dardarīka, a musical instrument; but we have in the Aryan vernaculars the word. damaru, a kind of drum, which may represent an original Aryan word not preserved in the literary language.

Now the question arises. Do all these words belong to the roots dṛ & dṝ or are they from an original root dar? There can be no doubt as to the answer. Nearly all, if not all, are avowedly children of the dṛ stock. It follows then that the roots of the dṛ family are one race with the roots of the da family, cousins perhaps, but members of a joint family who hold the same property in common & use it with a more than socialistic indiscriminateness. Dṛ. itself means to hurt or kill, dṝ means to tear, rend, split, separate, disperse, and to fear; dṛti, a skin, hide, or bag; dṛka, a hole or opening; dṛnphū , a snake, thunderbolt (dambha, dambholi, dalmi), wheel (dalbha also means a wheel), the shining or burning sun. Dṛp is to inflame, kindle or to pain, torture; dṛmp also means to torture, afflict, distress, dṛbh, to fear, & dṛbdham, fear, finally dṛś, to see with all its derivatives. That this sense of seeing which we find also in dṛp (darpaṇa, a mirror, darpaṇam, the eye) comes not from the idea of light in reflection but from the original physical idea of discerning, separating with the eye, is evident from the fact that das also means to see. There are two sets of associations in this word which are of considerable help to us in fixing the exact history of certain developments in this family. The word dṛp expresses any violent troubling emotion; it means to be greatly delighted, wild, extravagant, mad or foolish, proud or arrogant (without

Page 535

anything of that idea of ostentation attached to dambha); darpa means pride, insolence, rashness, heat, musk (from the strong, oppressive scent); dṛpta means proud. Dṛ* again means to care for, mind, desire, & so to worship or respect—its root sense is evidently care, anxiety or excitement of love or other favourable feelings. We see more clearly now why words of this root bear the sense of grief, fear, pain. The mind-impression of the seed-sound carries with it this possibility of expressing any emotion or sensation which is oppressive, troubling, disintegrating to the peace of the mind. To the pervasive root idea of strong division, we have to add the idea of oppression tending to division which is thus revealed to our observation.

But dṛpta also means strong and this sense is found again in dṛṁh, to strengthen, fortify, fasten, be firm, grow or increase; dṛḍha is firm, fixed, solid, dense, strong, hard; dṛḍham means iron, a fortress or abundance; dṛbh, to tie, fasten, arrange, string together; dṛṣad, a stone or rock; dṛh, to be fixed or fasten, to grow, increase or prosper. We have met some of these meanings in the da family. We have found words there which mean a mountain, and these may now be attributed to this root meaning of firmness, solidity, size and density. We may notice also a group of words which we have hitherto omitted; dāman, a string, thread or rope, a bandage, a girdle, which also means a line or streak (from the idea of cutting); dāmanī, a foot-rope, dāmā, a string or cord and dāminī, lightning, from the idea of shining. We may also note, as it now appears, that the kind of light indicated by this family, is only an oppressive or a sharp piercing light as in dāminī, dṛnphū, das, to shine, and the words which mean fire or to burn. I have to suggest that this idea of firmness, solidity, compactness comes similarly from a sense of close, heavy contact, pressing things together into firm cohesion.

What, then, is the result of this detailed examination of the dṛ family of roots? Always the same; first, that, whatever their varieties of meaning, there is no sense the words of this house-hold bear which cannot be paralleled from the roots of the da & dā household & does not either explain or get explained by them and, secondly, that these varieties resolve themselves to & derive

Page 536

from a common guna or mind impression variously applied.

Again, there are a certain number of compound roots with a base combined of d and r which it would be as well to examine here as possibly kindred to the dṛ roots. We find drakaṭa, a kettle drum; drāṅkṣ, to croak; druta, a scorpion (to sting); dru, wood, tree or branch (dāru); druma, a tree; druha, a deep lake; drāgh, to vex, torment, exert oneself, be weary, stretch, also to be able (cf dakṣ); drākh, to be able, to become dry (dagdha; dal, to wither), to adorn, grace (to shine); drād. , to split, divide or be pulled to pieces; drāpa, mud, mire (which recalls a meaning of dama, mud, mire), a small shell (dara also means a conch-shell); drāva, heat; druḍ, to sink or perish; druṇ, to hurt, injure, twist, bend; druṇa, a scorpion or rogue; druṇam, a sword or bow; druh, to hurt, bear malice; dru, to hurt or injure; drū, gold, from the idea of brilliance; drek, to sound (originally, a discordant sound as in drāṅkṣ), to grow or increase, to be exhilarated (dṛp); droṇa, a scorpion, a tree, a bucket. We have the idea of desire, wish or longing in draviṇam, wish, desire. We have the idea of solidity or density in dravya, substance, material, wealth, strength, draḍhiman, tightness, firmness, heaviness, & in drākh, to obstruct. All these form a goodly array of evidences, showing the family identity of these roots with the da and dr. groups.

There are a few isolated meanings whose connection is not so immediately clear, such as drā or drai, to sleep, drāṇam, sleep (cf nidrā); but this is probably connected in sense with drāgh, to be weary or heavy from exertion, & will then contain the common idea of heaviness or oppression; drāh, to wake; drāpa, heaven, either from shining or from the idea of covering; and one or two others of the kind. But these may all be traced with a little difficulty to the common significations and are extraordinarily few in number. One would expect in so ancient & long-lived a tongue as Sanscrit a far greater number of meanings which have wandered too far outside or too near to the farthest permissible verges of the country occupied by their race to be easily identifiable or exactly paralleled among their kindred.

Then we have a number of significations resulting from the root sense of motion which are of some importance to us. They

Page 537

start mainly from the two ideas of running and flowing. Dru, the most characteristic, means to run, flow, rush, attack, melt, ooze or simply by detrition of special force, to go or move. This root also means to hurt and to repent. We have also druṇ, to go, move; drū in the same sense; dram, to go or run about (Gr.δρóμoς); drapsa, a drop; drava, speed, etc, the noun proper to dru, but meaning also play, amusement (cf div, later); drāva, liquefaction, melting, running, flowing, flight, speed; drāvaṇam, distilling; dravantī, a river; drā, to run, make haste, fly (the same word which means to sleep); drāk, quickly, instantly etc; drāgh, to wander about. We shall find that the idea of motion is common to all Sanscrit root families but that in each case there are certain special significances kept in the words, where their special force has not suffered detrition, which tend to show that they originally indicated a particular kind of motion. It is possible & probable that swift, overcoming, forceful motion, “darting, dashing”, kindred to the idea of pressure & division, is the proper sense of motion in the roots of this family. It is even possible that the words drāva & drāvaṇam from dru—distilling, liquefaction by heat, etc—dakṣ, to do, go or act quickly, keep the original force, & that the other shades of sense under this head show the gradual force of the influence of detrition, a phenomenon whose study is of as great an importance in the history of language as the study of detritions of sound rightly so much insisted on in Comparative Philology.

After such consistent & conclusive results a very cursory examination of the di & du families might be held sufficient. Nevertheless, in order that the full force of the evidence may be appreciated, I shall devote an equal care to these two households, fortunately not very numerous in their population, as well as to the compound bases, dy & dv & the modified forms de (dai) and do (dau). We start as in the dra roots with diṇḍi, a kind of musical instrument, & then come to dita, cut, torn, divided; diti, cutting, dividing, liberality; ditya, a demon (also daitya, cf dānu, dānava); dinv, to gladden, please (dṛp); dimp, dimbh, to accumulate (dambh), also to order, direct; div, to shine, play, sport (cf drava); squander (from the sense of waste, scatter);

Page 538

to throw, cast; be glad; be sleepy (drā, drai); be mad or drunk (dṛp); to wish; to vex, torment, lament, suffer pain; & two new meanings, to sell & to praise,—the one associated with the idea of giving, delivering, distributing; the other with the idea of love, respect, homage (dṛ). Proceeding we find div, diva & divan, heaven, sky (which helps, perhaps, to solve our former difficulty drāpa, though I believe that to be connected with Vedic drāpi, a cloth or robe), day (also dinam), light, brilliance (the original meaning); divya, divine etc; deva, divine, a god, quicksilver, a sense we have also in [    ], a lover; sport, play; dev, to sport, gamble, lament, shine, throw or cast; devanam in connected senses, but also meaning praise, motion, beauty, and an affair or business which connects it with dakṣ & perhaps with the Gr δράω , I do & δρμα; diṣṇu, a giver, donor; dih, to increase, augment, and to smear, from the idea of rubbing, pressing; de, to protect, cherish; deha, anointing; body (to contain); dehī, rampart, wall (to cover or to strengthen); dai, to protect, brighten, cleanse, purify; dī, to perish, waste; dī, decay, ruin; dīti, dīditi, splendour, lustre; dīna, poor (daridra), distressed, wretched, sad (dagdha), frightened, timid (dara, darita); dīp, to shine & its derivatives; dīrgha, long (cf drāgh), dīrghikā, a lake, big pond or well. Finally we have diś, to give, grant, pay, assign, allot, show, point out, teach, direct or order (cf dimp above); diṣṭa, deśa,diśā, a direction, quarter. The last root, identically with Gr δείχνυμɩ , at once throws a light on daśā, understanding, dakṣa, Grk δóξα,δoχέω Latin doceo, I teach. It is the same idea of discernment, discretion or separation, allotting things to their place, showing, teaching—the family of special significances which have since had so important & brilliant a history.

Once again what is the upshot of this substantially exhaustive statement of the significances in Sanscrit of the di family? Once again the result never varies. It is precisely the same. It is as if this particular family in Sanscrit, at any rate, were insistent on proving the theory with which we started, on declaring themselves all one family, with the same spirit, the same temperament, the same intellectual equipment, the same physical features. Absolutely, we have arrived at hardly a single new significance &

Page 539

none which can be isolated from the rest of the family.

We turn to the du roots. We start with du, to burn, torment, afflict, give pain, be pained; also to go or move! Then there is dūna , pained, burned, agitated (the essential idea in all emotional senses in this family, good or bad); duḥkh, to pain, afflict, distress & duḥkha, pain, grief, trouble, difficulty, unpleasant, difficult, uneasy, with its derivatives; duḍi, a small tortoise (duli also means tortoise); duṇḍuka, dishonest, fraudulent, bad-hearted; dundama (but probably from Rt ḍam, cf ḍamaru), dundubhi, dundu, dundubha, a drum; dundumā, the sound of a drum; dudh, to kill, hurt, injure, propel, with its derivatives; dur, a particle prefix with the sense of hard, bad; durv, to hurt or kill; duvas, active (cf dakṣ, devanam,δράω ); dul, to toss up, swing; duṣ, to corrupt, spoil, destroy; to censure, annul; to be bad, impure, sinful, & its derivatives (cf daṁśa, a fault or defect); dūṣikā, a paint brush (cf dih, to smear), rheum of the eyes; dūṣya meaning corruptible, pus, or poison, but also cotton, a garment, a tent,—the common root sense to cover suddenly turning up in this unexpected quarter as if to point out the entire identity of these families; duh, to milk or squeeze out (here we have the original sense of violent pressure), to yield or grant, to enjoy, to hurt, pain, distress, & its derivatives (cf also dogdhṛ & doṣaka, both meaning a calf); dū, to afflict, be in pain; dūrvā , kusha grass (cf darbha); dūra, far, distant (cut off, separated); dūśyam, a tent; and finally dūta, a messenger, which must derive from the sense of impelling, sending we have already found in this family. We have also do, to cut, divide, mow, reap; dora, a rope; doṣas, doṣā, night, darkness (to cover, hide); dos, doṣā, doṣan, the arm, forearm, the side of a figure (probably, to cover, contain, ewmbrace); doha.... Once more, we receive nothing but confirmation of our theory.

There are, finally, two connected families, connected, as we might say, by marriage with the seed sounds y and v. In the first we have dyu, to encounter, attack; dyu, day, sky, brightness, heaven, sharpness, fire, with a number of kindred words, dyauḥ , dyumat, dyumna etc; dyut, to shine, elucidate, express, with its derivatives; dyūta, gambling, play, battle; dyūna, sportive, sorry;

Page 540

dyai, to disfigure, despise (cf div, to make sport of, squander, make light of). In the second we have dvi, two (to separate) & its derivatives; dvandva, a couple, strife, duel, fortress (to strengthen), secret (to hide); dvār, dvāra, door, gate, aperture; dviṣ, to hate, dislike (cf druh etc) with its derivatives; dvīpa, shelter, protection, refuge, an island (to divide, cut off, separate), a division of the world, continent; dvīpin, a tiger, leopard (to tear, rend); dvṛ, to cover, hinder (obstruct), disregard, misappropriate (cf dasyu, a robber). Again, an absolute confirmation.

We have completed our survey of this great D clan of Aryan words, so far as the Sanscrit language holds them & introduces them to us in its classical form. No one, I think, can regard this evidence without being driven inevitably to the conclusion that here we have no chance aggregation of words, no language formed by chance or arbitrarily, but a physico-mental growth as organic, as clearly related in its members, species, families, subfamilies as any particular species of physical fauna & flora. The words claim each other for kinsmen at every step. Not a single family, not one small group fails to bring forwards its claim, its documents, its oral evidence. All stand together, shoulder to shoulder, as closely as any Highland clan or savage tribe. The most opposite meanings meet in a single word, but always there is the evidence borne by the rest of the family to their common origin not only in body but in spirit, not only in physical sound form, but in mental sense origin & development. The proof is complete.

We have then a single great family with a common store of sense-property which each uses according to his needs.We have a number of meanings all going back to a few radical significances. What are those significances? First, forceful, effective or violent division or separation; second, swift oppressive overbearing motion; third, heavy pressure or oppression; fourth, violent, oppressive, strongly agitated or simply emphatic emotion; fifth, strong, heavy sound; sixth, strong, overpowering scent; seventh, strong or swift action; eighth, strong, brilliant or oppressive heat or light; ninth, close, solid & heavy contact or cohesion. I have stated them at random, but I think a little reflection will

Page 541

show us that these nine fundamental ideas resolve themselves into the single idea of a heavy, decisive pressure, sometimes the idea of weight, sometimes the idea of decision predominating, applied to the fundamental experiences which would recommend themselves to the newly awakened and virgin observation of mankind; viz sound, contact or touch, (form), light, (taste), smell, motion & action, sensation objective & subjective. From the da family form & taste seem to be absent; either they have lost it or never applied themselves to these provinces of human observation. But we cannot yet say this precisely; for we have the word deha, body, the Greek δέμας, shape, body; δέμω, to build. It is obvious also that words expressive of taste must necessarily be fewer & more limited than the words expressing sound or touch. It is possible that words of form & taste were drawn by a figure from other primary senses & were not in themselves a primary application of the original mind-impression to the terms of intellectual appreciation.We shall have to examine languages more widely before this question can be decided. Another idea that hardly appears in this clan is that of human speech itself as distinguished from sound in general.

I, therefore, add an additional hypothesis to those I have already formulated, viz that the original guna or mind-impression created naturally & automatically by the seed sound (in this case the consonantal sound d) was applied primarily to the simple categories of sense observation, contact, sound, light, motion & action, including speech, sensation, and perhaps taste & form. It is no more than a hypothesis at present; for other sound families will have to be examined before this hypothesis can be either established or dismissed as untenable. I put it forward here for the sake of completeness.

It is not necessary to suppose that this perception of mental or sensational sound values, of the particular impression on mind-sensation of a particular inarticulate sound defining & separating itself on the human tongue or even its systematic application to the categories of sense observation was willed, conscious or intellectually reasoned out in the men who first framed their utterance into the vocables of Aryan speech.

Page 542

Nature, whatsoever Nature may be, guides the unconscious tree & flower, the unreasoning insect & animal to self-expression, to self-organisation, to self-evolution, & the result exceeds the best efforts of the deliberate human intellect. Why should she not have done the same for human speech? Instead of saying that men applied the guna of the particular sound to the sensations they wished to express, let us say that as in plant & tree & animal the sound itself, by the force of Nature, by the law of its own activity, svadhām anu,ṛtūn anu, “according to its own self-arrangement, in the straight line of the truth of things inherent in it,” and helped by the half conscious responsive awakening mind, applied itself in the service of mind to the various classes of sense observation which his awakening mentality demanded. I do not say this is the final & complete truth of the matter. But it is the only part or aspect of the truth of it at all consonant with our present way of approaching Nature as a blind force working in matter out of unconscious through half conscious into fully conscious action. It is the only theory which, provided it can establish itself, deserves, as it seems to me, to be called in the modern sense, rational & scientific; for it takes its stand on the two natural movements which constitute speech, the physical movements of articulation & the mental movement, partly sensational, partly discriminatory which attends the physical movements. And it seeks to establish itself by reducing the relation between these two motions to classified order & ascertained rule.

Chapter III

We have not, however, approached even yet the last step of our theory. For as there are families of words, families [of] root sounds, families of simple sounds, so also are there families of seed-sounds. These families are known to all grammarians & in Sanscrit they have been distinguished with a faithful & peculiar care & related to those parts of the organ of speech which play the decisive part in their articulation; but their relations to

Page 543

meaning seem never to have been studied. The seed sound D belongs to the group called dentals, which consists in Sanscrit of the hard t, the soft d, the aspirates th & dh & the dento-nasal n—for every consonant group except the liquids, the sibilants and the isolated aspirate h is composed of these five members, the hard consonant leading, the soft following, each attended by its corresponding aspirate and a nasal bringing up the rear.Sanscrit has three sibilants & even these it attributes to the complete groups; s to the dental, ś to the palatal and ṣ. to the cerebral family. The question then arises, are these groups only related in sound? or are they related also in guna and therefore in signification tendency? If there is any soundness in the theory I have been advancing, then as we have found the word-families united in a single root family with a single paternal root (as, dal, dah, dabh etc), these root families united through the paternal roots in a single primitive root family, phratria or brotherhood (as, the da family) with one paternal simple root, & these primitive brotherhoods united through their paternal root (da, di, du, dṛ) in a single clan with one paternal seed sound (d), so also we ought to find kindred clans united through their ruling sound into a single tribe based on the kinship of the paternal seed sounds. The sound d being closely related to the sound dh, must hold a similar guna and therefore carry with it similar intellectual significances, and, though they may coincide in a less degree, t, th and even perhaps n, though this is more doubtful, ought to be not far in guna & sense from the d & dh word clans. We must now proceed to examine the facts & perceive how the theory fares in this last & final test.

We will take first the aspirated soft dental dh. In the last chapter I have taken the reader very much at random through the word-jungle, pointing out as we went, how the different trees fell into groups, all belonging to families of one species. In this chapter, the theory having once been found, springing up of itself as we progressed, we can afford to proceed with more order & method; we can collect our specimens and present them ready assorted for examination & even speak with greater confidence about the precise nature of their connections.

Page 544

The commonest sense of the d roots, a significance which we found so pervasive that we were first inclined to take it as the root significance,—was the idea of violent dividing or rending pressure, especially in the senses “to hurt, kill, injure, destroy; to afflict, distress, give pain; to burst open, cleave, split; to deceive, cheat, etc.” Do we find the same senses or the same tendency in the dh family? We find dhrāḍ , to divide, split or pluck (flowers); we find dhvāṇkṣa, a carpenter; but we do not find any other words with the precise idea of splitting or breaking open or cutting—a deficiency of some importance for the proper appreciation of the guna of the seed sound dh. On the other hand the sense of hurting, injuring, killing, giving pain, is sufficiently common. We have dhakk, to destroy, annihilate; dhanus (dhanu), a bow or an archer (Nb dharma also occurs in this sense, but this does not prove that the idea of bow is “the thing held”, for dhṛ has other senses, “to drink, to flow”, & its secondary roots mean to hurt, kill, injure); dhṛs. & its derivatives, to hurt, injure, offend, outrage, attack, violate; dhāṭī, attacking, assaulting; dhū, to treat roughly, injure; dhur, distress or affliction; dhurv & dhūrv, to hurt, injure, kill; dhūr, to hurt, kill; dhūrta, dhattūra, dhustura, the white thorn-apple (with its intoxicating & stupefying drug); dhūlaka, poison; dhorita, injuring, hurting, striking; dhru & dhvṛ, to kill; dhvaṁs & its derivatives, to perish, fall, sink; also to scatter or sprinkle. We find dhiks., to be harassed, weary; dhyāma, soiled or unclean (spoiled, withered); dhrākh, to be dry or arid; dhūka & dhava, a rogue or cheat; dhipsu, deceptive; and dhī, dhīti, to disregard, disrespect. We find the simple sense of heavy or strong pressure in dhāv and its derivatives, to rub, brighten, polish; & in dhvṛ, to bend, which we have already had with the sense of killing. This harvest is not so plentiful, and that has its significance, but neither is it entirely scanty. We may notice also the sense of giving in dhartram, a sacrifice, dhāyu, liberal, and dhenu, a gift, present, but we must also notice that the impression here seems rather to be that of placing than of distributing.

But then we observe that the sense of pressure so scanty otherwise gives more liberal results in two special senses, two

Page 545

particular kinds of strong & insistent pressure,—to shake or agitate and to blow. We have dhū & dhu with many derivatives meaning to shake, agitate, shake off, blow away to kindle or excite &, directly from the sense of pushing, to resist or oppose; we have dhūnana, dhūka & dhavāṇaka in the sense of wind; dhūli, the driven dust or ground powder & derivative senses of smoke, fog, incense etc in dhūpa, dhūma & their derivatives,— dhūma meaning also eructation and dhūp, to obscure or eclipse; dhmā, to blow, with its derivates &, connecting these roots with the sense of hurting or giving pain, we have dhamana, cruel. A certain idea of action, labour or effort appears vaguely as in the d family, in dhmā, to manufacture & dhūma, a place prepared for building (cf Greek ίδρύω, I build), but as in the da family, the sense does not prominently emerge. We seem to have also the sense of covering, cutting off in dhārā, night (also meaning edge) & perhaps dhvāntam, darkness. But dhvāntam may also come from the idea of thickness, crassness more proper to this family. Here again we find a difference between dh & d.

Another class of meanings which we noticed in the da family were those which expressed some kind of motion & we perceived that a swift, overpowering pressure of motion was the original idea in that family. In the dha clan also there are a number of words conveying directly or indirectly the idea of motion. As we have dru, to run, there, so here we have dhāv, to run, glide, charge, to flow, to give milk, to wash. We have dhunayati, flows; dhuni & dhenā, meaning river; dhena and dhīra, the ocean; dhārā, meaning a stream, current, shower, the pace of a horse, a wheel (cf dalbha etc); dhūma, a meteor; dhor, to run or trot (of a horse) with its derivatives; dhoraṇi, series, tradition; dhārā, tradition, fame or rumour, line, series (but here the idea of continuing may be the source); dharuṇa, water; dhras, to toss up; dhūr, dhvaj, dhrā, dhṛj, all in the sense of going or moving; dhraj, to go or move; dhraji, a gliding, persistent motion; dhrāji, impulse, storm or wind. It is evident that here there is a great stress not on the force of the motion, though this sometimes emerges, but on its persistence.

Page 546