When the first edition of Sampson’s book was published in the 1980s, there were relatively few books, or other publications for that matter, acknowledging the written language to be a valid object for serious linguistic analysis. Sampson’s book was one of the first publications in modern linguistics to challenge this consensus and also suggested that the ‘tide was beginning to turn’ in favour of this view. The second edition to some degree reflects this turn of the tide, in its many new references to relevant work that has been published in the meantime (e.g. Daniels & Bright Reference Daniels and Bright1996, Cahill & Rice Reference Cahill and Rice2014). Sampson mentions the emergence of journals in the field as further manifestation of the trend, and he could also have mentioned the presence of chapters on written language in newer introductory linguistics textbooks (e.g. O’Grady, Dobrovolsky & Katamba Reference OˈGrady, Dobrovolsky and Katamba1997). However, Sampson’s claims of nearly twenty years ago seem to have been a little strong in that many theoretical linguists still adhere to the view that the spoken language is the be-all and end-all of linguistic study. If the first edition of the book did not quite manage to reverse this trend, the second edition may still be influential, even if it appears now in a rather different intellectual context.

One side effect of the increase in interest in writing systems is that not all of Sampson’s assumptions and distinctions may meet with the agreement of experts in the field. His insistence on using the terms ‘script’, ‘writing system’ and ‘orthography’ interchangeably would not be acceptable to many working in the field, although he provides good justification for this position (8), and in several places, especially in the ‘Introduction’ chapter, makes clear that he sees particular descriptive decisions not as a case of right or wrong, but as a matter of choice. Examples of this include his choice of how to represent phonetic/phonemic forms (12), and the use of the terms graph, grapheme and allograph.

The structure of Sampson’s book (in both editions) broadly reflects the typological aspects of writing systems and includes reflections on the history and psychology of the systems as he goes along. That is, the chapters mostly focus on specific types or families of scripts (predominantly using a single example as a case study to illustrate in detail the principal characteristics of that family) but within these chapters, questions of the history of those systems and their practical pros and cons are discussed at length.

The differences between the first and second editions fall broadly into three categories. First, the overall writing style and presentation are more reader-friendly and more appropriate for the likely audience in the twenty-first century. Secondly, there are significant revisions in the form of new chapters as well as additions to and deletions from the first edition’s chapters. The second edition is around fifty pages longer than the first and has four additional chapters (although two of these are to some degree the result of splitting a chapter from the original into two). Finally, there are revisions to some of the arguments put forth in the first edition; some of these correct notions that recent evidence has weakened, others tone down the sometimes combative tone of the earlier edition.

With respect to presentation, there are some essentially cosmetic changes, such as the move from notes at the end of the book to notes at the end of individual chapters and, interestingly, given some of the discussion, the book is printed in a different font. There are also changes that reflect the received view on how to represent certain non-English names, such as Hangul (formerly rendered as Han’gŭl) and Ju Si-gyeong (formerly Chu Si-gyŏng). The division into paragraphs and sections (separated by lines of asterisks) all seem designed to make the book easier for the modern reader. Sampson’s writing style in the original, while immensely readable, was in places very old-school academic. This has been adjusted throughout so that, for example, ‘In East Asia matters were quite otherwise’ becomes ‘In East Asia it was different’. Such changes make a big difference to the readability and the overall impression and are likely to make the second edition even more accessible.

The additional chapters include a conclusions chapter, one on information technology and two that are expansions of material covered within other chapters in the first edition. The chapter on ‘European alphabetic writing’ (called ‘The Graeco-Roman alphabet’ in the first edition) has been refocused to more clearly explain the justification for the claim that the Greek, Cyrillic and Roman scripts should be considered as typologically equivalent. That this assumption was taken as read in the first edition is reflected in the chapter heading – it will undoubtedly have come as a surprise to many to see the Greek and Roman alphabets referred to as a single alphabet. In this regard, it is unfortunate that the errors highlighted in the review of the first edition in this journal (Collinge Reference Collinge1987) have not been corrected. The typological justification is supplemented by the historical developments of these scripts, with Semitic writing presented as the historical forerunner of the others, and only distinct from them in lacking vowels. Some of the material from this chapter of the first edition has been separated into a new chapter, ‘Influences on graph-shape evolution’, which succeeds not only in explaining the development of the European alphabets in terms of the linguistic requirements of the scripts for particular languages, such as the distinct phonotactics and phoneme inventories, but also in presenting some fascinating explanations of more prosaic reasons for choices of specific script features. The practicalities of writing with a stylus (sharp or blunt) on clay or of writing on relatively fragile leaves with veins going in all directions are used to explain the development of the Cuneiform and Devangari scripts.

An interesting change of view is demonstrated in the discussion of the role of aesthetic considerations in the evaluation of different typefaces. In the first edition, Sampson bemoans the lack of scientific assessment of the readability of different fonts, and the consequent reliance on aesthetic judgements. In the second edition, he reports on the (quite extensive) research that has been done in this area since the 1980s, observing that there is still no scientific consensus and concludes: ‘Considering the large part played in our lives nowadays by the printed word, it may be welcome news that no better guide than beauty is available’ (141).

The final additional chapter discusses the pros and cons of logographic scripts. This is, in fact, largely material that was in the chapter on Chinese writing in the first edition, with the equivalent chapter in the new edition expanded, chiefly to cover more of the history. This leads to an interesting discussion of the fact that certain scripts or types of script may be more suitable for certain languages, or types of language.

There is a nice touch in the chapter on the Korean script, where the sample he provides of Hangul running text has been replaced with a passage from the Korean translation of the first edition of this book. This, as other chapters, has been largely left intact in terms of content, although a large part of the section on consonants has been removed, with further detail about the writing of words of Chinese heritage added. This chapter also hosts a discussion of the question of deep/shallow orthographies and the debate around the SPE claim (Chomsky & Halle Reference Chomsky and Halle1968) that writing should represent (some form of) underlying rather than surface representation of a language. This is revisited in the chapter on ‘English spelling’. In this edition, as in the first, Sampson makes an unabashed defence of the English spelling system, although without resorting to full agreement with Chomsky & Halle. There is also an updated discussion of the attempts at spelling reform of English.

Sampson is assiduous in relying on scientific evidence wherever possible, and this comes partly in the form of psycholinguistic experiments that test the reading skills (speed and accuracy) of both adult and child readers of different languages. There has been a significant quantity of such research in the intervening years since the first edition was published, and Sampson refers to this in his argument about the acceptability (if not desirability) of the English spelling system. He has thankfully removed his somewhat offensive suggestion that adult illiterates are probably illiterate because of their own laziness, but makes a cogent argument that the evidence suggests that, once the initial stage of acquiring the system has been passed, the English spelling system is actually very efficient for typical adult readers, who make up the majority of the ‘users’ of written language. He also makes an important point about the relative importance of ease of writing versus ease of reading. As he points out, each written document has typically a single writer but potentially many more readers. Since the invention of the printing press, this imbalance has only increased, and in the present day, with the use of the internet, that increase has become probably exponential. Therefore, it makes sense that the primary consideration in choosing a script is the needs of the reader. This point is made in relation to spelling conventions but also in the discussion of font styles. Sampson notes that thirty years ago, when the first edition was written, few people had any knowledge of different fonts. Now word processing packages allow us to choose from hundreds if not thousands of different fonts.

The additional chapter on ‘Information technology’ addresses some of the questions about the influence of technology on the way we write. Sampson is convinced – like Crystal (Reference Crystal2009) – that young people will be able to distinguish the appropriate use of different writing styles, just as they learn to use different registers of spoken language. However, most of the chapter is taken up with a lengthy explanation of the Unicode system and a discussion of the relationship between the linguistic and technological (practical) considerations involved in its development. The level of detail in explaining the assignment of numbers to distinct characters is probably more than most linguists will be interested in and it is perhaps unfortunate that a shorter, less technical, explanation was not provided in order to allow explanation of the failure of Unicode to properly reflect significant linguistic differences relating to graphs and graphemes. Sampson is largely sympathetic to the decisions forced on the developers of Unicode, although his claim that ‘the problem of representing written documents electronically, it is safe to say, has now been definitively solved’ (247) may be a little strong, given that in the current versions there are still significant problems with combining left-to-right and right-to-left scripts; for example, Alqadah (Reference Alqadah2012) explored the rendering of documents prepared in HTML code containing Arabic writing in a range of different editors and found significant inconsistency.

Sampson maintains his stance that semasiological systems are not properly considered in writing, according to his own definition: in Chapter 2, ‘Theoretical preliminaries’, writing is defined as ‘a system for representing utterances of a spoken language by means of permanent, visible marks’ (18–19). He illustrates this point using the example of Blissymbolics, replacing the example of illustrated instructions used in the first edition. The second edition also loses the Yukaghir ‘picture writing’ example (replaced with washing symbols), which is a shame, as this was an interesting and unusual example, but perhaps justified precisely because of its rarity. Most readers of the book will be able to relate to the washing symbol example much more easily.

This book remains one of the most interesting, useful and balanced works on the typology, history and psychology of written language, and the revisions made for the second edition bring it nicely up to date, at a time when the linguistic study of written language, if not quite as widely accepted as Sampson expected at the time of writing the first edition, is nevertheless much more widely acknowledged and undoubtedly gaining ground.