Why Sir William Jones got it all Wrong,

By Evelyn Boyd,2014-11-13 18:02
7 views 0
Why Sir William Jones got it all Wrong,

    Why Sir William Jones got it all Wrong,

    or Jones’ Role in how to Establish Language Families

    Lyle Campbell

    University of Canterbury, New Zealand

    1. Introduction. Sir William Jones (1746-1794) is often credited with establishing the Indo-European family of languages and founding comparative linguistics. Since many have argued that there are methodological lessons to be learned from Jones for establishing language relationships, this paper shows why this view of Jones‟ work is mistaken. Larry Trask has contributed much to exposing methodological and other errors in long-range comparisons intent on defending questionable proposals of distant genetic relationship (see for example, Trask 1995). This paper follows his lead in this arena; its goals are to set the record straight with respect to Sir William Jones‟ ideas, to examine the methods Jones used and to put them in proper perspective, and in so doing to contribute to understanding of appropriate methods for establishing genetic relationships among languages.

     Jones‟ (1786[1789]) famous „philologer passage‟ – that most momentous sound-bite of

    yore which declared a relationship between Sanskrit and several other Indo-European languages, is often cited as the beginning of Indo-European and of comparative-historical linguistics in general (cited below). Nearly all introductory textbooks on historical linguistics repeat this claim, as do many other linguistic publications. On all sides we see statements such as Bentson and Ruhlen‟s (1997:3) that Sir William Jones “discovered the method of comparative linguistics and with it the Indo-European family.” Murray (1998:3) reports as general

    knowledge that Jones “had founded comparative philology, or historical linguistics.” Cannon (1990:246) asserts that Jones‟ “was the first known printed statement of the fundamental

    postulate of Indo-European comparative grammar; more than that, of comparative linguistics as a whole.” Gray (1999:116-7) speaks of Jones‟ “earth-shattering discovering,” asserting that

    “Jones used careful etymological analysis to demonstrate that [here he repeats the philologer passage] ... The significance of this statement cannot be overestimated. It rested on an empirical demonstration ...”; “Jones stimulated a stampede of philologists.” Trautmann

    (1998:105), immediately after repeating the famous philologer passage, remarks that “the modernity of the formulation is remarkable … these are exactly the views historical linguists hold today.” (See also Emeneau 2000:545.) Even in his own day Jones‟ “reputation was such

    that intellectuals were literally expecting major discoveries in colonial India, since Jones‟ Persian

    grammar and his translations from Greek, Persian, and Arabic were well-known”(Cannon

    1991:23). By 1772, Jones had “established himself as the foremost exponent of Oriental studies

    in England and as a scholar and writer of rare attainments” (Arberry 1946:10). Nevertheless, with respect to the discovery of Indo-European as a language family and the founding of the comparative method, we will see in this paper that such views of Jones‟ philologer statement are erroneous.

     In this paper, the following well-known facts are brought into the picture, facts which gainsay the commonly repeated but mistaken view of Jones‟ role in the development of Indo-

    European and comparative linguistics:

    (1) Connections among Indo-European languages had been observed long before Jones (cf. Comenius 1657, Dante 1305, Gelenius 1537, Goropius 1569, Ihre Giraldus Cambrensis 1194,

    1769, Jäger 1686, J.J. Scaliger 1599[1610], Stiernhielm 1671, Lhuyd 1707, among others).


    (2) The relationship of Sanskrit with certain other Indo-European languages, especially with

    Greek and Latin, had also been recognized prior to Jones (for example, Thomas Stephens

    [1549-1619] 1583 (see Muller 1986:14-5), Filippo Sassetti [1540-1588] 1585, Jean François

    Pons 1743, Benjamin Schultze [1715-1790] 1760, Gaston Laurent Coeurdoux 1768,

    Nathaniel Halhed 1778, Lord Monboddo 1774-1809 (cf. Zeller 1967). Moreover, there

    were at least 47 published accounts of Sanskrit before Jones‟ statement on the matter (Muller

    1986:14). Jones was well aware of the views of some of these predecessors. (3) Finally, Jones‟ procedures bear little resemblance to the comparative method practiced by

    later linguists, and in any case they were not original to him.

    (4) Jones‟ interpretation of affinity among Asian peoples and their languages reflects not so

    much the linguistic facts as the biblical framework with Mosaic chronology in which Jones

    couched his thinking; this interpretation, based on the descendants of Noah, naturally

    involved a genealogical orientation, and this both reflected and imposed views of how

    languages could be related to one another this was Orientalism directed in defense of

    Christianity (Trautmann 1998:107, 109).

    (5) Jones‟ philologer passage is usually read out of context, with its interpretation based on too

    much of present-day understanding (Mukherjee 1968:95), with little real understanding of

    Jones‟ own intentions or of the intellectual environment in Jones‟ day.

     2. Jones’ plan. Jones‟ (1798:415) declared that his “design” (intent) was to prepare for the annual meetings of the “Asiatick Society of Bengal” (which he founded, later called the Royal Society of Bengal) “a series of short dissertations” (presidential addresses), the theme and purpose of which he specified as:

    The five principal nations who have in different ages divided among themselves, as a

    kind of inheritance, the vast continent of Asia, with the many islands depending on it, are

    the Indians, the Chinese, the Tartars, the Arabs, and the Persians: who they severally

    were, whence and when they came, where they now are settled, and what advantage a

    more perfect knowledge of them all may bring to our European world, will be shown, I

    trust, in five distinct essays; the last of which will demonstrate the connexion or diversity

    between them, and solve the great problem, whether they had any common origin, and

    whether that origin was the same which we generally ascribe to them. (Jones 1798:417-


    Jones saw his essays delivered before the Society as interconnected parts of a whole, and it is only in treating them as a whole that Jones‟ methods, claims, and conclusions can be fully apprehended. Jones‟ grand plan was to write a history of humankind in Asia. These eleven

    essays were published in Asiatick Researches, the journal of the Asiatic Society. Jones‟ direct

    interest was not in historical linguistic matters; language was but one sources of information relevant to his goals:

    We seem to possess only four general media of satisfying our curiosity

    concerning it [(pre-)history]; namely, first, their Languages and Letters; secondly,

    their Philosophy and Religion; thirdly, the actual remains of their old Sculpture

    and Architecture; and fourthly, the written memorials of their Sciences and Arts.

    (Jones 1798:421.)


    Jones maintained that languages were not worth study for their own sake, but only as means to a higher end: “I have ever considered languages as the mere instruments of real learning, and

    think them improperly confounded with learning itself” (Quoted by Godrey 1967:58; cf. also Rocher 1980:178). As Trautmann (1998:106) notes, “it is ironic that Jones is best remembered for his contributions to linguistics, given that he said more than once that he did not wish to be considered a mere linguist, and always regarded language as a means to other ends. Restored to its own context, the famous [philologer] passage on the Indo-European languages reveals its extra-linguistic ends very clearly.” “The Anniversary Discourses taken together form a wide-

    ranging essay on Asian ethnology” (Trautmann 1998:106). The Third through Seventh discourses follow a formula, first, in each discourse, one finds a description of the boundaries of the area in question, followed by sections dedicated to each of the “media,” language, religion, monuments, and arts and sciences. In the Third Discourse, the philologer quote comes in section I., the section on language and letters.

     Jones‟ more general interest in the history of the human races rather than in language per se was not unusual for 18th and 19th century linguistic scholars. It was shared by Leibniz, Hervás y Panduro, Monboddo, Vater, Schlegel, Grimm, Humboldt, and others. For all of these scholars, linguistic comparisons were seen as part of the means for getting at a broader history of the nations and races of the world ()see Campbell and Poser in preparation).

     3. The philologer passage. Jones pursued this grand plan in his anniversary discourses, delivered to the Royal Society of Bengal each February from 1784 to 1794 (see Jones 1798, 1799,1979a-f). The third to the ninth discourses were dedicated to solving the “common origin of the five principal Asiatic nations: India, Arabia, Tartary, Persia, and China,” with each

    nation allotted a distinct essay (Teignmouth 1805:387, Cannon 1952:44). The „philologer‟ citation is from the Third Discourse (On the Hindus), given in 1786 (Jones 1798). We cite it

    here in “restored to its own context” (Trautmann 1998:106), that is, in its fuller form including

    connected material immediately preceding it which is never quoted with the so-often repeated philologer passage itself:

    Five words in six, perhaps, of this language [Hindustani (= Hindi)] were derived from the

    Sanscrit ... but the basis of the Hindustánì, particularly the inflexions and the regimen of

    verbs, differed as widely from both these tongues [Sanskrit and Hindi-Urdu] as Arabick

    differs from Persian, or German from Greek. Now the general effect of conquest is to

    leave the current language of the conquered people unchanged, or very little altered, in its

    ground-work, but to blend with it a considerable number of exotick names both for things

    and for actions ... and this analogy might induce us to believe, that the pure Hindì,

    whether of Tartarian [Turkic and other central Asian peoples] or Chaldean [i.e. Semitic]

    origin, was primeval in Upper India, into which the Sanscrit was introduced into it by

    conquerors from other kingdoms in some very remote age; for we cannot doubt that the

    language of the Véda’s was used in the great extent of country which has before been

    delineated, as long as the religion of Brahmá has prevailed in it.

     [Here begins the “philologer” passage as normally cited] The Sanscrit language,

    whatever be its antiquity, is of a wonderful structure; more perfect than the Greek, more

    copious than the Latin, and more exquisitely refined than either; yet bearing to both of

    them a stronger affinity, both in the roots of verbs and in the forms of grammar, than

    could possibly have been produced by accident; so strong indeed, that no philologer

    could examine them all three without believing them to have sprung from some common


    source, which, perhaps, no longer exists. There is a similar reason, though not quite so

    forcible, for supposing that both the Gothic and Celtick, though blended with a very

    different idiom, had the same origin with the Sanscrit; and the old Persian might be

    added to the same family, if this were the place for discussing any question concerning

    the antiquities of Persia. (Jones 1798[1786]:422-3; cf. Teignmouth 1805:388, Pachori


     This passage is understood accurately only when seen in the context of Jones‟ overall 1thought and that of his times. It is to this I now turn.

     4. Jones’ mistakes. Jones erroneously grouped a number of languages and peoples (Egyptians, Chinese, Japanese, ancient Mexicans and Peruvians, see below) with Sanskrit and other Indo-European languages. Moreover, subsequent historians extol Jones for seeing Sanskrit‟s connection with other Indo-European languages but neglect his misinterpretation in

    the immediately preceding paragraph of the Sanskrit-Hindi relationship, which he saw as due to diffusion, to the introduction of Sanskrit into a pre-existing Hindi geographical context. Moreover, even within the famous paragraph, Jones‟ view of Gothic and Celtic as “blended with a very different idiom” (i.e. mixed with non-Indo-European languages) is passed over without

    comment, and his leaving open the possibility that the “common source” of these Indo-European

    languages may still survive is forgotten (cf. Cannon 1990:245). Also, Slavic was misassigned; Jones believed it belonged with non-Indo-European languages of Central Asia. Historians have also failed to given attention to the concluding paragraph of Jones‟ Third Discourse, where he

    presents his conclusions, which is much more revealing of his real thinking and of his errors

    than the isolated philologer passage, from which it differs significantly:

    Of these cursory observations on the Hindus ... this is the result; that they had an

    immemorial affinity with the old Persians, Ethiopians, and Egyptians; the Phenicians,

    Greeks, and Tuscans [Etruscans]; the Scythians or Goths, and Celts; the Chinese,

    Japanese, and Peruvians; whence, as no reason appears for believing that they were a

    colony from any one of those nations, or any of those nations from them, we may fairly

    conclude that they all proceeded from some central country, to investigate which will be

    the object of my future Discourses. (Jones 1798:431; cf. Pachori 1983:178.)

     As this shows, Jones incorrectly classified many languages, based on his four “media” or sources of evidence. These errors bear further attention, for they are methodologically instructive.

     In the famous Third Discourse, based on certain similarities between religions, Jones

    found that “it is very remarkable, that the Peruvians, whose Incas boasted of the same descent

    [i.e. the sun] ... whence we may suppose that South America was peopled by the same race

    [as the Hindus]” (Jones 1798:426). Jones repeated his assumed Peruvian connection and added Mexico to the picture in his Ninth and Tenth Discourses, for example:

    Nor is it unreasonable to believe, that some of them [from India] found their way from

    the eastern isles into Mexico and Peru, where traces were discovered of rude literature

    and mythology analogous to those of Egypt and India. (Jones 1979e[1792]:491,



    Also in the Third Discourse, based on the remains of architecture and sculpture, religion, and “letters on many of those monuments,” Jones concluded that “all these indubitable facts may induce no ill-grounded opinion, that Ethiopia and Hindustàn were peopled or colonized by the

    same extraordinary race” (Jones 1798:427). Ethiopian languages belong primarily to the Semitic and Cushitic language families, not to the same Indo-European family as those of “Hindustan”

    (in the Indian sub-continent) as those with which Jones dealt do.

     In the Fifth Discourse, on the Tartars, Jones (1979b[1788]:25) asserted an affinity between “Indian” (Indo-Aryan) and “Arabian” (Semitic) languages:

    I will not offend your ears by a dry catalogue of similar words in those different

    languages; but a careful investigation has convinced me, that, as the Indian and

    Arabian tongues are severally descended from a common parent, so those of

    Tartary [all the other languages of Central Asia] might be tracted to one ancient

    stem essentially differing from the two others.

     In Jones‟ (1979c[1789]) Sixth Discourse, on the Persians, we find that even Jones‟

    more linguistic methods led him astray in several cases, since he failed to distinguish loans, basing his conclusions on a fairly superficial comparison of the languages involved. For example, Jones misidentified Pahlavi, an Indo-European language of the Iranian branch, as Semitic:

    [I, Jones, and my friend Bahman, were] convinced after full consideration, that

    the Zend [Avestan] bore a strong resemblance to Sanscrit, and the Pahlavi to

    Arabic ... This examination gave me perfect conviction, that the Pahlavi was a

    dialect of the Chaldaic [Semitic family, especially Aramaic]; and of this curious

    fact I will exhibit a short proof. By the nature of the Chaldean tongue most

    words ended in the first long vowel, like shemià, heaven; and that very word,

    unaltered in a single letter, we find in the Pazend, together with lailia, night; meyd,

    water; nira, fire; matra, rain; and a multitude of others, all Arabic or Hebrew,

    with a Chaldean termination; so zamar, by a beautiful metaphor, from pruning

    trees, means in Hebrew to compose verses, and thence, by an easy transition, to

    sing them; and in Pahlavi we see the verb zamrúniten, to sing, with its forms

    zamrúnemi, I sing, and zamrúníd, he sang; the verbal terminations of the Persian

    being added to the Chaldaic root. Now all those words are integral parts of the

    language, not adventitious to it like the Arabic nouns and verbals engrafted on

    modern Persian; and this distinction convinces me, that the dialect of the Gabrs,

    which they pretend to be that of Zerátusht ... is a late invention of their priests,

    or subsequent at least to the Muselman invasion. (Jones 1979c[1789]:41-2.)

    Thus it has been proved by clear evidence and plain reasoning … that the

    language of the first Persian empire was the mother of the Sanscrit, and

    consequently of the Zend, and Parsi, as well as of Greek, Latin, and Gothic; that

    the language of the Assyrians was the parent of Chaldaic and Pahlavì. (Jones


    Of his Pahlavi no more need be said, than that it strongly confirms my opinion


    2 concerning the Chaldaic origin of that language. (1979c[1789]:43.)

     In the Seventh Anniversary Discourse, on the Chinese, Jones concluded:

    All the circumstances, which have been mentioned under the two heads of Literature and

    Religion, seem collectively to prove (as far as such questions admit of proof) that the

    Chinese and Hindus were originally the same people. (Jones 1799:378.)

    It is very true that the Chinese differ widely from the natives of Japan in their

    vernaculary dialects, in external matters, and perhaps in the strength of their mental

    facilities; but as wide a difference is observable among all the nations of the Gothic

    family; and we might account even for a greater dissimilarity, by considering the number

    of ages during which the several swarms have been separated from the great Indian hive,

    to which they primarily belonged. The modern Japanese gave Kaempfer the idea of

    polished Tartars; and it is reasonable to believe, that the people of Japan, who were

    originally Hindus of the martial class, and advanced farther eastward than the Chinas,

    have, like them, insensibly changed their features and characteristics by intermarriages

    with various Tartarian tribes, whom they found loosely scattered over their isles, or who

    afterwards fixed their abode in them. (Jones 1799:380-1.)

    Jones concludes this discourse with a statement of his belief that he has “now shown in five discourses, that the Arabs and Tartars were originally distinct races, while the Hindus, Chinese,

    and Japanese proceeded from another ancient stem” (Jones 1799:381; cf. also Mukherjee 1968:98; Teignmouth 1805:395).

     Jones‟ Eighth Discourse is on the “borderers, mountaineers, and islanders” of Asia. In it he arrived at several classifications known today to be erroneous. Repeating from his Sixth

    Discourse, Jones (1799c:5) “believe[d] on the whole, that the Ethiops of Meroë were the same

    people with the first Egyptians, and consequently, as it might easily be shown, with the original Hindus.” He also mistakenly classified other Iranian languages as Semitic, in addition to Pahlavi (seen above): “there is very solid ground for believing, that the Afghans[identified by

    Jones as Patans (Pashtos ?) and Balójas (Baluch), i.e. speakers of Iranian languages] descended from the Jews; ... principally, because their language is evidently a dialect of the scriptural Chaldaick” [Aramaic] (Jones 1979d[1791]:4, cf. also p.7). Jones (1979d[1791]:8) also

    mistook Malay as Semitic: “As to the Moplas, in the Western parts of the Indian empire, I

    have seen their books in Arabick, and am persuaded, that, like the people called Malays, they

    descended from Arabian traders and mariners after the age of Muhammed.” Jones also

    mistakenly regarded other Austronesian languages as connected with Sanskrit (Indo-European):

    If Mr. Marsden [see Chapter 6] has proved (as he firmly believes, and as we, from our

    knowledge of his accuracy, may fairly presume) that clear vestiges of one ancient

    language are discernible in all the insular dialects of the southern seas from Madagascar

    to the Phillipines, and even to the remotest islands, lately discovered, we may infer from

    the specimens in his account of Sumatra, that the parent of them all was no other than

    the Sanscrit. [Our emphasis, LC/WP.] (Jones 1979d[1791]:10.)

Jones also wrongly regarded Tibetan as Sanskrit (Indo-European):


    for, although it [Tibetan] was anciently Sanscrit, and polysyllabick, it seems at present,

    from the influence of Chinese manners, to consist of monosyllables, to form which, with

    some regard to grammatical derivation, it has become necessary to suppress in common

    discourse many letters, which we see in their books, and thus we are enabled to trace in

    their writing a number of Sanscrit words and phrases, which, in their spoken dialect are

    quite undistinguishable. (Jones 1979d[1791]:13.)

     Jones began his Ninth Discourse with a “short review of the propositions, to which we

    have gradually been led”:

    that the first race of Persians and Indians, to whom we may add the Romans and Greeks,

    the Goths, and the old Egyptians or Ethiops, originally spoke the same language and

    professed the same popular faith, is capable, in my humble opinion, of incontestible proof;

    that the Jews and Arabs, the Assyrians, or second Persian race, the people who spoke

    Syriack, and a numerous tribe of Abyssinians, used one primitive dialect, wholly distinct

    from the idiom just mentioned, is, I believe, undisputed, and, I am sure, indisputable; but

    that the settlers in China and Japan had a common origin with the Hindus, is no more

    than highly probable; and, that all the Tartars, as they are inaccurately called were

    primarily of a third separate branch, totally differing from the two others in language,

    manners, and features, may indeed be plausibly conjectured; but cannot from the reasons

    alledged in the former essay, be perspicuously shown, and for the present, therefore, must

    be merely assumed. (Jones 1979e[1792]:479-80.)

    This summary of Jones‟ conclusions from his interconnected discourses is strikingly different from the image usually derived from the philologer passage. We concur with Rocher (1980:179-80):

    Linguistic evidence led him [Jones] to postulate a common source for Sanskrit and other

    languages later known as Indo-European, but extra-linguistic arguments made him

    expand to non-Indo-European speaking peoples the list of nations with which the Hindus

    “had an immemorial affinity”. It was the same term “affinity” which he used to

    describe the linguistic kinship of Sanskrit, Latin, and Greek, and it is evident that he did

    not consider linguistic factors as qualitatively different from others he used. What Jones

    summed up as the “result” of the observations made in the discourse, was that the Hindus

    had a common origin with a vast array of nations, including the Egyptians, Phoenicians,

    Chinese, Japanese, even Peruvians ... This was a far cry from the method of

    linguistic reconstruction which Bopp and later scholars were to develop. [My

    emphasis, LC.]

    (See also Koerner‟s 1990:255.)

     His Ninth Discourse is quite different from the previous ones; it is almost wholly an attempt to accommodate Jones‟ conclusions about the nations of Asia within a biblical framework, one influenced heavily by the writings of his friend, Jacob Bryant (1774-1776), to whom Jones made occasional reference throughout his discourses and whom Jones praised highly in the beginning of the Third, where Jones began his treatises on the five principal Asian nations. Here Jones seems to abandon the better part of the comparative linguistics of the day


    in favor of a very long-standing biblical account, speculating about descent from the sons of

    Noah and about Mosaic chronology:

    Three sons of the just and virtuous man, whose lineage was preserved from the general inundation, travelled, we are told, as they began to multiply, in three large divisions

    variously subdivided: the children of Ya‟fet [Japhet] seem, from the traces of Sclavonian names, and the mention of their being enlarged, to have spread themselves far

    and wide, and to have produced the race, which, for want of a correct appellation, we call Tartarian: the colonies, formed by the sons of Ham and Shem, appear to have been nearly simultaneous; and, among those of the latter branch, we find so so [sic] many names incontestably preserved at this hour in Arabia, that we cannot hesitate in

    pronouncing them the same people whom hitherto we have denominated Arabs; while the

    former branch, the most powerful and adventurous of whom were the progeny of Cush, Misr, and Rama (names remaining unchanged in Sanscrit, and highly revered by the

    Hindus,) were, in all probability, the race which I call Indian. (Jones 1979e[1792]:485-


    From testimonies adduced in the six last annual discourses … it seems to follow that, the only human family after the flood established themselves in the northern parts of Iran;

    as they multiplied, they were divided into three distinct branches, each retaining little at first, and losing the whole by degrees, of their common primary language …; that the branch of YA‟FET was enlarged in many scattered shoots over the north of Europe and

    Asia … and had no use of letters, but formed a variety of dialects [languages], as their

    tribes were variously ramified; that, secondly, the children of HAM, Who founded in Iran

    itself the monarchy of the first Chaldeans, invented letters … they were dispersed at various intervals, and in various colonies, over land and ocean; that the tribes of MISR,

    CUSH, and RAMA, settled in Africk and India; while some of them, having inproved the art

    of sailing, passed from Egypt, Phenice, and Phrygia, into Italy and Greece … whilst a

    swarm from the same hive moved by a northerly course into Scandinavia, and another, by

    the head of Oxus, and through the passes … as far as the territories of Chin and

    Tancut … nor is it unreasonable to believe that some of them found their way from the eastern isles into Mexico and Peru, where traces were discovered of rude literature and

    mythology analogous to those of Egypt and India; that thirdly, the old Chaldean empire

    being overthrown by the Assyrians … other migrations took place … while the rest

    of Shem‟s progeny, some of whom before had settled on the Red Sea, peopled the whole

    Arabian peninsula. (Jones 1979e[1792]:490-1.)

    In the Tenth Discourse, this is reaffirmed:

    we cannot surely deem it an inconsiderable advantage that all our historical researches have confirmed the Mosaic accounts of the primitive world … Three families migrate

    in different courses from one region, and, in about four centuries, establish very distant governments and various modes of society: Egyptians, Indians, Goths, Phenicians, Celts, Greeks, Latians, Chinese, Peruvians, Mexicans, all sprung from the same immediate stem. (Jones 1979f[1793]:xv.)


     It must be asked how such a mistaken classification, forced to conform to preconceived biblical interpretations, could have been so misunderstood by generations of scholars as the foundation of comparative linguistics, how it could have had such a monumental impact in the linguistic literature?

     5. Jones’ methods. Let us look closer at Jones‟ linguistic methods, particularly with the claims in mind made by some scholars who attempt to justify their own methodological beliefs by calling upon Jones‟ authority.

     It may be helpful at the outset to recall Max Müller‟s (1861:162) assessment, that “it was impossible to look, even in the most cursory manner, at the declensions and conjugations, without being struck by the extraordinary similarity, or, in some cases, by the absolute identity, of the grammatical forms in Sanskrit, Greek, and Latin.” Because the relationship of Sanskrit to these other Indo-European languages was so unmistakable, it was obvious to Jones (and others who looked at it) without the application of any particularly sophisticated historical linguistic method. This being the case, Jones‟ methods (or lack of method) may, in fact, not be particularly instructive when it comes to looking at more challenging cases of potentially related languages where the relationship may not be so obvious.

     In assessing Jones‟ historical linguistic methods, it is important to bear in mind not only the cases in which he mistakenly grouped unrelated languages (mentioned above), but also the cases of related languages which his methods led him to dismiss (for example Hindi and Sanskrit, or Pahlavi and other Iranian languages), and even cases he correctly grouped together but for the wrong reasons.

     In this regard, it is instructive to contrast Jones‟ view of the role of grammar for showing language relationships with that of his contemporary Nathaniel Brassey Halhed (1778). Halhed had remarked in the preface of his Bengali grammar:

    I have been astonished to find this similitude of Sanskrit words with those of Persian and

    Arabic, and even of Latin and Greek; and these not in technical and metaphorical terms,

    which the mutation of refined arts and improved manners might have occasionally

    introduced; but in the main groundwork of language, in monosyllables, in the names of

    numbers, and the appellations of such things as could be first discriminated on the

    immediate dawn of civilization. (Cited in Müller 1861:162-3.)

    Both Halhed and Jones recognized “the same facts about Sanskrit and the modern languages of northern India” (Robins 1990:93), but drew diametrically opposite conclusions from them:

    Both saw the etymological links between Sanskrit and the modern vocabularies, and both

    noted the structural differences exhibited by the verbal inflexions and the verbal phrases.

    Halhed (1778:ix) declared that on the evidence of the etymologies the Hindustani

    language(s) were „indubitably derived from the Sanskrit‟ although „the inflexions by

    which the words are affected and the modes of grammatical regimen are widely different‟.

    But Jones [1798: 422], despite his admission that „five words in six‟ in Hindi are derived

    from Sanskrit, argued that the typological diversity of the languages in „the inflexions and

    regimen of verbs‟ precluded any relation of descent, asserting that Hindi was a surviving

    original language of India which had been heavily invaded by Sanskrit loanwords.

    (Robins 1990:92.)


    There is methodological irony in this: the same structural criteria which allowed Jones to group Sanskrit with the other Indo-European languages as well as with several unrelated languages

    prevented him from accepting the correct genetic relationship between Sanskrit and Hindi, i.e. the relationship between Sanskrit and the Indo-Aryan languages which Halhed had postulated and “conveyed to Europe” (Trautmann 1998:97-8).

     This should constitute sufficient warning to anyone who would praise the methods Jones used to propose linguistic families too enthusiastically. For example, Joseph Greenberg claims Sir William Jones as a historical precursor; he claims, in effect, that, Jones practiced „multilateral comparison‟, that is, Greenberg‟s own method:

    With Jones‟s background knowledge of Arabic and, no doubt, Hebrew, on the one hand,

    and Latin, Greek, Germanic, and so forth, on the other, the addition of Sanskrit to his

    repertoire enabled him to see a valid grouping based on differential resemblances. In

    later work he accurately outlined the Semitic and Finno-Ugric families ... In other

    words, even though he did not state it explicitly, he was in effect applying what I called

    earlier the method of mass comparison and more recently multilateral comparison.

    (Greenberg 1991:127.)

    (See also Greenberg 1949:79, repeated in Greenberg 1955:1.)

     Contrary to Greenberg‟s claim, Jones relied on notions about typology which led him to misinterpret the Hindi-Sanskrit relationship. Jones relied on grammatical evidence when speaking of the “affinity ... in the roots of verbs and in the forms of grammar.” However, the

    criterion of grammatical evidence for family relationships was already widely employed in historical linguistic studies before Jones and was accepted by almost all practitioners (see Campbell and Poser in preparation). In Jones‟ case, we need to recall the influences which oriented him to think as he did. As Cannon points out, Jones was influenced by William Robertson‟s typology with its view of social evolution from savagery to barbarism to (European ethnocentric) civilization. Such views of social evolution later strongly influenced views of language classification and genetic relationship (see Campbell and Poser in preparation). The following, then, are, I believe, some of the underlying reasons for why Jones spoke of “roots”

    and “forms of grammar”:

    (1) Jones studied Sanskrit with natives of India and they taught him using the Hindu grammatical

    tradition. This tradition held that the language was composed of amalgamations of lexical

    roots and derivations.

    (2) His British associates in India who had studied before him and had influenced his thinking

    had also obtained this orientation to the structure of Sanskrit from the pundits with whom

    they studied.

    (3) There was a well-established tradition in European linguistics (derived from Semitic

    grammars), with which Jones was familiar, which saw roots as older and basic. It is

    certainly not the case that Jones wrote about “roots” and “grammatical form” in this way

    because he had somehow discovered this aspect of comparative grammar on his own. (Cf.

    Hoenigswald 1985:65).

     Jones was already familiar with the claims made by his predecessors and contemporaries concerning historical linguistics in general. He corresponded with and associated with David Ruhnkenius, Everardus Scheidius, Hendrik Albert Schultens, Henry Thomas Colebrooke, Anne Robert Jacques Turgot, Sir Charles Wilkins, and others concerning such matters and about Sanskrit‟s relationship to various Indo-European languages, even before he left England for India


Report this document

For any questions or suggestions please email