|
| 阅读上一个主题 :: 阅读下一个主题 |
| 作者 |
 ZT:东亚文字是如何妨碍东亚人的创造性的.(英文,较长,需耐心读) |
 |
启明
[博客]
[个人文集]
游客
|
|
|
作者：Anonymous 在 罕见奇谈 发贴, 来自 http://www.hjclub.org
From W. Hannas'
The Writing on the Wall: How Asian Orthography Curbs Creativity (2003).
Chapter Eight: The Concrete Nature of Asian Writing
"Cognitive science often carries on as though humans had no culture, no significant variability, and no history."
---Merlin Donald
East Asian orthography is enormously complex. As noted in the previous chapter, the number of units that make up the inventory of Chinese symbols and the complexity of their design is mind-boggling. Literacy in Chinese means being able to read and write--as a base line--some 4,000 to
6,000 different characters. Although many graphs share common elements,there is no efficient way to describe how these elements combine into characters, as evidenced by the many indexing methods used in dictionaries and by the hundreds of coding schemes that have been designed for
inputting characters into computers.
Complicated as the characters are, their relationship to Chinese Mandarin is straightforward. Most characters are pronounced as a single syllable.' The combination of a distinct character and a syllable sound identifies
a morpheme. This formula can be stretched pretty thin, for example,when foreign words taken into Chinese are analyzed as having as many morphemes as syllables, or when morphemes whose meanings have been lost are shielded by their unique representation. By and large, however, Chinese does hold to a one-syllable-per-character-and-morpheme format,
which though debilitating in terms of its sensory effect,2 does mitigate some of the complexity.
Japanese inherited the Sinitic complexity and added convolutions of its own. Unlike Chinese, a character in Japanese normally has multiple pronunciations depending on the number of indigenous and Sinitie morphemes associated with it-usually two but sometimes a dozen or more.
Although other orthographic clues such as the use of okurigana (kana after the character) and furigana (kana wedged between columns of characters) help identify a word, the correct reading ultimately depends on one's ability to interpret context. Moreover, Japanese uses not one but three
scripts mixed together according to etymology and other conventions. Complex as the system appears to be, the real difficulty lies in its operation, which has provoked unflattering comments even from those who have learned it well.
3 Koreans also mix Chinese characters with a phonetic script, but there are differences in the way the formula is applied. Japanese expect that characters will be used wherever they can be used. Koreans tend to use them only as needed to clarify a text. Of course, this need is selffulfilling, since the availability of Chinese characters discourages Koreans from developing real, phonetic solutions to ambiguity. Another difference be-
tween the two systems is that Korean uses characters only for Sinitie terms. Indigenous words are not written with Chinese graphs, so the complex mapping system used in Japanese is absent in Korean. That said, there is
still the problem of learning 1,800 or more characters and the added complexity (from an information processing standpoint) of adjusting the size, shape, and position of hangul letters to fit in a square matrix.
Vietnamese, the fourth major East Asian language to have used Chinese writing, replaced the characters entirely with a phonetic alphabet. Yet for all its simplicity relative to other East Asian writing, Vietnamese orthography retains a degree of complexity uncommon in an alphabetic system.
Unlike English, which uses just 26 letters to represent the language's 40 phonemes, Vietnamese uses 29 letters for 30 phonemes plus five diacritics placed over and below vowels to represent tone. Since some of these vowel letters already bear special markings, the combination produces a text which, in appearance at least, seems almost as cluttered as the Chinese based writing it replaced. This impression is strengthened by the Vietnamese practice of putting spaces between syllables, so that viewed at a
distance the units do superficially resemble character text.
4 Why does Vietnamese orthography emphasize syllables instead of words?
One could argue, as indeed many have, that the syllable is the most natural unit for representation. I have already discussed the abstract quality of subsyllabic segments and pointed out that their recognition for most people
depends on access to a prior analytical model, usually an alphabet. The same is true of the larger-than-syllable-sized units that comprise most of a language's words. What is or is not a word is no more apparent from natural speech than is a phoneme. Both are artifacts of alphabetic literacy, established by introspection and convention, although once established they become part of one's representation of reality.
The decidedly unnatural depiction of words in writing is a product of a long tradition of abstraction, comparable to the abstract thinking that underlies phonemic orthography. The two devices-letters and word division-are complementary features of a system that has gone beyond the representation of concrete elements into the abstract design of the language. Vietnamese orthography's lack of word division marks the script as a transitional artifact, one that imported the trappings of the Western orthographic tradition a fully elaborated phonemic alphabet-but not the
other major feature common to alphabets, because there was and still is no consensus on what are the language's words.
There is no reason why the word Viêt nam could not he written in Vietnamese as a linked sequence, as it is in English and other alphabetically written languages.5 Doing so would allow the script to shed its obligatory use of diacritics and perhaps other symbols as well, since the information given by word boundaries would make much of the detailed phonetic specification redundant. But with no agreement on what constitutes words, Vietnamese has no choice but to represent the only other linguistic units
on which, by virtue of its concrete nature, agreement does exist, namely, syllables, and represent them in enough detail to compensate for what they lack in abstraction. In so doing, it betrays its cultural, if not etymological
history as a descendent of the Sin itic family of syllable-based scripts.
This failure of East Asian writing to transcend concrete syllabic units is obscured by the variety of systems in use there, ranging from full use of characters (in China), through partial use (Japan, South Korea), to nonuse (North Korea, Vietnam). Yet a close look at these individual systems reveals that they all share this common characteristic. East Asian writing, in principle and practice, is and always has been syllabic.
As we have seen, Chinese in its modern báihuà (colloquial) and traditional wényán (classical) styles is written in what can best be described as a morphosyllahic writing system.6 Graphemes correspond to morphemes and both correspond to syllables. There is some question about which of these connections is the primary link between the language and writing system. Since several dozen characters may have the same monosyllabic pronunciation one can argue that characters map primarily onto the morphemes of the language, where the fit is nearly one-to-one; their identification with syllables would be a consequence of the fact that the morphemes are monosyllabic. On the other hand, it is easy to find exceptions to the one character = one morpheme formula, whereas almost no characters in
Chinese are pronounced as more or less than a single syllable.
Nor does any part of a character correspond to a subsyllabic element. To the extent that Chinese characters can be usefully analyzed at all, one often finds graphic components that reappear in different characters with the
same or similar sound, and hence are associated with a given sound or class of sounds. These socalled "phonetics," like the full characters in which they appear, equate to a syllable and nothing smaller. The only analysis that
can be said to have occurred is the de facto willingness of Chinese to ignore inconsistencies in tone, initial consonant aspiration, and other types of homorganic variation that arose in the language's evolution (or were allowed for convenience's sake when the characters were being constructed) so these elements could be shared more widely throughout the corpus.
Conceivably, the development of phonetic variation between syllables written with the same sign, coupled with China's own tradition of phonological study, might have led to changes in the graphic representations to reflect, in a more or less principled fashion, the evolving sound changes
if the "phonetic" were the only part of the character and if attention had been drawn exclusively to this feature. Users would have been forced by the growing mismatch between symbol and sound to create written distinctions which, lacking any other motivation, might have been keyed to changes in the pronunciation itself. It is not hard to imagine how such a trend could lead to a segment or even feature-based orthography. Had this occurred, history would have been different.
Instead, Chinese writers and lexicographers preferred to clarify the ambiguities that developed between written symbols and their referents by attaching components borrowed from other characters on the basis of their
ascribed meanings. Intended to remove ambiguity in cases where a character had or developed multiple meanings, the more important effect of these semantic-plus-phonetic (xíngsheng) characters, I believe, was to direct attention from phonetic inconsistencies that, unmasked, would have been intolerable. With the introduction of this new type of symbol, which now comprises 85 percent of the corpus, vague or misleading cues about the pronunciation of a character could be accepted given the character's dual role as a morphological unit. Buoyed by its semantic association, the accuracy of its phonetic representation became less critical.
These semantic determinatives (known as "radicals") prevented Chinese writing from following the path toward full phoneticization (and ultimately alphabetization) taken by the world's other major orthographies. With less pressure on the system to develop phonetically, Chinese writing locked itself into an iconic relationship with the elements it represents, where all parts of the sign relate to each other as whole units without the benefit of abstraction. Graphemes lack a discrete correspondence with the elements of the sound they represent, and both graphemes and syllables equate to ill-defined "units" of meaning. The system from top to bottom is holistic and utterly concrete.
What masquerades in Chinese as complexity is simply the result of the writing system's failure to transcend the concrete world of perceived phenomena. The situation is only marginally different for the other East Asian
languages that Chinese influenced. The thousands of Sinitic loans that inundated Japanese, Korean, and Vietnamese inculcated users with the same concrete linguistic habits that paralyzed Chinese and with the same unrelieved penchant for iconic representation. Units of meaning were lined up with natural syllable sounds. Abstraction and analysis were not part of the equation.
Orthographically the situation was no different. Japanese use of Chinese characters is entirely, syllable-based. Sinitic readings (onyomi) of characters differ from the original Chinese sounds by the addition of an extra syllable
to cover final consonants in some of the original syllables. Like Chinese, the representation is syllabic. Native readings (kunyomi) of Chinese characters also follow a strict syllabic format. Although many Japanese morphemes end with consonants, the practice is to render the consonants orthographically in kana as part of a consonant-vowel (CV) syllable. Now if the representational rules for kanji were based entirely on meaning, the whole root of an inflected morpheme, including its consonant final, would
be subsumed by the character and its derivational inflection would be handled by one of the a-i-u-e-o kana signs instead. The phonetic mandate to depict CV syllables apparently was strong enough to override this morphological feature.
The above remarks apply to the kanji part ofJapan's mixed script. kana, for its part, is the archetypal syllabary. Derived from Chinese character shapes, the two kana systems are able through combinatorial formulas to render the language's spoken syllables with just 48 symbols, without compromising the syllable format. For example, the long syllable kyO, which has no single kana sign, is written with the three kana きょう corresponding to ki + yo + u. The English equivalent would be to write the word "today" as
"tee" + "oh" + "dee" and so on using syllables for each phonemic segment. Geminate consonants, which also lend themselves to representation as segments, are handled in Japanese the same way by syllable signs. Far from reflecting the language's syllable-based phonology, kana helps create it by imposing syllables where none exists. To call Japanese writing syllabic is an understatement.
This brings us to Korean. Chinese characters are used in South Korea, if at all, to represent monosyllabic Sinitic morphemes. These morphemes and the words they make up constitute some 60 to 90 percent of the Korean lexicon, depending on the type of text and how the words are counted. Although Sino-Korean vocabulary is not always written in characters, it is taught with characters in school, the theory being that words can be understood better by learning their morphosyllabic roots. This philosophy also governs education in the North, where instruction in Chinese characters forms a significant part of the school curricula, despite their nonappearance in most types of text.
Hangul, the phonetic orthography that is used alone or in conjunction with Chinese characters, never escaped the most salient feature of Chinese writing, namely, its focus on syllables. For all the ingenuity that went into designing hangu4 its inventors forewent many of the cognitive and me-
chanical advantages of an alphabet by arranging its letters not serially, in imitation of speech, but as small square blocks that emulate the shape of characters. While excused by some as an effort to highlight the syllabic aspects of speech (which indeed it does), historical evidence suggests that the convention was mostly a ploy to assuage the sensitivities of a China dominated court.
7 Instead of freeing Korea from China's linguistic straightjacket, the promoters of hangul went on to explain their creation as an instrument to help master the sounds of Chinese characters! From its inception, hangul was
linked linguistically and socially with the dominant East Asian writing system. If this were not enough, hanguPs designers diluted whatever cognitive benefits might have been had from its analytic nature by requiring that a
letter's size, shape, and relative position be defined by its location within the syllable block. This practice shifted attention away from segments and toward syllables, so much so that the syllable is regarded today by printers
and educators alike as the basic unit on which hangul fonts and formal instruction are based. It is, in essence, treated as a syllabary, albeit one whose structure is well motivated.
Ch'oe Hyon-bae (1894-1970), linguist, educator, and Korea's most outspoken advocate of replacing Chinese characters with all-hangulwriting, was aware of this problem with hanguPs syllabic orientation. He later designed a sequential system whose segments were modeled on a combination of hangul-like and other alphabetic letters (1946:145-87) to help overcome the limitations of what would later be called hangulword processing. I some-
times wonder if he was aware of this larger, cognitive issue as well.
The conclusion to be drawn is that all officially sanctioned writing in East Asia is syllabic in design or in practice, from the Chinese characters that began East Asia's literary tradition to the nominal alphabetic systems
that might have become truly alphabetic under more congenial circumstances. The complexity these systems display on the surface merely compensates for their failure to achieve the level of abstraction attained by the
two dozen-odd letters of a Western alphabet.
The Syllable in East Asian Speech
The prevalence of syllabic orthography in East Asia, and in the early development of writing everywhere, is a result of the syllable's concreteness.8
Unlike segments and words, whose identities are based on abstraction, syllables can be distinguished directly from speech itself. They are perceptually prominent and recognizable within the context they appear. There
is no need to enlist outside criteria. As such, they are natural units for representation.
The prominence of syllables in the history of writing is repeated in the ontogeny of linguistic awareness. Clinical observations and studies of preliterate children, adult illiterates, and people literate in nonalphabetic or-
thography point consistently to the ease with which speakers are able to isolate syllables, in contrast to the difficulty they have identifying segments.
Jacques Mehler in his study of early speech perception observed that infants discriminate syllables "from each other on the basis of a physical difference that is insufficient to allow babies to discriminate an identical difference in a nonsyllabic environment" (1989:197). Goswami and Bryant (1990) noted that preliterate children have no difficulty identifying the initial
sounds and rhymes of syllables (the same level of distinction achieved in traditional Chinese phonology), but go on interpreting alphabetic letters as syllables for some time into their reading instruction. Liberman et al. (1974)
found that half the American four-year-olds they tested could distinguish the syllables in English words, whereas none could distinguish phonemes.
Bertelson and Dc Gelden argued on the basis of their own experiments and a comprehensive review of the literature that adult illiterates in societies where illiteracy is the norm have trouble identifying phonetic segments and larger discourse units. By contrast, rhyming and other types of
syllable identification tasks are carried out by illiterates reasonably well.
The authors explain "rhyme decisions do not require analysis into segments proper and can be carried out at the level of syllables by appreciating some holistic sound identity or similarity" (1989:13-14).
If syllables are natural phonological units, some syllable types are more natural than others. Specifically, CV syllables are the first to be acquired in the normal course of speech development and the last to be lost in cases of
pathology. David Ingram (1978), who studied children's acquisition of syllable types, found that CV syllables are learned earliest, followed by those of the CVC and VC pattern. Conversely, Blumstein observed from the distribution errors of aphasic speech that CCV syllables are reduced to CV, and if a single consonant is lost in a CVC syllable, "it tends to be lost in the final position" (1978:195).
This elemental CV canonical type is also the one most commonly encountered in the Sinitic and indigenous phonologies of East Asian languages, and the form to which East Asian phonology generally aspires. As far as I am aware, there are no syllables in the area's languages with initial consonant clusters of the CCV type. Unlike the phonology of many European languages where initial clusters are common and at times complex (try saying the Russian word vzglyad meaning "look"), no contemporary East Asian language, including Vietnamese or the so-called Chinese "dialects," has this feature.9 Although present in earlier forms of Chinese, consonant clusters were lost entirely or submerged into other parts of the syllable.
Nor are there consonant clusters in syllable final position of the type VCC, despite their commonness in European languages (e.g., German strumpfs ifs for "of a stocking"). Although Korean hangul displays syllable final clusters, this is a morphological convention and not part of the phonological syllable. Either these hangul clusters are pronounced as a single consonant (for example, [tak], not *[talk] for the hangul form "chicken") if in isolation, or the "second" consonant becomes the initial consonant of
the following syllable: [tal] + [ki] meaning "chicken (nom.).10 Although syllable final single consonants of the type (C)VC do exist in most East Asian languages, they are perceptually less distinct than when they appear
in the initial position and are severely restricted in number. Their presence in the orthography in many cases reflects distinctions that have been lost or that applied to the extended community of speakers when the writing convention was laid down and were never part of any one speaker's competence.
For example, hangul permits a wide range of syllable final consonants in the orthography to replicate a word's underlying morphophonemic structure. Phonologically, however, only seven such final consonants exist, including the most commonly encountered set of oral and nasal stop
consonants -p, -t, -k, -m, -n, and --ng, that is, a maximally distinct set, and the consonant -l. According to Martin and Lee, "When the basic form of a word ends in something else, the 'something else' must be reduced to one
of these consonants, unless the word is followed . . . with a vowel" (1969:xxiii). Similarly Vietnamese writing distinguishes only eight orthographic consonant finals, which reduce in speech to the standard six-consonant set.11
Japanese phonology, the simplest of the lot, arguably has
no syllable final consonants. As noted above, consonant finals inherited from early Sinitic borrowings were given a vocalic supplement [ i ] or [ u ] and treated as the second syllable of a two-syllable morpheme. Final -n in Japanese is not treated as a consonant at all but as a "syllabic nasal."
Chinese phonology from the earliest period for which reliable reconstructions exist (circa An. 601) to the present evidences the same progressive loss of syllable final consonants. Cantonese and Southern Mm, which
most nearly approximate Ancient Chinese phonology, are both limited to the oral and nasal consonant set -p, -t, -k, -m, -n, -ng (Taiwanese and some Cantonese dialects additionally have a final glottal stop). Since none of the
oral stop finals are released, the sounds are more difficult to distinguish than in English and other Western languages, and in fact were treated in traditional Chinese phonological study not as consonants but as a class of
tones. The obligatory nasalization of vowels preceding final sounds -m, -n, and -ng in Southern Mm suggests the eventual erosion of this series in at least some of the Mm dialects.
In other Chinese languages the loss of final consonant distinctions is further pronounced. Wti reduced the -p, -t, -k to a single glottal stop. The -m final has merged with the other nasals, as in Mandarin, and in one dialect
(Ningbo) the -n and -ng endings merged into a single phoneme with its two phonetic values in complementary distribution. Mandarin, the national standard, has eliminated all vestiges of stop finals including the
glottal stop. Although the -n and -ngremain, they merge for some speakers after the high front vowel. 12
The prevalence of CV(C) syllables in East Asian languages to the exclusion of all other types contrasts with the full range of syllable forms present in Western languages. This fact is doubly striking given that three and possibly all four of these East Asian languages have no verifiable genetic connection with one another. Although extensive sharing of Sinitic vocabulary accounts for some of the phonological similarity, the same restriction on syllable types also applies to the indigenous part of non-Sinitic East Asian languages. The absence of consonant clusters and the tendency to eliminate all but the simplest syllables is characteristic of the area's phonology as a whole, an example of what linguists (for want of an explanation) label
an "area tendency."
I am not going to insist there is a causal relationship between this fact and the parallel use of syllabic orthographies in East Asia. Verifying such a hypothesis would require taking into account detailed historical data that are beyond the scope of the present study. Still, one cannot help but be struck by the observation that the alphabetically written languages of the West all have complex syllabic structures, while those of East Asia, whose
orthographies are syllabic, have syllables with relatively simple segmental phonologies. Causation, if it exists, could work in either or both directions: the reduced number of spoken syllable types in East Asia is one factor that
makes the operation of syllabaries mechanically feasible. On the other hand, the enforcement of a syllabic paradigm by the orthography could motivate literate speakers-and the rest of the population by example-to focus more strongly on the languages' syllabic aspects. Reversion to the canonical prototype would be an outcome of that process. Since alphabetically written languages can reflect the phonology of complex syllables as easily as they can the simpler types, neither of these influences would occur.
Whether there is a direct relationship between the prominence of syllables in Asian orthography and phonology, an indirect link, or simply a psychological one (reinforcement), the important fact for the present study is
this: East Asians when they are writing or speaking are focused on syllables. As such, their linguistic awareness is kept within circumscribed parameters. There is no need for them to project beyond the concrete level to analyze
smaller units or to synthesize larger ones. Despite their fabled complexity, East Asian languages use the most simplistic configuration possible: basic units of meaning (morphemes) map onto syllables of the simplest type,
and both are expressed by holistic, nonanalytic graphemes. Abstraction plays no part. Even Korean hangul and Vietnamese quoc ngu, which are analytical by design, adopt the syllabic configuration in practice.
To put the matter in bald terms, what qualifies in one (restricted) sense as dyslexia in the West represents the East Asian psycholinguistic norm. Pinker defines dyslexia as "a difficulty in reading that is often related to a
difficulty in mentally snipping syllables into their phonemes" (1994:322). In a famous article tilled "Teaching Reading by Use of a Syllabary," Gleitman and Rozin suggested that this conceptual hurdle could be overcome
by separating phonemic analysis from the basic reading process. Instead of requiring children to master two novel processes simultaneously, the authors proposed that the concept of reading be introduced through syllables, which are natural and present no new obstacles to the learner. The
problem of segmenting the syllables into phonemes could be tackled later (1973).13
Gleitman et al. went on to note in another study that five-year-olds "have the greatest difficulty of all in segmenting words or syllables into
phonemes .... this generalization holds for a variety ofjudgmental linguistic tasks (so-called metalinguistic tasks), of which reading acquisition is only one" (1989:93-94). This fundamental problem with literacy acquisition in the West has been noted by many linguists and psychologists. According to Bell and Hooper, "the structure of culturally based activities under the headings of language use, language play, and language art con-
sistently show that segments are less accessible than the groups of segments that serve as prosodic domains and rhythmic units. The dominance of syllabaries in the development of writing systems is a celebrated instance"
(1978:19). Morton called phoneme segmentation "the most important single aspect of learning to read" (1989:56). Scholes and Willis (1991:218) found that phoneme identifications could
Quote: be done easily by subjects who could read alphabetic orthographies. They could not be done, however, by prereading English-speaking children, nor by older children who read languages whose orthography is not alphabetic.
be done easily by subjects who could read alphabetic orthographies. They could not be done, however, by prereading English-speaking children, nor by older children who read languages whose orthography is not alphabetic.
The low incidence of dyslexia in East Asia can be connected specifically to the absence of any phoneme segmentation requirement in the orthographies. The price East Asians pay for the lack of this impediment, however, is an inability to identify abstract units of language. In an article
titled "The Ability to Manipulate Speech Sounds Depends on Knowing Alphabetic Reading," Read et al. (1986) tested two groups of Chinese-one literate in both the Chinese character script and alphabet pinyinl, the other literate in the character script only-for their ability to add and delete consonant phonemes. The study found that those literate only in Chinese characters did no better than Western illiterates at identifying phonemes, while those who had learned the pinyin alphabet did as well as Western subjects who were literate in their own alphabetic script.
This striking affirmation of the irrelevance of phonemic analysis for most East Asians reminds me of a story told to me by the late A. Ron Walton, who taught me Chinese linguistics when I was a graduate student. Many readers of the present book will remember Dr. Walton for his contributions to Chinese language pedagogy, but his major theoretical work was done in Wti (Shanghainese) phonology. Walton claimed, and I have no reason to doubt him, that one of his former classmates at Cornell who was also working with Wit phonology had his dissertation held up for months
by his refusal to include a phonemic analysis in the study. Now this was a pretty gutsy thing to do, especially at Cornell, but the student, who had acquired a deep appreciation of the psychology of Wit speakers, insisted that phonemes had no relevance to the community, and that the only linguistic unit that mattered was the syllable. The student ended up compromising with his mentor by appending a phonemic analysis to the dissertation.
The story is consistent with my own experience with nonstandard Chinese languages, which unlike Mandarin lack officially sanctioned alphabetic notations. Most of the dozen or so native tutors I had the pleasure of work-
ing with had exposure to an alphabet through English, but were otherwise linguistically naive. In most instances14 my tutors could not understand why I was so intent on cataloging segmental (and tonal) distinctions, and
were completely baffled by my search for phonemes. Although they had no trouble recognizing distinctions in their own speech at the syllable level, including some I could not hear myself, most were quite unable to tell me why such-and-such a syllable sounded different, even with their latent
knowledge of the English alphabet.
Typically what happened was as the studies progressed and we were able to record utterances in an alphabetic notation my tutors would start telling me to what part of a syllable they thought a particular distinction pertained and what the most appropriate letter would be to transcribe it. Finally (and incredibly) I noticed them altering their own pronunciation depending on the letters and tone marks I used to record what they had said! In the span of a few months, an informant would replicate the history of alphabetic awareness, moving from concrete syllables to segments and abstract phonemes, and ending at the point where orthography starts affecting pronunciation (confirming meanwhile what my professors had told me about
keeping my notes out of sight).
Since my interest was more in learning the languages than in doing accurate field studies, I was willing to overlook some of this sloppiness for the sake of expediency. But here is the whole point: unlike the Mandarin speakers in Read et al.'s study whose literacy in pinyinl biased them toward phonemic analysis, these informants initially were unable to analyze their own languages into segments, and in fact resisted the exercise (or just barely tolerated it). Despite their ability with the English alphabet, they
had never applied this understanding to their native speech, and hence behaved at first in the same way as those whose Mandarin literacy depended wholly on Chinese characters.
East Asians literate only in the traditional orthography share with preand illiterate populations in their society and the world over a fundamental inability to identify phonemic distinctions. We have seen in preceding
chapters that the ability to make these distinctions serves as an exemplar for analysis and abstraction in general. Just as the presence of an orthographic model for analysis in the West has stimulated the development of an analytic mindset, so has its absence in the East deprived East Asians
of comparable skills or, at least, of an early definitive opportunity to acquire them.
Significance of Tibetan and Vai Studies
Claims about the psychological effects of alphabetic literacy have been based largely on the impact of this orthography on Western society. Progressive changes in the dominant mode of orthographic representation in
the West--from pictographs through logographs (word writing), syllabaries, consonant alphabets and, finally, to fully phonemic writing--led to changes in the way literate Westerners address reality. Abstraction and analytic
thought, the sine qua non of Western philosophy, had their roots in the alphabet, which promotes these skills as a by-product of its operation. Habits of thought that began as linguistic functions were generalized to other ar-
eas, taking their place within the matrix of skills that form the basis for scientific creativity.
Others have noted the absence until recently of a creative ethic in East Asia coinciding with a neglect of analytic thought in the area's philosophical traditions. In the material arena as well, East Asia's technological
achievements show little of the creative discontinuity associated with the "breakthrough" science performed in the West. Innovation has been incremental, and as we saw earlier highly derivative, owing to East Asians' dispo-
sition to concrete thought, reinforced by the orthography. Unlike Western alphabets, which force an analytic mindset on their users, East Asian syllabic writing makes no such demand. As a consequence, abstract and analytic skills are less diffused throughout the population.
These parallel arguments, one the mirror of the other, are supported by events that together strongly suggest some element of mutual dependency. Objectified science and speculative philosophy coincided with the appearance of a phonemic alphabet, and failed to appear where the dominant
orthography lacked the alphabet's analytic characteristics. A connection between orthography and creativity is also supported by a review of the cognitive basis for creativity, which shows a close match between skills ac
quired through alphabetic literacy and those needed for creativity at each of its stages. The two types of data support the thesis that alphabetic writing has a reordering effect on thought and stimulates creativity.
The problem with this thesis is that I am attributing to one feature of a multifaceted culture, namely, its orthography, the wherewithal to induce effects that define the same culture. The argument would benefit significantly by examples of different orthographies in one culture producing effects associated with the other. It would be better still if the same cognitive effects claimed for one orthography were manifested in a culture that has no connection to either of the two under study. The creative skills displayed by first- and second-generation Asian scientists in the West who have learned an alphabetic script, read the literature, and exchange thoughts
in an alphabetic milieu constitute one such example. But I want to reach outside the languages and cultures we have already looked at for fresh material to support my hypothesis.
The reader will recall from Chapter 4 how Tibetan thought is distinguished from Chinese and other East Asian philosophy by its emphasis on logic and metaphysical speculation. According to Nakamura (1964:339),
Tibetans "main interest was in the strict definition of technical terms, establishment of scholastically detailed rules for their use, and the expression of all kinds of scientific thinking by means of syllogisms." These are
the very features Logan and others claim exist in Western thought but are lacking in East Asia's cognitive tradition. Logic, the apotheosis of analytical reasoning, in fact formed the core curriculum in Tibetan monasteries and
was taught to children at age ten (340).
A parallel difference is found between Tibetan and other East Asian orthographies. Tibetan writing, an adaptation of an Indian alphabet, is widely characterized as a conservative system whose spellings preserve historic dis-
tinctions lost in today's standard dialect. Such criticism seems rather lame when read in English, where the fit between sound and spelling is notoriously poor. In reading Tibetan one quickly learns to discount the surplus symbols and eventually to depend on them to show morphological distinctions not apparent in speech. The importance of Tibetan for this study, however, is not the quality of its alphabet, but the fact that it has one at all, in contrast to other languages under Chinese influence, which use some
form of syllabary.
The 28 consonant symbols of the Tibetan alphabet'5 have as their default value a consonant plus the vocalic element [a]. Although the consonants are pronounced with the vowel when named (just as in English), in practice the default [a] is replaced by one of four other basic vowels as
needed, each represented by its own unique symbol. Ko is written with the two symbols k(a) + o, ku is written with the symbols k(a) + u, and so on. Ka itself is written with just the one symbol. One thinks of them as symbols for
pure consonants as one's skill in the language develops. This attitude is reinforced by the use of the symbols as simple consonants (without the implied vowel) when they appear in syllable final position. There is no question that the system is and functions largely as an alphabet.
Although some Tibetan sounds appear in the writing as composite blocks,composed like Korean hangul of one to four letters, this is not the only possible configuration. Root elements consisting of a consonant and a vowel written above or below the consonant with other super- or subscript letters, are often preceded and followed by one or two more letters, in the serial fashion common to alphabets the world over, which serve as phonetic affixes affecting the core pronunciation. To drive the point home, Tibetans add small marks at syllable boundaries (there is no word division), yielding a format one to four spaces long per syllable, unlike the one-to-one correspondence
in other East Asian scripts. In Tibetan, there is a great deal of rigorous--not to say pedantic--phonological analysis forced on the user by the writing, a
fact that coincides with the analytic nature of Tibetan thought.
The second example I want to bring up is the Vai, a Liberian people who invented a syllabary to record their language. Some Vai are literate in the syllabary, others in English, still others in Arabic, while a number remain il-
literate. The Vai were the subject of a famous study conducted two decades ago by Sylvia Scribner and Michael Cole, who sought to test Goody, Havelock, and Vygotsky's hypothesis that literacy affects cognition (see Chapter 6).
Scribner and Cole's finding that "there is no evidence in these data to support the construct of a general 'literacy' phenomenon" (1981:132) had a significant impact on the academic community. Even some scholars identified with what Olson (1994:41) calls the "Toronto school" were motivated
to temper their claims).16
Now one man's goose is another man's gander. Scribner and Cole's study showed that literacy in the Vai syllabary was not enough to produce the cognitive effects claimed for writing in general. As they put it, "Our results furnish little support for speculations that literacy is a precondition or prime cause for an understanding of language as an object" (157). The two found no evidence of the cognitive effects that abstract reflection on language is supposed to produce, neither on individual patterns of thought, nor in the achievements of the group as a whole. In their words,
Quote: Although the Vai invented an original writing system and a social mechanism for transmitting it, Vai society has not gone beyond the kind of restricted literacy described for northern Ghana by Goody (1968) or "craft literacy" discussed by Have-lock (1976).
Vai script literacy does not fulfill the expectations of those social scientists who consider literacy a prime mover in social change. It has not set off a dramatic modernizing sequence; it has not been accompanied by rapid developments in technology, art, and science; it has not led to the growth of new intellectual disciplines. (1981:238)
What Seribner and Cole did show--and this needs to be shouted loudly--is the inability of syllabic writing to generate the same cognitive effects that the Toronto school, working within the alphabetic tradition, had claimed
for literacy generally. Their negative finding, in other words, had nothing to do with alphabetic literacy and everything to do with the point I have been making here and throughout this book about the "semi-oral" nature of syllable-based literacy. To the extent that their study involved alphabetic literacy at all, the researchers found that literacy in English combined with formal schooling in the alphabetic tradition in fact did produce the effects
that Goody, Vygotsky, and others had predicted (1981:132).
Not surprisingly, Seribner and Cole found that advanced Vai literates were better than other groups "in their ability to handle syllables as the unit of communication" (185). The script produced only those cognitive effects it was equipped to facilitate. By the same token, literacy in Vai,
which has no word division, did not enhance its users' ability to identify larger abstract units. The authors reported, "We quickly discovered that there is no lexical item that can be unequivocally identified with the En-
glish word 'word.' . . . [A]ll of the evidence suggests that basic units for Vai writers are meaning-carrying phrases (which sometimes consist of only one word) rather than words in our sense" (143, 149).
Vai users do not distinguish what their orthography fails to discriminate.17 As I stated earlier and will dwell on at some length in the following chapter, East Asians similarly have a vague concept of "word," and until recently had none at all, because their scripts made no provision for them.
Neither the Sino-centric cultures of East Asia nor the Vai of West Africa evidence any substantive ability to handle abstract concepts at either end of the linguistic spectrum, not at the phonemic level, which their writing by-passes, and not at the level of words, which the writing does not identify either.
By the same token, both groups display a marked preference for concrete thought, as shown in the Vai case by the content of their writing. Seribner and Cole observed that "Consistent with Havelock's predictions for writing systems with many representational ambiguities, we found that the content of Vai letters largely concern topics that draw on the recipient's background knowledge of the writer's circumstances. We also found little in the way of original expository text" (240). Olson, in reviewing the
study, also remarked on the concrete focus associated with Vai literacy, "While the non-schooled literate could distinguish things and names for things, the discourse about the relation was problematic" (1994:39).
Although Vai literacy failed to generate the level of cognitive facilitation connected with what Logan has termed the "alphabet effect," Seribner and Cole's conclusion that literacy in the Vai script does not lead to cognitive
differences with nonliterates (1981:258) may be overdrawn. The authors themselves acknowledged "Vai script literates are better conversationalists about formal features of sentences than their nonliterate neighbors" (159). They also found enhanced skills in some linguistic tasks, such as semantic integration, to be a common effect of literacy that cuts across all script types (184). Finally, the two conceded their study may not have been able to measure the full effects of literacy for the simple reason that Vai lit-
erates do not use their written language for many functions.
Quote: If the uses of writing are few, the skills they require are likely to be limited when compared with literacy functions in modern, technologically sophisticated societies. They may be used to accomplish only a narrow range of tasks in a few content domains. Such a pattern can be expected to give rise to specialized or specific
literacy-related skills-the pattern we found in our studies. (1981:258)18
Goody picked up on the same flaw, noting that the cognitive facilitation claimed by Vygotsky and others is "surely still possible if one does not assume immediacy of effect" (1987:225). Olson likewise argued that Seribner
and Cole missed the point by assuming "that the cognitive implications of literacy can be determined by examining the direct impact on the individual who learns to read and write. Learning to read and write is at best an
introduction to the world of literacy." Olson dismissed their argument that schooling-not literacy-produces the cognitive changes that distinguish literate from illiterate societies for not recognizing that schooling is a literate institution (1994:42).
Whether literacy has an immediate impact on cognitive structure or an indirect influence is of less concern to me than what Scribner and Cole did show, namely, that literacy in a syllabic orthography does not produce the cognitive changes that affect people in alphabetic cultures. Their work with this West African people, who owe nothing culturally or linguistically to East Asia, validates the hypothesis made here for syllabic writing generally, just as Tibetan demonstrates the ability of alphabetic writing to produce cognitive effects contrary to those of the wider civilization of which it is a part.
Hemispheric Processing of Chinese Characters
Alphabets differ from earlier forms of writing by the extent to which they encourage abstract thinking. By forcing one to analyze syllables into segments and phonemes, alphabets acclimate individuals and whole societies to the use of analysis as a cognitive tool. Concrete units are split into their abstract antecedents, which is the first step of the creative process. Synthe-
sis, the other part of creativity, is facilitated by the practice one gets grouping linguistic elements into larger wholes. This constant shifting between analysis and synthesis, between the discrete and diffuse patterning associated with consonant and vowel phonemes, and between different levels of linguistic abstraction conditions alphabet users to making these cognitive transitions generally.
Creativity also requires that data be exchanged between the brain's left and right hemispheres. Left brain skills nurtured and refined by alphabetic literacy are used to break problems into their constituents as a precondi-
tion for creative reordering. These abstract "target" patterns become the input for associative processes conducted in the right hemisphere. Provisional solutions assembled in the right hemisphere are evaluated in turn
by left brain analysis. It is the middle stage of creativity, where fragments are completed or recombined, for which the alphabet's contribution has been less well appreciated, although as we have seen its effects here are
equally significant.
Neglect of the alphabet's associative function and its ability to facilitate analogous processes in the brain has been paralleled by an overappreciation of the right hemisphere's role in creativity and by the popular notion
of the right hemisphere as the "seat" of creativity. This naive view, which differs from the integrated, bicameral model of creativity proposed now by specialists, marginalized creativity studies for many scientists because of the right hemisphere's identification with affective aspects of the mind. It seems that any cognitive process that is strange and poorly understood such as creativity, but also intuition and psychic phenomena-is attributed
reflexively to the right brain, where scientific accountability is held to be unnecessary or impossible.
Fascination with the right brain's associative powers coincided during the late 1960s and 1970s with a growth of interest in Eastern mysticism and in Asian civilizations, which were believed to represent an alternative to
the rationalism of the West. In the humanities, this trend was expressed, among other ways, by larger enrollments in East Asian language courses and by an uncritical acceptance of the systems used to write these languages. Chinese orthography, despite everything DeFrancis and other re-
sponsible scholars had written about it, was perceived as a negation of the left-brained culture that the students of that era were struggling against.
This intuition appeared to be confirmed by early psycholinguistic experiments that seemed to show Chinese characters processed in the brain's right half, in contrast to alphabets, which are processed in the left. Tachistoscopic studies conducted by Tzeng et al. for Chinese (1978) and by Hatta for Japanese (1977, 1978) on single-character stimuli showed strong right hemisphere involvement in their processing. This finding contrasted with experiments showing primary left hemisphere support for alphabetic scripts and for Japanese written in kana (Hirata and Osaka 1969; Endo et al. 1978). Jones and Aoki (1988) also found greater involvement of the right
hemisphere in kanji than in kana identification. These studies encouraged the notion that phonetic writing is processed one way, while Chinese characters, like images and other holistic stimuli, are processed in a unique
manner.
Such differential processing, if it did occur, would force us to adjust our thesis that Chinese characters are comparable psycholinguistically to other East Asian syllabaries and amenable to treatment as a single typological group. Moreover, if one can argue that alphabets promote creativity by sharpening left brain analysis, it would seem that an orthography processed
in the right brain would, by the same logic, enhance that hemisphere's own particular contributions. And indeed similar claims have been made.
Fritjof Capra, in The Tao of Physics, subtitled An Exploration of the Parallels between Modern Physics and Eastern Mysticism, wrote:
Quote: The classical Chinese word was very different from an abstract sign representing a clearly delineated concept. It was rather a sound symbol which had strong suggestive powers, bringing to mind an indeterminate complex of pictorial images and emotions. The intention of the speaker was not so much to express an intellectual idea, but rather to affect and influence the listener. Correspondingly, the written character was not just an abstract sign, but was an organic pattern-a 'gestalt' which preserved the full complex of images and the suggestive power of the word.
(1985:103-4)
Chinese characters are regarded here not as a vehicle to express language, but as a way to get around language, as symbols that offer readers a window into the whole essence of a thing by refraining from, well, spelling it out. Needham, who also spared no effort in drawing parallels between quantum physics and Eastern mysticism, likewise was fond of classical Chinese for this same reason, and of the Tao Te Ching in particular, which is undoubtedly the least precise and most widely interpretable text ever written.
Both writers found truth in the basic Taoist premise that analysis robs an object of its essential unity, clouds one's understanding of reality, and that language is especially guilty in this respect. Chinese characters by bypassing language and left brain analysis present a truer picture of reality, in this interpretation.
The reader will recognize this claim-that Chinese writing escapes the confines of language by presenting ideas directly-as an instance of the ideographic myth, which we saw in the previous chapter to be untenable.
The notion, moreover, is getting a bit old. As DeFrancis noted, "From the seventeenth century right down to the present scholars have debated just how to utilize the universal quality attributed to Chinese characters"
(1984:161). But there is no more truth to the myth now than when Peter DuPonceau first criticized it in 1838. Chinese characters represent linguistic elements, namely, syllables and morphemes. Like writing everywhere, the system is inextricably tied to natural language. Although the relationship can be masked depending on how the mapping takes place, allwriting depicts elements of a language and is subject to the same limits as language. Characters are no exception.
The argument for direct, intuitive access to thought through Chinese characters has also been disproved by subsequent studies showing that characters are processed-like all other writing systems-in the brain's left
half. Ovid Tzeng, who helped pioneer hemispheric studies of character processing, eventually became curious about whether the tests with single character stimuli had anything to do with what actually happens in read-
ing. Taking account of the fact that "the perceptual unit in reading may be much larger than single characters," Tzeng redesigned his experiments using two-character stimuli and found a strong left-brain dominance (Tzeng
et al. 1979). Tzeng and Hung later acknowledged "there is very little evidence, from either experimental or clinical studies, to suggest a stronger right hemisphere involvement in the linguistic analysis of Chinese logographs." Rather, the "data provide unequivocal evidence against any suggestion that Chinese logographs are processed in the right hemisphere"(1988:284-85).
Tzeng's finding of LH involvement in character processing based on tachistoscopic experiments with healthy subjects was replicated by studies of Chinese patients with right and left hemisphere lesions. As expected, his team found that patients with right brain lesions were unable to draw pictures properly. However, they "wrote the full and completely correct characters without any linguistic errors." Conversely, those with left brain lesions "produced good copies of geometric figures and well-configured (if simplified) drawings," but did terribly on the characters. Tzeng et al. concluded that Chinese characters are processed in the left hemisphere, like
all other forms of writing, and that "they may be represented in the same way as words transcribed by alphabetic letters" (1986:148-49).
Meanwhile, Hatta working with Japanese substituted two-character units corresponding to words for the one-character stimuli he had used earlier and achieved a complete reversal of effects. The RH dominance mani-
fested when the characters were viewed individually disappeared when they were presented in their linguistic context (1978). Summarizing these Japanese laterality studies, Paradis, et al. noted, "the only pattern that has
consistently emerged across a variety of studies is that two-character word stimuli written with any type of script will produce a RVF [left hemisphere] advantage" (1985:32). They also found "no clinically documented pattern of greater impairment for kanji with right hemisphere lesions" (166).
It's time to put to rest the notion that Chinese characters offer any unique cognitive benefits to compensate for their failure to promote the creative behavior found among alphabet users. The belief that the characters, by not specifying phonetic detail, let readers access ideas without distortion is a myth with no basis in linguistic or psycholinguistic research Chinese characters, like all other forms of writing, are processed in the brain's left hemisphere. As we shall see in the next chapter, not only do
they add little to one's ability to think creatively, but by representing linguistic signs concretely, they obstruct creativity by making it harder to separate concepts from the signs that hold them together. Far from facilitating
right brain association, the characters are instrumental in impeding it.
Syllabic Literacy and the Quasi-Oral Society
The belief that East Asians care more than Westerners for spiritual matters has less to do with "alternative" right-brain thinking than with the failure of the dominant orthography to impart cognitive skills associated with analytical thinking. Lacking an early and ubiquitous model of abstraction, such as that available to literate Westerners through the alphabet, Asians as a whole have acquired fewer characteristics of what Donald (1991:274) calls "theoretic" culture. As such, East Asian thought is heavily tinged with the mythic elements of preliterate society. Syllabic writing does not support a different view of reality, not in East Asia and not anywhere else. It simply fails to express reality to its fullest extent.
As Olson noted, "what the script-as-model does not represent is difficult, perhaps impossible, to bring into consciousness. What is represented tends to be seen as a complete model of what there is" (1994:260). Although syl-
labic literacy does provide-beyond the social and intellectual benefits of literacy itself-a measure of metalinguistic awareness as evidenced in the Vai studies, it stops short of the full complement of abstract skills that created the Western philosophical and scientific traditions. By representing language, syllabic orthography allows its users to think beyond the immediate context of events. That much is indisputable. But by failing to represent it abstractly, the orthography puts severe constraints on the scope of this awareness.
The marginal nature of Chinese character-based orthography is evident at each linguistic level. Speech sounds are represented concretely. They are not analyzed into segments or abstract phones. Morphemes--basic elements of meaning--are depicted instead of words that correspond to complete concepts. The lack of clearly defined words leads to illdefined sentences, not only in classical texts, which are often nebulous and subject to varying interpretations, but even in modern scientific writing. Sentences
run on interminably, much as in speech, with little thought given to segmenting ideas. The style is often vague, as if East Asian writers do not appreciate the need to present their work as an objectified, self-sustaining whole independent of personal context.19
Such is the legacy of syllabic writing. It supports literacy, but with few of the supplementary benefits that are associated with literacy in the West. East Asian orthography thus stands midway between pre-and fully devel
oped writing, not only in an evolutionary sense but also in terms of its impact on the cognitive habits of its users. The strength of this link between orthographic type and culture can be appreciated by citing once again the
features that Donald, writing within an alphabetic tradition, associates with "theoretic cultures," namely,
Quote: formal arguments, systematic taxonomies, induction, deduction, verification, differentiation, quantification, idealization, and formal methods of measurement. Arguments, discovery, proof, and theoretical synthesis are part of the legacy of this kind of thought. (1991:273-74)
How many of these characteristics describe the kind of thinking done in East Asia traditionally, or for that matter today? On the contrary, they represent precisely that set of cognitive features that both Needham and Lo-
gan agree are missing in China, and which Nakamura says are absent from East Asia entirely. Qian Wen-yuan, the physicist and intellectual historian, reached the same conclusion. Here is what Qian had to say about Chinese
scientific thought:
Quote: we can seldom find the dimmest consciousness of scientific axiomatisation ... the great deficiency in old Chinese mathematical thought was the absence of rigorous
proof, in particular, the absence of a system of deductive geometry. This configuration correlates with the lack of formal logic and the dominance of associative (organicist) thought. From our Sino-European comparison, it is clear that the deficiency was notjust in mathematics; it hindered the development of modern science as a whole. In other fields of science, the Chinese way of thinking generally lacked accuracy in defining, exactness in formulating, rigor in proving, and logic in explaining. (1985:66-67)
It is evident that we are looking at two distinct types of cultures, whose intellectual foundations are qualitatively different. If writing, as Donald and many have argued, was responsible for the changes in thinking that attended the shift from mythic to theoretic culture, then consideration must be given to the type of writing involved in the process, since the cognitive effects of this shift have not been uniform. Neither the mystical orientation of East Asian thought represented by Taoism and Zen (Chan) Buddhism,
nor the concrete practical strain of East Asian philosophy expressed by Confucianism, have much in common with the theoretic culture created in the West by alphabetic literacy. On the other hand, both such tendencies--
the mystical and the pragmatic-are characteristic of oral cultures and the preliterate society Donald describes.20
East Asia's syllabic writing could not lay the groundwork for a complete transition to the theoretic culture on which modern science depends. In particular, it has not supported the cognitive skills needed for scientific creativity, neither in its analytic nor its synthetic phases. I have described some of the technical reasons why this is so. In the next chapter we will consider another such dimension to the problem.
作者：Anonymous 在 罕见奇谈 发贴, 来自 http://www.hjclub.org |
|
|
| 返回顶端 |
|
 |
| |
|
|
您不能在本论坛发表新主题
您不能在本论坛回复主题
您不能在本论坛编辑自己的文章
您不能在本论坛删除自己的文章
您不能在本论坛发表投票
您不能在这个论坛添加附件
您不能在这个论坛下载文件
|
based on phpbb, All rights reserved.
|
