To teach Confucian morality, Sejong ordered the compilation of the Samgang haengsil (“Illustrated Guide to the Three Relationships”), a book of stories illustrating Confucian virtues. Unlike its earlier models, Sejong’s book contained examples drawn from Korean as well as Chinese sources. The king took care to have the book provided with large illustrations, aware that the vast majority of his subjects could not read. He issued statements encouraging the teaching of literacy to all classes of society, including (with a generosity unusual in his time) to women; but with a writing system as complex as idu and an educated class intent on keeping the privileges of literacy to themselves, nothing much happened.
Despite his scholarly bent, Sejong was no ivory-tower philosopher. His people needed more than morality and court ritual. Most importantly, the growing population needed food. Sejong set his academy to work on numerous scientific and technological projects directed toward increasing agricultural production. He commissioned a geographical survey of Korea; invented and distributed a rain gauge, instituting a nationwide meteorological network; and implemented irrigation systems. The optimal timing of sowing and harvest required an accurate calendar, he realized, not one calibrated to a Chinese latitude. He would have to derive a Korean calendar from scratch. In the course of his calendrical research Sejong invented (and/or had invented by his academy) an astrolabe, a water clock, and a sundial. His Publications Office printed seven books on agriculture and 32 calendars; it must have weighed on him that his farmers couldn’t read them.
King Sejong opened a medical school and even encouraged the education of female physicians so as to improve the health care available to women (who could not with propriety see a male doctor unless very seriously ill). He sponsored a compendious work of herbal medicine in 56 volumes, which again broke with Chinese tradition by emphasizing native Korean herbs, their uses, and where to find them.
For many years, beginning in 1422, Sejong worked on revising the Korean law code. Aware that bad law would make for bad precedents, he personally reviewed each article of the code with his legal scholars. Throughout his reign he showed a passion for justice, working to improve prison conditions, set fairer sentencing standards, implement proper procedures for autopsies, protect slaves from being lynched, punish corrupt officials, set up an appeals process for capital crimes, and limit torture. Nevertheless, one problem continued to vex the king: the litigation process was carried out in Chinese. Were the accused able to adequately defend themselves in a foreign language? Sejong doubted it.
Despite his achievements in agriculture, music, science, printing, and jurisprudence, Sejong was repeatedly stymied by the fact that his subjects couldn’t read. How could they learn about advances in technology if they couldn’t read? How could they benefit from moral philosophy if they couldn’t read? How could they defend themselves properly in a court of law if they couldn’t read?
Something must be done about it. Quietly, without telling his academy what he was about, Sejong set out to invent a script that matched the Korean language and could be easily learned by everyday people. The existing script he knew best was Chinese, but he knew that there were other ways to write. His government’s school for diplomats offered classes in Japanese, Jurchin, and Mongolian, besides spoken Chinese. Mongolian was taught in both the traditional vertical Mongol script and the newer ’Phags pa script. ’Phags pa was a squarish akara-based script modeled on Tibetan that had been developed at the command of Kublai Khan by the lama ’Phags pa Blo gros rgyal mtshan and completed in 1269. Kublai intended the script to be a universal writing system that could encode all the languages in his empire; in practice it was not used much. Nevertheless official edicts were often issued in ’Phags pa and it was useful for transcribing the correct pronunciation of Chinese characters. When the Mongol Yuan dynasty lost its hold on China in 1368 most people dropped ’Phags pa with relief.
Sejong may well have studied ’Phags pa. He probably realized that some scripts directly represented their languages’ pronunciation rather than whole morphemes, thus using far fewer symbols than the logographic Chinese script. Sejong decided that he too would create a phonological script. He studied all the phonological science available at the time. He also learned spoken Chinese, becoming one of the few Koreans who could speak the Mandarin dialect of the time as well as read Classical Chinese.
Chinese phonologists had recognized that syllables contained two parts, an initial (or onset) and a final (or rhyme). Sejong, perhaps inspired by the representation of consonants in ’Phags pa, realized that a rhyme could have two parts (today known as the nucleus and the coda), and that the sounds that could end a syllable were the same sorts of sounds as those that could begin a syllable. In other words, the coda and the onset were both filled with consonants. Thus in a word like kuk, “country,” the [k] sound at the end and the [k] sound at the beginning were “the same thing” and could be represented with the same symbol. Sejong had discovered the phoneme. He went well beyond this discovery and established that phonemes fall into a number of different classes according to traits, or features, which they possess, such as whether they are vowels or consonants, where in the mouth they are pronounced, and whether they are aspirated. Although Sanskrit grammarians had organized the Indian alphabets according to these same sorts of classifications, Sejong took the unusual step of incorporating these phonological features into the design of the individual letters he created.
Known today as han’gl, the writing system Sejong invented was a wonder of simplicity and linguistic insight. To make it, he first systematically analyzed the phonology of his language – a job that, like the creation of a Korean calendar, he had to undertake from scratch, as no one had ever yet cared to subject vernacular Korean to linguistic study. First he divided consonants from vowels; that is, he divided the sounds that occur in the onset or coda of a syllable from those that constitute the nucleus of a syllable. Then he divided the consonants into five classes according to where in the mouth they are pronounced (the place of articulation, in modern terms), guided by the Chinese philosophical principle of analyzing almost everything into five classes to match the five elemental agents of Water, Wood, Fire, Metal, and Earth. The categories he arrived at were the labials (made with the lips, e.g. [m]), the linguals (what we now call alveolars, made with the tongue tip touching the alveolar ridge behind the top teeth, e.g. [n]), the dentals (made with the tongue tip touching or close behind the lower teeth, e.g. [s]; this class is today known as the sibilants), the molars (what we now call velars, made with the back of the tongue against the soft palate inside the back teeth, e.g. [k]), and the laryngeals (made in the throat, e.g. [h]; these are also known as glottals).
Having thus identified and classified the consonants of Korean, the next task was to assign them each a graphic shape. At this point Sejong had a stroke of absolute genius. He knew that the most basic Chinese characters had originally been pictograms and ideograms, though most characters of the developed script were compounds made up of a semantic radical and a pronunciation clue. The pictograms at the root of the Chinese script were pictures of things, as indeed were the pictograms at the root of cuneiform, Egyptian hieroglyphs, and even the Semitic alphabet. Sejong, conscious that he was departing from Chinese practice in designing a script that recorded individual sounds rather than whole morphemes, created a wholly new form of pictogram. He drew pronunciations. Choosing one consonant in each of the five categories as basic (the one he considered least harshly articulated), he drew schematics of the formation of the consonant sounds in the mouth (see figure 11.1).
For the basic labial, [m], he drew a pictogram of a mouth, , influenced by the Chinese character/pictogram for mouth, . For the basic alveolar, [n], he drew the shape of the tongue as seen from the side, with its tip reaching up as it does when it touches the alveolar ridge: . For the basic dental, [s], he drew a schematic of a tooth, , perhaps influenced by the Chinese character for tooth, , which depicts the incisors inside an open mouth. For
the basic velar, [k], he again drew the shape of the tongue, this time with its back raised upward to touch the soft palate: . For the basic laryngeal he drew a circle depicting an open throat: . What precisely this letter stood for in Sejong’s time is unclear, as it now stands for the absence of a consonant in the onset of a syllable. What Sejong may have had in mind was the open throat preparatory to the voicing of the upcoming vowel, considering this to be a type of syllable onset in contrast to the closed throat of the glottal stop or the tensed throat that yields the breathy [h].
Armed with five basic shapes for the least harshly articulated consonant of each class, he proceeded to build the remaining consonants around these basic shapes. Plosive consonants were harsher than the nasals, while the affricate [t∫] was harsher than the fricative [s]. These harsher sounds were given an extra stroke: was [p], derived from [m], (written so as to require one extra stroke); was [t], derived from [n], was [t∫], derived from [s], . The glottal stop, , was similarly derived from the open-throat onset with a line on top of the , but the glottal stop and its symbol have since gone out of use in Korean.
Figure 11.1 The derivation of the han’gãl letters from their pronunciations. Above, the positions of the lips, tongue, and throat that Sejong used to derive the basic shapes of the consonants. Below, the derivation of further consonants from the basic ones. Note that and are now a single letter.
Korean has a set of strongly aspirated stops and affricates, and to mark this set of yet harsher sounds the basic symbols were given a further modification, usually a further stroke: was [ph], was [th], was [t∫h], was [kh], and was [h], the aspirated version of the open throat. Korean also has an unusual set of consonants that are held longer than regular consonants and are similar to doubled consonants in other languages except that they are pronounced with a tense vocal tract. These are sensibly written with double letters: and . (Except for , these tense phonemes seem not to have existed in Sejong’s time, and he used the doubling of a consonant letter to indicate the voiced consonants of Middle Chinese. The Modern Korean tense consonants have arisen since then out of earlier consonant clusters.)
A few consonant phonemes still needed symbols. One of them, pronounced either as [l] or [], depending on context – [l] when in the coda of a syllable or when doubled, and [] elsewhere – has an alveolar articulation, so Sejong gave it yet another variation on the basic shape: . Another exception was the velar nasal, []. It is not entirely obvious why Sejong did not consider this the basic sound of the velar series, since the other nasals were considered basic in their classes, but perhaps it was because [] is unique in not being allowed to begin a Korean morpheme – making it a poor exemplar for the alphabetic principle. Instead he classified this sound with the laryngeals – a reasonable decision (though at odds with modern classification systems) given that in the pronunciation of [] the mouth is blocked off completely at the very back, leaving the sound to resonate only in the throat and nose. Sejong gave it a symbol based on the open-throat/absence of consonant: plus a short vertical line on top, . Since Korean morphemes do not begin with [] and the absence of consonant is not indicated at the ends of syllables, the two symbols have become conflated since Sejong’s time: at the beginning of a syllable stands for no consonant at all, while at the end it stands for [].
Having created symbols for consonants, Sejong turned to vowels. Again, he went about it philosophically and systematically. Of the vowels he chose three as basic: the one made with the lowest placement of the tongue [Å], the one made with the most forward placement of the tongue, [i], and one of the two made with the tongue drawn up and back, [] (the unrounded vowel transliterated as ). He gave these three vowels symbols representing the mystical triad of heaven, earth, and humankind. The first he represented as a dot, ·, to represent the round heavens (this phoneme is no longer used in Korean). The second he made a vertical stroke, |, representing humankind, standing upright. The third became a horizontal line, the flat earth: Other vowels were formed as combinations of these. The heavenly dot has evolved into a short line, easier to write with a brush. Thus is [a], (transliterated ), is [u]. Diphthongs starting with a [j] onglide received a second dot (nowadays a line) to make . Other diphthongs were made by combining vowel symbols, though two of these are nowadays pronounced as single vowel phonemes, namely (made from and originally [aj] but now [ε]) and (originally [Δj] but now [e]). Two other combined vowels, , originally [oj] and [uj], are pronounced [we] and [wi] in some modern dialects but as single vowels [ø] and [y] in others.
It was time to put the consonants and vowels together. In Chinese each character is pronounced as a single syllable, and all are written the same size, regardless of complexity. Accordingly, Sejong grouped his newly created letters into syllabic blocks. The initial consonant went at the top left-hand side of the block. Next came the vowel, written to the right of the consonant if it contained a vertical stroke and otherwise below it. Thus is the syllable ku, and is ka, while a syllable without an initial consonant receives the null sign: ! is a. A final consonant is written at the bottom: is kuk, “country,” and , mal, is “language.” Syllabic blocks together make up words: is han-kuk-mal, the Korean language.
If a single consonant occurs between two vowels it is generally assigned to the onset of the second syllable, unless it is []; thus the word ha’nl, “sky,” is spelled , but pang’ul, “bell,” is with [] at the end of the first syllable block and the null consonant starting the second. Exceptions in modern orthography (following Sejong’s personal practice but revised from most other earlier spelling traditions) arise in order to keep constant the spelling of a morpheme. Thus the verb root (mk, “to eat”) will be spelled the same, regardless of whether the suffix added to it starts with a vowel or a consonant; the root is clearly visible in both . The same morphemic convention allows consonant clusters to be spelled in syllable codas, even though they are not pronounced. The Korean word for “price” is , pronounced [kap], with only a single final consonant. The extra is there because if a vowel-initial suffix is added, there will be room to pronounce it: , with the -i suffix marking the word as the subject of the sentence, is pronounced [kap∫i] (s being pronounced [∫] before [i]). A similar principle drives the spelling in English of a word like iamb, with a silent unpronounceable letter that is heard only when a vowel-initial suffix is added, as in iambic.
The constant-morpheme principle is also applied to the spelling of single final consonants. The number of consonants that can be pronounced in Korean syllable codas is quite restricted. The phonemes represented by are all pronounced as , [t], at the end of a syllable, but they will reappear with their distinctive pronunciations if a vowel-initial morpheme is added. Similarly the tense and aspirated plosives sound like their plain versions in codas but will be spelled according to how they sound when they are followed by a vowel and allowed to show their true colors.
In addition to symbols for all the phonemes of Korean, Sejong developed a way to record the pitch-accent of the Korean of his time, placing one dot, two dots, or no dots to the left of a syllable to indicate the pitch with which it was to be pronounced. (Pitch-accent is similar to tone except that only one syllable of a word receives a distinctive pitch and the pitch of other syllables is predictable from that one.) Standard modern Korean has lost its system of pitch-accent, and the side dots are no longer used.
While he was about it, Sejong also incorporated extensions of the new script so as to be able to transcribe Chinese. With these modifications, han’gl could be used to teach the proper pronunciation of Chinese.
The system that Sejong ended up with is indeed the paragon of scripts (figure 11.2). It matches the phonology of the Korean language perfectly, and it is elegant and easy to learn. It encodes a large range of linguistic insights. In common with the Indian alphasyllabaries it recognizes the difference between vowels and consonants, a difference that is entirely ignored in linear, Greek-descended alphabets. It encodes individual phonemes, but also provides phonologic
al information on a smaller scale than the phoneme, again in contrast with Western alphabets that recognize only phonemes. Similar phonemes are given predictably similar shapes; this property not only reflects the linguistic insight that phonemes are composed of smaller distinctive features, but is a crucial factor in making the han’gl alphabet so easy to learn. By contrast, the Roman alphabet abounds with pitfalls for the learner: the graphic similarity between the letters E and F corresponds to no phonological similarity at all, nor does the similarity between O and Q, while the similar phonemes written T and D look entirely dissimilar.
Figure 11.2 Han’gl, the Korean alphabet, listed in the South Korean order. The vowels are here presented separately, but in the alphabetization of actual words all vowel-initial words begin with the dummy consonant and will therefore appear after words beginning with and before words beginning with . The Korean names of the letters are given in Romanization. The name of each consonant contains that consonant twice – as the initial and final letter – which serves to indicate how the letter is pronounced when syllable-initial and when syllable-final. In the phonetic transcriptions given here, the tense consonants of Korean are transcribed with an asterisk. While common among linguists, this usage is not official IPA; these consonants do not yet have a standard IPA transcription.
The Writing Revolution Page 24