THE BRAHMI FAMILY OF SCRIPTS AND HANGUL: Alphabets or Syllabaries?

A great deal of disagreement exists as to whether the writing systems of the Brahmi family ol7 scripts from Southeast Asia and the Indian subcontinent and the Hangul script ,..+f Korea should be classified as alphabets or syllabaries. Both have in common a mixture of syllabic and alphabetic characteristics that has spawned vigorous disagreement among scholars as to their classification. An analysis of the alphabetic and syllabic aspects of the two writing systems is presented and it is demonstrated that each exhibits significant characteristics of both types of writing systems. It is concluded that neither label entirely does either writing system justice.

Tn s. document has been reproducea as rece,ved from the person or organaatton Linguists studying the writing systems of the world have traditionally classified them according to three categories, those of logographic, syllabic, and alphabetic scripts.
The Brahmi writing systems found throughout the Indian subcontinent and Southeast Asia a. well as the Korean Hangul script, however, both defy classification.The two have in common a mixture of syllabic and alphabetic characteristics that has spawned vigorous disagreement among the scholars discussing them.
For example, Lambert (1953) refers to the Devanagari script used to write Sanskrit and its daughter languages as a syllabary, Shamasastry (1906) as an alphabet, Coulson (1976:3) :;:escribes it as 'halfway in character between an alphabet and a very regular syllabary,' while Cardona (1987) simply calls it a script and avoids the issue in his overview of Sanskrit.An examination of the various alphabetic and syllabic aspects of these writing systems is therefore in order, and indeed the results of such an investigation would seem to indicate that neither label fully does justice to them.
The Brahmi family of scripts, so named for their descent from the Brahmi script which is first attested in the third century B.C.,1 are distinctive in having in common, to a greater or lesser extent, a number of characteristics that begin to surface in their progenitor.Foremost of these is what Masica (1991:136)  hails as 'The great innovation of the Brahmi script, its indication of vowels other than A ([8)) by modifications added to the basic c.:.,risonant symbols.' The vowel corresponding to (a) itself is regarded as assumed or inherent to each consonant in its most basic form, and any vowel pronounced after the consonant is represented by a marker appended in some fashion to the consonantal symbol.Vowels also tend to have distinct allographs when they occur in an initial position.A consonant standing alone must be so indicated by a special diacritic, and consonants otherwise not followed by any vowel, as in consonant clusters, tend to appear in some altered or abbreviated form.
Descendants of the Brahmi script are most commonly associated with the Indic and the Dravidian languages of India.
They are also represented in the two primary members of the Tibeto-Burmese family, as well as in significant members of the Khmer and Kam-Tai families.Brahmi-derived scripts have also made their way to such scattered locales in time and place as Sumatra, the Philippines, and the extinct Tokharian language.2The most widely known member of this family of scripts, however, is the Devanagari script, most particularly as it is employed in writing Sanskrit.
It was also at the hands of the grammarians who adapted Devanagari to the writing of Sanskrit that the aforementioned qualities peculiar to Brahmi writing systems become perhaps most pronounced.While an analysis of Brahmi scripts should consider a representative sampling of them, Sanskrit Devanagari is generally taken as the most representative case, and is therefore the best point at which to begin.
The characters f the Devanagari script are elegant not only in appearance but also, in Sanskrit at least, in operation as well.As mentioned above, Devanagari consonantal characters are considered to include in their basic, 'unmarked' form the vowel [a], corresponding to [ ] in Sanskrit and most of its daughter languages, pronounced after the articulation of the consonant itself.
Thus, the characters for Sanskrit's voiceless unaspirated plosives,EB , and IT , stand for the syllables [ka], [ca], Eta], Eta], and [pa], respectively.When the consonant has no following sourd, as utterance-finally or in isolation, a diacritic known as a vir5ma is placed to the lower right of the character, so that -c.7. and -cT, indicate [c] and [t] alone.
The vowel [a] is overtly indicated only in an initial position, by the character JT .All other vowels and diphthongs have one allograph used initially, and another, smaller one when pronounced following a consonant.These latter allographs may be attached to the consonantal sign at almost any portion of it, such as to the right, as for TW [ta], -m- [ti]; belo , as inir [tu], [ta], [q]; above, as in tr [te], [tai]; above and to the right, as for rr [to], [tau]; and even to the left of the consonantal symbol, as in fh- [ti].
The signs for these vowels in an initial position, on the other hand, are 37r [a] [i],t Two diacritics frequently modify vowels.The anuamara ( ) indicates vowel nasalization and is customarily transcribed, e.g., -am.The visaraa (37: ; h) is an aspirated echo of the vowel it modifies (Coulson 9).
Although Devanagari does not readily lend itself to the representation of consonant clusters, such clusters are quite common in Sanskrit.These are represented by ligatures known as conjunct consonants, wherein two or more consonantal characters are modified to fit together in a larger conglomeration.The two most common means of effecting these combinations are horizontally, which generally involves deleting the vertical stroke where present for non-final members of the clusters, as in RT.
[sta], from "R is] and Tr [ta], or aZi [byal, from "Et [b] and Zr [ya]; and vertically, as in ---;;F [iga], from [g) and 71.[gal, or [dva) , from [d] and 1 [val.Some combinations may be made in either fashion, as inr-r4 or [cca], although the advent of printing has made the former method more desirable.These conjuncts can appear quite formidable and bewildering; Conjuncts involving the flap [r] are of particular interest.
[r] following a consonant is represented by a short diagonal mark to the lower left of the consonantal character, as in [kra].
However, when [r] precedes a consonant, it is indicated by a small hook above and as far to the right of the character as possible, as in Tf [rta].
In syllables involving the diacritic anusvara, p this hook appears even to the right of it, as in <4..rii'zir  [yajfiartham], 'for sacrificial purposes.'A question commonly invoked in determining::whether a script might be considered alphabetic or syl;lAbic is whether or not its most basic unit corresponds more or less with the phoneme; that is, whether it approaches an ideal principle of 'one sign per phoneme.' (Gaur 1985:119; see also Kim 1987:888-9).However,:this principle would seem to be for the most part irrelevant in Sanskrit Devanagari.Two points support this view.The first is what Masica (146) refers to as 'phonemic overkill' in the inventory of characters.He argues that the visaraa is in fact an allophone of /s/ t ), and argues that the velar and palatal nasals (;!,and,:q )  were 'largely predictable' in their distribution.It is true that they virtually never appear apart from a homorganic obstruent, and this would tend to indicate that they are less than full-fledged phonemes in Sanskrit and may have been included in the script to provide symmetry by nasals with the velar and palatal series of stops along with those of the retroflex, dental, and labial series ( , T( , R respectively).
The second of these points is embodied in the phenomenon of sandhi.Devanagari was adapted to Sanskrit with the goal of reproducing as faithfully as possible exact pronunciation (see , and the term sandhi, meaning 'juncture,' refers to all of the assimilation in voicing and place of articulation among consonants and the coalescence and glide formation among vowels at word boundaries and between lexical stems in compounding.A word-final segment analyzable phonemically as /t/ may be written, with pronunciation in mind, as 7r , , , , or [1], depending on the initial sound of the following word.Words are not separated from one another within clauses in written Sanskrit unless the first word ends in a vowel and the second begins with a consonant, or the first word ends with a visaroa and the second begins with a voiceless consonant, or unless the regular and predictable sandhi rules result in hiatus between two vowels. In attempting to separate strings of words into their component members, students of Sanskrit must work their way backward through these sandhi rules.The rules for sandhi given their predictability and their application across word boundaries, bear a striking resemblance to the post-lexical rules of the theory of lexical phonology (see J. T. Jensen 1990:84-7, 174-6).It must be concluded from the practices of regular Sanskrit orthography that the script was not adapted to the language with units corresponding to phonemes in mind.The implications of this fact would seem to that, while the orthography was organized to capture each sound as it passes from the lips of the speaker, these individual sounds were not considered meaningful in and of themselves.One is therefore left with no unit of analysis between the phonetic segment and the syllable conceived as a vowel preceded by any number of consonants (see Coulmas 1989:41-2).
A sample of written Sanskrit, accompanied by a transcription and translation, follows (adapted from Katzner 174):
In Hastinapura there was a washerman named Vilasa.His donkey was near death, having become weak from carrying excessive burdens.So the washerman covered him with a tiger-skin and turned him loose in a cornfield near a forest.The owners of the field, seeing him from a distance, fled away in haste, under the notion that he was a tiger.
Most of the modern Indic languages employ Brahmi scripts, and indeed most of these scripts are fairly closely related to Devanagari.Aside from some relatively minor languages, however, only Hindi, Marathi, and Nepali are generally written in the Devanagari script.Masica explains this great number of different scripts by noting that there was no unifying political or religious force, such as the Roman Empire and Catholic Church in western Europe or the Koran in the Islamic world, over most of Indian history (137), so that the sundry language communities tended to develop their own scripts.Then, 'What may have been the high water mark of script differentiation unfortunately coincided with the introduction of printing, which had a tendency to freeze and accentuate many minor differences (144).' He also observes that in the linguistic hodgepodge that is India, languages are under tremendous pressure to maintain a distinct identity, so that 'there is a widespread feeling that a self-respecting language should have its own script. (27).'  Even Hindi and Nepali have some divergent orthographic customs for the script they share (145).For the purposes of this discussion, the Devanagari of Hindi and Marathi will be considered, along with the closely related but visually more distinct GujaraLi script and the somewhat less closely related Bengali script.
The Devanagari characters as used for Hindi and Marathi are essentially identical to those of Sanskrit.4 The most significant innovation in shape involves the importation of non-Indic segments such as the Arabic [q] and [f] from Arabic as well as Persian and English.
In these cases a subscript dot is added to the characters phonetically closest to the new sounds.

Thus E [ka] becomes
[qa] and There are, however, two more fundamental changes in the script, pertaining to the manner in which it is mapped onto the spoken language.The first of these renders the script less imposing in appearance.Sandhi rules are no longer taken into consideration, so that separation between words is always maintained.Such rules are not effective within words, either; the modern languages under discussion allow two consecutive vocalic syllable nuclei within a word, with the second represented by the initial allograph, as in c4 [kat] 'several,' or Epqr[bu'A], 'paternal aunt.' In Sanskrit, any such sequence would.havebeen coalesced together, or reduced to a glide-vowel sequence.While individual words in Hindi and Marathi are easily distinguishable, the pronunciation of these words is rather less accessible to the non-native reader than in Sanskrit.
In some, but not all environments, the inherent vowel (a) is deleted.In these instances, a consonantal character stands for its corresponding segment alone, and no additional diacritic is necessary.
The most easily predictable environment is word-finally, as inlql [Par] 'but,' or mo [ksan] 'moment.' Word-medial environments are less obvious.The best discussion of this phenomenon is in Ohala (1983).She argues that he most basic environment for deletion of the inherent vowel is VC_CV (121).This is fairly readily apparent where the two vowels are overtly marked, as in 79 [ kohni] 'elbow,' or -M:14-'1'41 (cunna] 'to choose.'More troublesome are cases where one or both of the the vowels are also the inherent vowel.Ohala argues that the deletion rule then applies right to left from a morpheme boundary.She bases her conclusion on such data as the following: The rare word pronounced [godnagin] 'adopted' is derived from /god+na6in/ 'lap+sitter' but is written in Devanagari as 7.1-/-----RZT (godanain).
If a speaker knows the word is /god+npsin/ he will not pronounce the Z; (d)  of /god/ as a CV syllable (i.e., [dal), but will correctly render it as simply the consonant [d]; he will also retain the /a/ in /nasin/.
However, if he doesn't know the true morpheme boundary then he applies his adeletion rule from right to left and pronounces it as [godansin] (124).
Conjunct consonants do occur in Hindi, but they are rare relative to Sanskrit.
Lambert indicates that they do not occur across morpheme boundaries (77); when they do appear, they are often in environments where a- deletion cannot be predicted by Ohala's rule, such as word initially: /sneh/ 'love.'However, they also quite frequently occur where a-deletion is predictable, as in c,,h1 /kacca/ 'raw, uncooked, triiir/taiksi/ 'taxi,' or 54WITQFT/janmadin/ 'birthday.'Many of these are geminates, and Lambert takes pains to make clear that a-deletion cannot occur in loanwords from other languages, particularly Sanskrit (78-83). Nevertheless, while anyone who has internalized Ohala's rule should be  Horl use jate dekhta hua apna kalefd thandha karta ra'.Ab larke ki sagai ne der na karni cahie.
Gobar said nothing more.He put his staff on his shoulder and walked away.
Hori looked with pride at the receding figure of his son.He was growing into a fine young man.
The Gujarati script is fairly close in appearance to the Devanagari.
It differs chiefly in the absence of the distinctive headstroke.The phonotactics of Gujarati are quite similar to those of Hindi, except that consecutive vowels are not allowable within a word.A sample of written Gujarati follows, accompanied by a translation and transcription, adapted from Katzner The appearance of the the Bengali script is quite different from that of the Devanagari; broadly speaking, its characters can be described as tending toward a rather triangular shape.The Bengali language itself differs from the majority of Indic languages in that its vowel corresponding to a has drifted in articulation to This then is its inherent vowel, and so the consonantal character Z is taken to stand for [to).The vowels :DI and J , corresponding to Devanagari tr and 31 Hai] and [au)), are pronounced Eoi) and [ou).
One noteworthy feature of the Bengali script is that, in addition to T7) (tii, other non-initial vowels are written br7Dre the consonantal character: C6 [tel, et   [toil.Tu..) others are written to either side of it: C5 Signs for other non-intial vowels are not greatly different from their Devanagari counterparts Unlike Hindi words, whose pronunciations are predictable from their written form but not the reverse, in Bengali neither is fully predictable, since inherent vowel-deletion is not regular.Thus, the written form 1U , orthographically [motD], may denote either /Rut/ 'idea, opinion,' or /moto/ 'similar, like' (Lambert   185).
Further compounding difficulties, as is apparent from the latter example, the inherent vowel may also be pronounced (o), so that it overlaps with 3 [o].Ray et  al. (1966:15) states that there 'are no simple rules' for this alternation of D/o/O, and Lambert (185) asserts that the proper realization can be understood 'only by a knowledge of spoken Bengali.'A sample of written Bengali, with a transcription (albeit without taking into account the shift in pronunciation of the inherent vowel) and translation follows (from H. Jensen 379-80): ("1 .1 kr-q?rtc-k-t ice.IT-C9T qiC31-47q-5 "1-51.
Among the rich in the old days was a man called Amad Sultan.He possessed great wealth and also a numerous army.
Certain features of some non-Indic Brahmi scripts are worth noting, at least in passing.
Of note in the Tamil script is the pu,li.This is a raised dot corresponding in function to the Devanagari vir-ama, but, unlike its counterpart, as Stevens (1987:734) observes, 'The use of the milli is instrumental in the correct representation of Consonant clusters: g4u/.represents ilma 'now,' not l'ipapa.'Thus, in Tamil conjunct consonants are unnecessary.
The Thai script offers an example of diacritics used to indicate a fairly complex tone system.A consonant sign falls into one of three classes, and this class in conjunction with any of four diacritics or the absence of one determines the tone for thai: consonant's syllable (Hudak 1987:766).
Thai also appears to be unusual among Brahmi scripts in that consonantal characters have no inherent vowel; Yd stands simply for /n/.
In any comparison between the Brahmi family of scripts and the Korean Hangul script, that of the Tibetan language is particularly worthy of note as it is often mentioned as possibly having had some influence on the shaping of Hangul (Gaur 85, Diringer 1968:354, Lee 1983:7).
In this connection perhaps its most significant feature is the tsheg, a syllable-ending point.Otherwise, a narrow space separates each consonant character.Beyond this, it is fairly similar to the Devanagari script in appearance.In contrast with Devanagari, however, Tibetan syllables contain a staggering number of apparently superfluous consonantal signs called pre-, super-, sub-and postscripts, relics of the changes in spoken Tibetan since the script was invented, 'with auxiliary significance or none (Miller 1956:6),' which 'allow for variety in the writing of one and the same phonetic shape;' these 'just have to be memorized word by word: there is no rule to guide in their usage (8).'The Tibetan script does have largely the same system of vowel indication as the Devanagari.
Even if you d n't understand your neighbor, make allowances for him and his peculiarity.
There is no lack of scholarly opinion concerning the question of whether members of the Brahmi family of scripts should be considered alphabetic or syllabic.;Ireement alone is lacking on this topic.Masica refers to the scripts used for modern Indic languages as alphabets (145), while Snell & Weightman (1989:5) introduce Hindi Devanagari as a syllabary.Kachru (1987:474) also writing on Hindi, states that the script is 'syllabic in that every consonant symbol represents the consonant plus the inherent vowel /;/,' but then on the next page the characters of the script are listed under the heading of an alphabet.Klaiman (1987:493), writing on Bengali, describes its script as 'organised according to syllabic rather than segmental units,' and Ray et al. declare that 'It is a syllabary, modified somewhat towards becoming an alphabet' (12).Lambert maintains that all of the Indic scripts set forth in her work are syllabaries.Hudak (764) refers to the Thai script as an alphabet, and Miller (1) calls the Tibetan system of writing 'an alphabetic script on syllabic principles.' Wheatley, writing on the Burmese Brahmi script, declares that the inherent vowel 'sometimes leads to Indic writing systems being incorrectly labeled "syllabic "' (1987:844), but Steever, discussing Tamil's Indic script in the same volume refers to it as a syllabary (1987:734).
Disagreement among scholars of writing in general on the typological classification of Brahmi scripts arises in large measure from their differing definitions of alphabetic and syllabic systems.Gaur stresses that 'in alphabetic scripts... vowels and consonants have equal status' (119) and, since this is clearly not the case for Brahmi scripts, they are classified as syllabic.
Gelb (1965) -s on the whole unwilling to commit himself.He declares, 'The main characteristic of the alphabet is the existence of special signs for both consonants and vowels' ( 184), but then observes that in Indic writing systems the vowel indicators are 'attached to the respeccive syllabic signs' (187).He describes the inherent vowel as an 'abnormal development' (239) and relinquishes the question by calling for 'sharper typological definitions' for future discussions (188).DeFrancis (1989) draws a sharp distinction between syllabic scripts such as that of Japanese which represent syllables by means of Unitary syllabic signs, and Indic scripts which are 'syllabic' only in the quite different sense that they represent phonemes by means of nonunitary signs graphemes representing phonemes which are grouped together to form a syllabic bundle.Such scripts must still be classified as basically phonemic systems.He goes on to observe, 'The unit of writing, the syllable, is not the same as the unit of underlying analysis, the phoneme.' For both Coulmas and DeFrancis, then, it is this unit on analysis that establishes a script's typological status.
H. Jensen and Diringer both gravitate toward the alphabetic viewpoint.
Jensen writes regarding the classification of Brahmi scripts as syllabic: There is some justice in this point of view; on the other hand, however, two things must be emphasized, first that there are no syllablesigns for [e.g.] ki, ku, kg, kg, etc., on the contrary, in these cases a vowel sign is added, and the sign concerned thus has to lose its A and and become a pure consonant-sign; and secondly that when several consonants come together... the many ligatures themselves... show that the signs are first and foremost pure consonant-signs and that the inherence of an represents, not something essential, but a peculiarity. (362-3)  Diringer, too, argues the individual representation of sounds in the absence of an inherent vowel gives the Brahmi scripts an alphabetic classification: 'Syllabic forms of writing... are ultimately based on the fact that the smallest unit into which any spoken word or series of sounds can be subdivided is the syllable' (1962:23).
Later, however, he comes to view the inherent vowel as a flaw in the writing system and therefore calls the Devanagari script a 'semi- syllabary.' (1968:283) Both alphabetic and syllabic arguments regarding the typological classification of Brahmi scripts unquestionably have merit.With the exception of postconsonantal /a/, every phoneme receives an explicit segmental representation and, as the scripts were originally conceived at least, // could invariably be considered as present in the absence of any other mark.Still, it should be borne in mind that the existence of this inherent vowel is not some sort of aberration, but has been a part of these scripts from their origin.down into the syllable-based units of which they are composed.
In contrast, in an unambiguously alphabetic script words are constructed directly from their member segments, and these segments always appear in the same linear order relative to pronunciation.
In Brahmi scripts, within syllabic units, although every individual segment may be in evidence, the reader must have at least some ability to arrange these items into the proper order of pronunciation, as the signs 3Tr themselves may appear in virtually any order within their syllabic bundles.
In the Sanskrit word [arthin] 'wanting, petitioning,' the sequence r-th-i appears in reverse order relative to the left-to-right direction of the script.The assessments of Coulson and Diringer that the Devanagari script is neither wholly alphabetic nor wholly syllabic may therefore be said to possess considerable insight, for neither classification does the writing script complete justice.
The Korean Hangul writing system has been widely praised for the logic and straightforwardness with which it was devised.Gale (1912:14), for example, writes, 'In simplicity, the Korean [script] has perharz no equal, easy to learn and comprehensive in its power of expression.'Although it has forty signs corresponding to individual sounds, many of these are formed by regular principles from the more basic signs.The basic consonantal signs are: /k/, L /n/, t. /t/, /1/ ([r] initially),0 /m/, 8 /p/, -4 /s/, 6 (0 initially), 'X ./c/, t /h/.Aspirated plosives are indicated by adding a stroke to the symbols for the unaspirated ones: /kh/, /ph/, Z /ch/.
In like manner, there are eight basic vowel signs: Symbols for two other 'pure vowels' (N.K. Kim 889), /11/ and /0/, are formed by adding ) to the signs for their back counterparts and are alternately analyzed as /we/ and /wi/ (Lukoff 1982:xvi).
One other diphthong combines /i/ and /t1/: BEST COPY AURAE These individual signs are grouped together to form syllable-based blocks, again according to regular principles.The vowel-sign always occupies the central position, thus becoming the 'nucleus' for the syllabic group.Then, depending on wilether the vowel-sign is vertical or horizontal, the syllable-initial consonant is indicated either above or to the left of it: IF- /ca/.This initial position is never left empty; if there is no syllabic onset, a silent 6 appears in the initial position: 6) /i/, /yo/.
The final position may be left empty; when it is filled, it always appears at the bottom of the block, beneath the other two signs:4 /can, /, /to1/, .b/wan/.
These syllabic blocks have customarily been written vertically, although they sometimes are arranged horizontally to accommodate printing.
A sample of written Korean follows, accompanied by a transcription and translation (adapted from Katzner na po-ki-ka yak-kya-ik' ka-sil e-run mal-aps-i ko-hi po-ri u-ri-ta.
From Mount Yag of Yongbyon An armful of azaleas I shall pick, And strew them in your path.
Go now, I pray, with short steps!Let each footstep gently tread The flowers which I have strewn for you.
When you take your leave, Tired of seeing me, Though I should die, I shall not weep.
The pronunciation of the individual signs is not unvarying.
For example, the alternation of /1/ with [r) has been noted, unaspirated stops are voiced word-medially, and in a syllable-final position -4 /s/ is pronounced [t) and the laryngealization contrast is neutralized.
All of these alternations, however, are completely predictable in any given environment, a fact which has by no means been lost on those analyzing the Hangul script.Taylor (1980:68), discussing the script's alphabetic aspects, comments, 'In Hangul the ideal of one symbol for one phoneme is almost realized.'Coulmas writes, 'Of all the systems that were actually invented as writing systems, the Korean script comes closest to treating distinctive features as the basic units of representation°( 120).
DeFrancis goes even further, declaring, 'Korean as written today is more accurately designated as morphophonemic.That is to say, changes in pronunciation are generally not indicated in the spelling if they can be predicted from the environment' (193).
In Hangul, every spoken segment is accounted for in the script, and the phonetic value of any given sign can be ascertained from its environment.Such characteristics would not only tend to indicate that the Hangul script is an alphabet, but a very good one at that.
Taylor, however, stresses the syllabic aspects of the script as well,5 finding certain advantages to the fact that the prixiary visual object is a syllable rather than a phoneme: Sequencing and grouping sounds can be stages in word identification.Problems associated with these stages can be minimized in a syllabary where the syllabic breaks within a word are immediately apparent and a word requires only a short array of letters... Another advantage of a syllabary is that a syllable is a stable and concrete unit to compare with a phoneme.Often a consonant phoneme by itself cannot be pronounced or described until it is paired with vowels to form a syllable.Not surprisingly, a syllabary is easier to develop and to learn than an alphabet.Young children find it easier to segment words into syllables than into phonemes.(70)Coulmas, too, notes the advantages of the script's syllabic arrangement after observing its phonemic accuracy (120), and does not venture to classify it as either alphabetic or syllabic.Among other comentators, Gaur emphasizes the syllabic organization of the Hangul (84-5), although few scripts better meet the criterion of approaching the ideal of one sound per phoneme (119).
In DeFrancis' view, Hangul is no more syllabic than he sees the Indic scripts as being (193); he goes so far as to assert, 'Korean can be called syllabic only in the same sense that English can be called logographic because it groups its letters into words' (192).This, however, would seem to overlook Taylor's arguments regarding the different approach to the script necessitated for the reader by this different arrangement.H. Jensen calls Hangul a 'pure alphabetic script' (211), while Diringer describes it as 'practically an alphabet ' (1968:352).
One apparent source of disagreement is terminological.To DeFrancis, Lukoff, and N. K. Kim, the component members of the syllabic blocks are letters of an alphabet, while for Taylor the blocks themselves are the letters, and J. P. Kim (1983) seems to use the term interchangeably.
Kim does also use the term 'syllabigraph' to refer to these units; he credits typographic designer Ann Sang-oo for coining this word, 'for lack of an existing one to express the way Korean units are constructed... Hangul combines the features of an alphabet and syllabary' (22).
A factor which may impel scholars to typologize such a script as an alphabet is that such prominent theorists of the subject as Gelb ( 201) and H. Jensen (52-3) explicitly regard alphabetic scripts as more evolved and therefore more advanced.To acknowledge the syllable-based aspects of a script might therefore seem to diminish its prestige by implying that it is somehow more 'primitive.'In this connection, it is worth noting, with Gaur, that some scripts do not shed their syllabic characteristics to evolve into full-fledged alphabets simply 'because syllabic scripts are an excellent vehicle for the representation of a large number of languages' (119).
It also remains true that the Korean script is a work of genius by whatever name one chooses to refer to it.DeFrancis aptly describes King Sejong, the script's reputed inventor'who ruled during the fifteenth century, as 'a monarch who, if rulers were ever measured by anything besides military exploits, would surely rank among the foremost of those who have appeared on the stage of history' (188).
In any event, while the Hangul writing system's phonemic representation is nothing short of remarkable, its syllabic orientation, as is true of the Brahmi scripts, is significant enough that it cannot be ignored.
Neither Hangul nor the Brahmi family of scripts may be classified as either alphabetic or syllabic with complete accuracy.One might therefore pause to consider where they fit relative to one another on a continuum between the two script types.
A particularly striking contrast between the two writing systems is the inherent vowel of the Brahmi scripts as opposed to what in Hangul might be considered an 'inherent initial consonant.'No syllabic block may appear with its initial position unfilled; if there is no pronounced syllabic onset, 6 /u/ is written but remains silent.Gale in fact notes that the script originally also had three other silent initials: , o , a , but that 6 was eventually substituted for them (44).
As a result, every written Korean syllable must include an onset of some sort and a vocalic nucleus, although the coda remains optional.In Brahmi scripts such as Devanagari, however, the consonantal character conceived as the most significant element of a syllable may appear in certain circumstances with no following vowel if a virEma is attached.
This indeed is the fundamental difference between the two; in Hangul the vowel which modern theory refers to as the syllabic nucleus occupies the central and most prominent position, while in the Brahmi scripts, it is the consonant immediately preceding this vowel that is considered the basis upon which the rest of the syllable is built.
Immediately preceding consonants, conjoined to this segment, are considered part of this syllable, as Lambert (76) explicitly states.Also indicative of this is the fact that, if in Devanagari the vowel [i] is pronounced after a consonant cluster such as [str-], the vowel-sign is written before the entire cluster:*T Hangul holds a more "modern' conception of the syllable.It is also more regular and more linear in its organization of the syllable; consonants preceding the vowel are always written above or to the left of it, while those following are always below it.Brahmi vowel diacritics, on the other hand, may appear in any direction from the consonant, and even, in the Thai and Bengali scripts, on two sides of it.
It may therefore be concluded, on the whole, that while neither Hangul nor the Brahmi family of scripts is completely alphabetic, Hangul comes much closer to fitting this description.
Nevertheless, the relative typological similarity between the two writing systems, coupled with the recent origin of the Korean script, inevitably raises the question of whether any of the Brahmi scripts might have had some influence on the shaping of Hangul.Of course, by far the greatest outside influence on Korean culture was China, and the Hangul syllabigraphs certainly bear a greater casual resemblance to Chinese characters than to those of any of the Brahmi scripts.DeFrancis affirms, 'What Sejong did was to adapt the Chinese principle (f equidimensional syllabic blocks by grouping the letters that comprise a Korean syllable into blocks separated from each other by white space' (191).
The fact remains, however, that Hangul is much closer typologically to the Brahmi writing systems than to that of Chinese.
H. Jensen reports that before the invention of Hangul Koreans had obtained some utility from various Chinese methods of rendering unfamiliar sounds by adapting ,?.xisting characters to syllabic usage and assumes tAat the Koreans thereby became aware of the syllabic principle (179,211).Gale (1912) argues that one particular set of syllabic characters was in turn inspired by the Devanagari script (42,).An indirect relationship at least is thus demonstrated.
Moreover, a number of scholars, among them Gaur (85) and Lee (6-7) suggest that the Sanskrit and Tibetan languages as well as the scripts with which they were written would quite likely have been known to literate Koreans, and Lee points to these as likely sources for the alphabetic aspects of Hangul.
H. Jensen also mentions a Korean writing system known as the PumsQ script, developed before the time of Sejong, which is used 'in Buddhist ceremonies of prayer and sacrifice for the transcription of foreign Sanskrit words' (216).This script was apparently fairly closely modeled on the Tibetan script.
DeFrancis, too, names India as a likely, if perhaps indirect, source of alphabetic principles (186).
Indeed, unless we are to believe that Sejong and his assistants conceived of representing a single sound with each sign entirely on their own, it is most difficult to imagine from what other source they might have learned of this principle.
Finally, one other question remains from the anomalous typological status of these two writing systems, one of which represents a very significant portion of the world's languages and population, while the other, although isolated, nevertheless presents linguists with an impressive specimen of phonemic analysis.The failure of most commonly accepted definitions for syllabic and alphabetic systems of writing to include such important scripts and script families would seem to suggest that a new typological category is needed to fill this void.
Suggestions such as 'alphabetic syllabary,"alphabet on syllabic principle,' or 'semi-syllabary' might not be the worst compromise, for the time being at least, as they take into account the elements found in these writing systems.
Despite the differences that do exist between Hangul and the Brahmi scripts, they clearly belong together in such a category.Although there is no definitive evidence the majority of scholarly opinion is reasonably confident that the Brahmi script was derived from or at least inspired by a West Semitic source; see especially Shapiro 1969, Masica 1991:133-4, H. Jensen 1970: 368-70,  and Diringer 1962:144-5.In rather greater doubt is its precise date of origin.Diringer places it in the seventh century B.C., While H. Jensen (363) asserts that 'literary evidence shows it to have been in widespread general use in the fifth century B.C.' Masica, on the other hand, argues strongly that the script was still quite young in the time of Asoka, after whom the inscriptions bearing the first clear example of the Brahmi script are customarily named.

2.
For a comprehensive inventory of Brahmi scripts, see H. Jensen 361-404, or Diringer 1968:257-351.Lambert identifies those found in the text with Bombay printing houses and the Marathi language, preferring the latter for Sanskrit and Hindi (21,102).In practice, however, associations are less rigid; Coulson as well as Snell & Whitman (1989) use the Bombay characters for their respective textbooks on Sanskrit and Hindi, and Katzner's (1977) sample of Hindi includes the Bombay characters, while the Marathi sample includes the other set.
The Bombay characters will be used in this discussion as they seem both more esthetically pleasing and easier to produce.The subject of Hangul's logographic aspects is briefly entertained in Taylor's article as well (73).This is based largely on the fact that some Korean words are monosyllabic, so that one syllabic block stands for one word, such as g /talk/ 'hen.' This, however, might more appropriately be ascribed to the script's syllabic aspects.
mentioned by Masica 150 and Lambert 103 is an effort in Marathi to regularize initial vowel signs so that they consist of the basic 31 plus the post-consonantal allographs:f31[i], 3 [1],3g JT [5) , 34-[v], 3 [e],[ail.However, this has not gained widespread currency and is certainly not in evidence in the following sample, necessarily brief and tentatively transcribed due to the poor quality of the printed original, adapted from Katzner 189:A-1,^i 47-34-V4 3 341 c4ci .mala ugic nhuk nhuk asa gosti "dthvata..I have a sort of hazy recollection of certain events.5.