LINGUIST List 26.5483

Wed Dec 09 2015

Review: Writing Systems: Sampson (2015)

Editor for this issue: Sara Couture <saralinguistlist.org>


Date: 14-Aug-2015
From: Michael Cahill < mike_cahillsil.org">Mike_Cahillsil.org; mike_cahillsil.org>
Subject: Writing Systems
E-mail this message to a friend

Discuss this message


Book announced at http://linguistlist.org/issues/26/26-698.html

AUTHOR: Geoffrey Sampson
TITLE: Writing Systems
PUBLISHER: Equinox Publishing Ltd
YEAR: 2015

REVIEWER: Michael C. Cahill, SIL International

Reviews Editor: Helen Aristar-Dry

SUMMARY

This volume is not an encyclopedia of all the known writing systems used in the world’s languages. Its more modest but more achievable goal is to give the reader a typological view of the ways languages throughout the centuries have represented ideas and sounds in writing. Sampson also succeeds in dispelling a few popular myths about writing systems along the way.

Chapter 1 (“Introduction”) lays out Sampson’s purposes and broad conceptual organization, discussing the core concepts of typology of scripts, their historical developments, and their “psychology” – how people actually process orthographies. He also describes the transcription systems he uses.

Chapter 2 (“Theoretical preliminaries”) discusses such questions as “Are all written languages parasitic upon spoken languages?” and what main types of writing systems exist. The first main split is between logographic (writing that represents morphemes) and phonographic (writing that represents sounds). He distinguishes complete and incomplete writing systems, and discusses the more linguistic “deep” and “shallow” orthography distinction.

Chapter 3 (“The earliest writing”) gives an overview of cuneiform, first used by the Sumerians and then adapted by the Akkadians and others. Sampson notes that some of the patterns presented here appear later with other writing systems, so this particular system is examined in detail. Common issues include the physical limitations of the medium (here, stylus on clay), but pertinent to future chapters is the variety of ways symbols can be used. Some symbols represented concepts, but some of the symbols originally represented concepts, but were also used to represent words that sounded similar. Eventually there was a shift to symbols that represented sounds only. Also noteworthy is that the shapes of the symbols themselves shifted over the centuries (a natural development prior to the printing press). They generally became simpler, but also they were rotated 90 degrees early in their development. Besides the graphic considerations, Sampson also notes the actual usages of earliest writing, most generally for keeping records; complete sentences were basically nonexistent. As he notes in Chapter 4, cuneiform was “a complex system which rather untidily combined various orthographic principles in different proportions at different stages of its history.”

Another useful contribution of this chapter is debunking a theory of the origin of writing that has been widely accepted. Sumerians enclosed small figures of clay in clay envelopes called “bullae”. The bullae themselves had symbols on them, and the idea is that the symbols on the bullae represented the tokens inside, and these symbols then developed into writing, independent of the existence of the inner tokens. Sampson cites recent criticism of this theory that makes it untenable. We are left with no good theory about how specifically writing originated. Due to decipherment of the Mayan glyphs (unrelated to the Middle East), we can, however, affirm that writing has had more than one origin.

Chapter 4 (“A syllabic system: Linear B”) is a brief examination of the early Mycenaean orthography so painstakingly deciphered by Ventris and Chadwick and found to be archaic Greek. Linear B is a syllabary, and Sampson explicitly points out the characteristics of a true syllabary. Each symbol is independent in shape from other symbols, even if the sounds are related. This contrasts with Ethiopic script, which is often called labeled a syllabary. In Ethiopic, each symbol represents a syllable, but the symbols for /ta, te/ have a common basic shape, but modified for the differing vowel. Linear B did not represent all the phonemic contrasts found in the language, and so was “incomplete,” in Sampson’s terms, exhibiting what others would term “underrepresentation.” For example, the aspirated/unaspirated contrast was not represented. This is probably because Linear B was not initially developed for Mycenaean Greek, but for another language, possibly Luwian, and the Mycenaeans adapted it for their language. Sampson goes into detail about how the orthography actually worked, e.g. how a consonant cluster was represented when each symbol represented CV. All findings of Linear B to date have been for administrative or business records, leaving open the question of whether people actually used it for other purposes.

Chapter 5 (“Consonantal writing”) is the longest chapter, and covers three major topics: how an early Semitic language (represented by biblical Hebrew) operated, how the Devanagari script of India compares, and how Semitic modification of Egyptian hieroglyphics likely led to the first alphabetic writing system. For biblical Hebrew (hereafter “Hebrew” for conciseness), Sampson presents early and more fully developed forms of the letters. He discusses the sound system and phonology of Hebrew. For example, a single consonantal phoneme could be pronounced as a single stop, a geminated stop, or a fricative, but these variant pronunciations were all represented by a single grapheme. This representation means that early Semitic writing was the first truly alphabetic system, where one symbol represents one phoneme. Sampson details two possible reasons why early Semitic writing represents only consonants, but not vowels. First, there was considerable variation in vowel pronunciation, as in the first vowel of English “meter” and “metric”. Second, different grammatical constructions have different vowels, and again, these distinctions do not correlate with different lexemes. Sampson discusses the disadvantages of totally neglecting vowels, and discusses the use of consonants sometimes doing double duty as vowels, as well as the better-known “pointing” system to indicate vowels in children’s books and other contexts.

It appears that Indians borrowed their earliest alphabet from the Semitic language Aramaic. Though modern Devanagari script bears no resemblance to Semitic, the earliest forms of Devanagari do. Sampson uses Sanskrit as a representative language. The basic consonantal graphemes are marked with various modifications to indicate which vowel is wanted, and true consonant clusters combine the symbols with a ligature.

To conclude the chapter, Sampson raises the question of why the first alphabetic system arose in Palestine and not elsewhere. Palestine is located between two locations where possible predecessors existed: between the Fertile Crescent where cuneiform arose, and Egypt, the home of hieroglyphics. Until recently, there was no evidence to decide between the two. However, in the 1990s, inscriptions in the western desert of Egypt were discovered, and these had clear links to both early Semitic writing and Egyptian hieroglyphs. As Sampson writes:

“The discoverers suggest that the inscriptions could represent a script created by Semitic-speaking immigrants in Egypt (there are known to have been many of these), who grossly simplified the Hieroglyphic system in order to adapt it to their own languages, and who then took this simple script with them when some of them returned to their Palestinian homeland.” (p.100)

Chapter 6 (“European alphabetic writing”) begins by asserting the underlying unity of European alphabets which appear to have different scripts (e.g. Roman, Greek, Cyrillic). Many of the shapes of letters are identical, and they all have separate graphemes for consonants and vowels. Greek, the earliest, derived from some version of Semitic, with modifications in which sound was represented by what symbol (e.g. Semitic ejective ‘t’ was used for Greek aspirated ‘t’). A crucial innovation was the adaptation of six Semitic symbols for use as Greek vowels, since vowels were crucial in lexical contrasts in Greek. Sampson details, symbol by symbol, how this transition from the Semitic system to the Greek system took place, and how the right-to-left Semitic writing changed to left-to-right writing, via the “boustrophedon” system which alternated directions. Unlike Linear B, the Greek alphabet was used from the beginning for literary purposes. Sampson notes that Greek orthography represented more of a surface pronunciation than a morphological structure, as modern English tends to do. Sampson briefly presents the origin of Cyrillic, and ends the chapter by describing how the Greek alphabet was adapted by the Romans through Etruscan, giving us our present Latin alphabet.

Chapter 7 (“Influences on graph-shape evolution”) concentrates on three factors that have impacted the forms of scripts: writing materials, ideology, and the need for distinctive shapes. Runic alphabets, with their exclusive straight but angular lines, were written mostly on wood, a material less conducive to curves. Sampson also notes that the long-held stories of runes being associated with pagan religions and magic are nonsense; runes were commonly used by Christians. In contrast to runes, the curvy alphabets of southeast Asia were commonly executed on palm leaves; any straight line would likely coincide with a vein and split the leaf. Regarding ideology, nationalistic fervor is shown to be a huge factor in use of the Irish script, first used throughout the British Isles, but later relegated to Ireland. It became a symbol of Irish nationalism, with the Irish rejecting attempts by Queen Elizabeth I, for example, to use this script for her own politico-religious purposes, and for some time, the Irish resisted all attempts to print Irish language materials in a roman script. Similarly, the German “Fraktur” script was highly identified with German nationalism for a period of time. The last factor – distinctiveness of shapes – has proven to be less powerful than might be imagined – in Semitic alphabets. Several Hebrew letters are quite similar in shape, and several Arabic letters actually merged in shape historically, e.g. < z > with < r >, < p > with < k’ >, etc. (The sounds were later distinguished by a system of dots over the base shape.) The influence of distinctiveness is more important in European languages, as the history of upper and lower case and italic styles shows. Sampson finishes this chapter with a discussion of fonts, which computer-age people are much more familiar with than readers in past centuries. Characteristics of families of fonts are discussed, including serifs and line strength. Scientific research on font readability has not favored one type as being inherently more readable than another.

Chapter 8 (“A featural system: Korean Hangul”) focuses on what has been called the best alphabet in the world, demonstrating that symbols incorporate linguistic information, so that sounds made in the same place in the mouth have similar shapes, all stops have a horizontal line, etc. From a linguistic and orthographic point of view, one interesting facet of Hangul is that it manifests a “deep” orthography, where the orthography represents not a surface phonetic form, but a deeper underlying level. In terms of orthography change, King Sejong’s original Hangul script of the 15th century seemed to be intended as a shallow, more phonetic representation. Centuries of language changes made the orthography more difficult to use consistently, and in 1933 a major spelling reform was instituted. In this, the spelling deliberately represents an attempt at writing underlying forms, not surface ones.

A key connection to linguistic theory is also addressed here. Sampson argues that the historical move from a shallow to a deep orthography, as happened in both Korean and Greek, may call into question the idea of generative phonologists of how speakers store vocabulary in their memory. If speakers store a deep underlying form, then that deep form, one would think, would be the natural one to access for an orthography, even at the initial stages of developing it. But we find the reverse.

Sampson notes that Hangul is not an ideal system, having low distinctiveness in its characters, for example. He also notes some recent efforts toward using Hangul as the basis for writing previously unwritten languages, and questions the wisdom of that, largely for sociolinguistic reasons. Still, he calls Hangul “one of the greatest intellectual achievements of Mankind.”

Chapter 9 (“A logographic system: Chinese writing”) starts with the startling but plausible claim that until the 19th century, more than half of all books ever published were written in Chinese. The basic idea of a logographic system such as Chinese is that there is a separate character for every morpheme. Chinese has much homophony, e.g. the words for “parboil” and “leap” are pronounced identically. But their graphemes are quite distinct, and thus the written form clearly disambiguates them. The grapheme IS a morpheme. Sampson intriguingly sketches the origins and development of the symbols, showing how early characters were often shaped like their referent (though some of the connections are opaque to modern eyes). More complex characters were developed by combining two or more existing characters. Thus “military” was composed of the strokes for “dagger” and “foot,” indicating arms and marching. This can be looked on as a semantic-semantic compound character. Another development took advantage of the large numbers of homonyms or near-homonyms. Thus the character for *pek “prince” was also used for *pek “thin-sliced,” *bek “law,” and several others. This obviously left much ambiguity in the system, and thus “significs,” additional characters that fleshed out the semantics of the word, were added to aid in identification of the real meaning. The result was a phonetic-semantic grapheme, and most present Chinese graphemes are of this type. The system was essentially fixed by the dawn of the Christian era. However, since some graphs have become obsolete, and pronunciation has changed through the centuries, a contemporary Chinese reader is not likely to be able to give the meaning of each element of a character. Each character must be memorized as a unit. Sampson illustrates these with an analysis of ten random Chinese graphs.

Chapter 10 (“Pros and cons of logography”) is largely a defense against the common Western idea that Chinese is so cumbersome that it is inferior to an alphabetic orthography. Sampson briefly discusses the introduction of the pinyin alphabetic transcription system (not an official writing system) and the simplified characters (which Sampson criticizes on the grounds of reduced visual distinctiveness). Sampson argues that learning Chinese vs. learning an alphabetic system involves two unrelated types of difficulty: the difficulty of memorizing thousands of graphs (Chinese) vs. analyzing every written word into sounds to recognize the word (alphabetic). Sampson mentions two other drawbacks of Chinese. First is the immense difficulty in typewriting, which has been largely alleviated by computer systems. Second, and even less recognized by outsiders, is the difficulty in representing foreign names, since all the graphemes of Chinese represent Chinese syllables. Chinese has largely dealt with this, not by adapting a foreign word into its sound system, as other languages commonly do, but by coining a new compound word (e.g. “computer” is “electric brain,” or more fully, “electron calculate machine”). Proper names are represented by sequences of phonetically-based graphs, with clumsy results. But Sampson maintains that these drawbacks are more than outweighed by two advantages of Chinese. First, sound changes in the history of Chinese have created an incredible number of homophones. For example, the words/morphemes for “cheat, period, mountainous, creek, seven, varnish, to mash tea, wife, grieved, roost, kinsman” all had distinct pronunciations in Old Chinese, but are all identical today! In conversations, meaning is negotiated, and people can immediately clear up misunderstandings. However, in writing, there is no negotiation possibility, and a separate symbol to represent each word is an enormous advantage. The second plus for Chinese writing is that since it is based on morphemes and not pronunciation, it can be understood by different dialects across China and has been unifying for the country. Sampson’s point is that logographic writing is not obviously inferior to alphabetic writing.

Chapter 11 (“A mixed system: Japanese writing”) starts by labeling Japanese as the most complex writing system that exists, and after a few pages, one can appreciate this evaluation. Much of this is traced back to the Japanese aristocracy who developed Japanese writing. They had much leisure time and were more interested in the intellectual richness of the system rather than prosaic functionality. One huge factor is mismatches between the Japanese language and Chinese, from which much of the writing system comes. Some graphemes were borrowed with the Chinese meanings intact, but pronounced as Japanese, e.g. the graph for “man” is the same in both languages, but Japanese say /hito/, not the Chinese /rən3/. These are called “kun” readings. The second type of writing is “manyogana (kana)” writing, in which each syllable of a polysyllabic Japanese word was represented by a Chinese grapheme that had about the same sound, no matter if the semantics were related at all. For example, the three syllables of Japanese /fukushi/ “trowel” were represented by graphs which have the meaning “cloth-long.time-thought.” It was not always obvious whether the reader should use a “kun” or “manyogana” reading. Furthermore, Japanese borrowed words from Chinese, along with their graphemes. Such graphemes may be read as either the native Japanese word or as a Chinese loan word. The latter reading is called “on” and is distinct from the “kun” above. Sampson goes into bewildering detail on the single Japanese word “kimono,” showing all the history and possibilities for interpreting the written form. Resolving an ambiguity of the written form often must come down to one’s knowledge of Japanese vocabulary. Homophony in Japanese writing is greater than in Chinese; often distinct Chinese pronunciations merged into one in Japanese (Sampson gives one list of 13 Chinese characters which are all pronounced the same in Japanese.) To add yet more complexity, Chinese characters were borrowed in three waves historically, and with each wave, the pronunciation of the character in Chinese could be different. The three waves of “on” mean that a goodly number of Japanese characters have three distinct pronunciations and meanings.

The above only treats Chinese characters used in their full form (“kanji”). After some time, Japanese developed two simplified versions of some Chinese characters (“hiragana” and “katakana”), and these were assigned phonetic values as a syllabary, jointly called “kana” symbols. These are often used today to spell out function words, borrowed words, and foreign names. These factors do not exhaust the complexities of Japanese writing; Sampson discusses a few others. He observes that Japanese writing shows “just how cumbersome a script can be and still serve in practice,” noting the very high literacy rate of Japan.

Chapter 12 (“Writing systems and information technology”) is a new chapter for this edition, and begins with Sampson’s discussion of the younger generation’s innovations in texting, such as < c u l8er > for “see you later,” which some regard with “moral panic,” but Sampson is not very worried about. The bulk of the chapter discusses the history of the way computers have handled writing systems. In early computers, memory was expensive, and the 64 possible characters allowed did not even allow capital and lower-case distinctions. Later hardware handled larger characters sets. A huge problem in pre-internet days was that designers were not consistent in how to code the same symbol; thus documents passed between computers often came out garbled. The Unicode standard directly addressed that, with a standard code point to represent a specific grapheme. Interestingly, Sampson points out that Unicode operates on different principles than ordinary phonology, in terms of phonemes and graphemes. One example is the script < ɑ >. This is a variant of < a > in normal English writing, so < ɑ > and < a > would have the same Unicode code point for English, the actual shape depending on the font used. However, in the International Phonetic Alphabet, < ɑ > and < a > represent two distinct sounds, and therefore need two code points.

Chapter 13 (“English spelling”) raises the question “Why not reform English spelling?” Other European languages have done it, including Germany as recently as 1995. But there is no serious movement to do so. Sampson maintains that the logistical difficulties are easily surmountable; what is lacking is the will to do it. Why is English spelling so difficult? English was spelled fairly phonemically a millennium ago, and some of those spellings were frozen while the language changed. But more importantly in Sampson’s view is the Norman conquest. For three centuries, English was not used for official purposes, and so standardization was degraded. Furthermore, the official French had not been standardized at that point either. So spelling conventions used a mix of various English conventions and various French conventions; all were in use, and often the same word would have different spellings, depending on which convention was used. Also, printers often came from the Netherlands, and introduced spellings based on Dutch, as in the < gh > of “ghost.” By 1650, in spite of all this, English spelling had pretty much stabilized. Sampson next examines some defenses of English orthography. It is more of a deep than a surface orthography, but the early generative phonologists’ extreme ideas, such as that the < gh > in words like “night” reflect an underlying /x/, are not tenable. English spelling has moved into a deep phase, and even flirts with being on the verge of logographic, since each word has a more constant image than it would in a regularized phonemic orthography. This more constant visual image is actually an advantage of reading English, though he acknowledges that learners have a rougher time with this than they would a more phonemically based system. This system is biased toward the reader rather than the writer; a writer finds it easier to spell with a regular phonemic system. And thus the historical move of English toward a constant-image morpheme has favored the reader, which may have its own social consequences.

In Chapter 14 (“Conclusion”) Sampson briefly summarizes three themes of the book: that scripts (logographic and phonemic) are diverse, that since languages are diverse, some scripts fit them better than other types, and that the history of writing does not support the idea that an ideal orthography represents speech sounds perfectly. Thus English spelling may not be so bad after all.


EVALUATION

This is the second edition of the book, and some additions are obvious. Over 110 new (that is, more recent than 1985) references have been added, and more recent research has been incorporated into the text. As Sampson himself notes, concepts like information technology and references to computers were totally lacking in the first edition of 1985. Chapters 7, 10, and 12 are new. The book is well-edited, with very few typos (though I can’t personally vouch for the Linear B, runes, Chinese, and other non-Roman scripts!).

One of the challenges in discussing writing systems is what term to use for the characters in a system (symbol? character? letter?). Sampson spends some time in defending the use of “graph” in the first chapter, not “grapheme,” but then “grapheme” does pop up occasionally later on. Much of Chapter 2 may not be strictly necessary to benefit from the succeeding chapters, but as the chapter progressed, I saw the value of setting out explicitly the definitional and philosophical issues, and his assumptions and framework.

The book is quite readable, is clearly written, and the bulk of it is a pleasure to read. The reading becomes more difficult in some chapters when Sampson presents very detailed examples of particular languages and their systems; it is easy to get lost in the particulars of many words in an individual language. Nevertheless, this is rather a matter of concentration, more than of specialized knowledge. The reader does need a basic knowledge of phonology to follow the totality of the book, and what a morpheme is, but not more than that. Thus the potential audience is fairly broad. Even those with no linguistic background could benefit to some degree.

This book, as mentioned, does not aim at a comprehensive listing of the world’s scripts and orthographies. Thus it is only a third as long as the well-known Daniels and Bright (1996) volume “The World’s Writing Systems.” For those interested in yet more examples of writing systems, there is now ScriptSource (www.scriptsource.org), described as a “dynamic, collaborative reference to the writing systems of the world, with detailed information on scripts, characters, languages - and the remaining needs for supporting them in the computing realm.” It currently (August 14, 2015) contains 140 examples of scripts, and more contributions are invited.

By choosing key examples and representative orthographies, this volume quite nicely achieves its main goal of providing a typology of the main writing systems of the world’s languages.

REFERENCES

Daniels, Peter, and William Bright (eds.) 1996. The World’s Writing Systems. New York: Oxford University Press.


ABOUT THE REVIEWER

Michael Cahill (Ph.D. Ohio State University) was involved with the Konni Language Project in Ghana for several years before serving as SIL's International Linguistics Coordinator. He was recently appointed as the first Orthography Services Coordinator of SIL. His main linguistic interests are phonological, especially tone systems and the phonology, phonetics, and historical development of labial-velar obstruents, as well as African languages in general.


Page Updated: 09-Dec-2015