Most helpful critical review
10 of 11 people found the following review helpful
Badly Written on Several Levels
on 8 July 2013
Nicholas Ostler plainly knows a lot about languages. Unfortunately he is not very good at writing a book. To support that judgement here is a selection of authorial shortcomings.
Item 1. The book takes a long time to get going. There is a Preface, then a Prologue, then a Chapter 1, then Chapter 2, and Chapter 2 is still telling us what the book is going to be about - eventually. Four sections of a preludial character before the book actually gets into its stride is a ridiculous way to test the patience of the reader. The author or his editor should have spotted that poor piece of design.
Item 2. The book includes a fair number of textual extracts, such as an exchange of salutations in Maya (transliteration and English translation, p2), a Sumerian lovesong and lullaby (transliteration and English translation, p51-52), a traditional maxim in Sanskrit (actual characters, transliteration and English translation, p174), a poem in Greek (actual characters and English translation, p265) and many others. This feature of the book may have seemed a nice idea early at the early brainstorming stage, but it should not have survived very long. A sensible author would have realised long before publication that these extracts would add nothing whatsoever to a reader's intellectual grasp of the book's subject.
Item 3. There is an account (p297) of the constitution of ancient Rome (consuls and praetors and so on). This extends for almost two pages and I began to wonder how the author would make this information relevant to the topic of the Latin language. He doesn't even try. If he had looked at his text critically he would have seen that the passage could be excised completely with no loss whatsoever.
Item 4. The chapter on Sanskrit begins (p174) `The word Sanskrit ... is a term for the language as formulated in the grammar books, contrasting it with its colloquial dialects, known as the Prakrits.' After reading the whole chapter twice I think I know what the author means by that clumsy sentence; ie something like this There was a language, Sanskrit, that had numerous dialects (which were called Prakrits). In everyday life people normally spoke one of those dialects. A more refined form of the language also existed.
With that made clear the reader will want to know: So when was the refined form of the language used? Soon the author gives an indication, albeit rather clumsily. He starts a paragraph: `Although it is religious tradition which has proved the most reliable preserver of Sanskrit .... .' It would have been much clearer if he'd written a sentence along the lines of: `The refined form of the language was used as follows:...'
From the above it follows that the term `Sanskrit' has two different senses: Sanskrit1 (my terminology), a whole family of related dialects, used in everyday speech; Sanskrit2, a certain cultured language, not normally used in everyday speech. It is pretty vital to grasp this distinction, but the author doesn't make it explicitly. Consequently the alert reader has to waste energy by constantly judging which of the two the author means when he refers to `Sanskrit'. In the text: `..by the beginning of the fifth century BC the language was spoken in area extending as far west as Bihar.. (p176) that is presumably Sanskrit1. ' When he says: ` .. in these inscriptions .. Sanskrit is used for the verse, Prakrit for the prose ...' (p188) that must be Sanskrit2. When he says: ` .. major Sanskrit-speaking states were set up all other South-East Asia ... (p204)', um, that is a puzzle. From the context I'm pretty sure he means Sanskrit2, but it isn't at all clear by whom or in what circumstances Sanskrit2 was spoken in South-East Asia.
In fact one simple question that will occur to the thoughtful reader remains unconsidered throughout this 50-page chapter: outside the context of religious ceremony and the recitation of literature to what extent was Sanskrit2 a spoken language?
The whole chapter is just poorly written. A skilled writer would have recognised the key distinctions and issues to get across in this chapter, and would have made sure that happened.
Item 5. The author says a good deal about Latin and just occasionally slips in a mention of `vulgar Latin'. There seems to be no definition or explanation of `vulgar Latin'. (I can't say for sure since the index doesn't contain `vulgar Latin'). If you as an author are going to use two different but cognate concepts such as `Latin' and `vulgar Latin' it is just basic good practice to include a clear explanation of the relationship between the two terms.
Item 6. (p532) .. the two largest languages in Indonesia, Javanese (75 million) and Sundanese (27 million) ... have recently sustained a massive switch to the national language, Bahasa Indonesia (now with over 200 million speakers, if mainly as a second language)' This is very poorly worded. Does it mean that the totals for Javanese and Sundanese have gone down because many speakers now speak Bahasa Indonesia instead? Or does it mean that Javanese and Sundanese have maintained their figures despite the fact that many speakers have recently acquired Bahasa Indonesia as well? Since I happen to know this field I'd say that the second is more likely to be true, but it is impossible to be sure of that from the clumsy text.
Item 7. (p532 same quote as Item 6). The term `second language' is used as if it were a straightforward, easily intelligible term. But it isn't. Suppose a man speaks Sundanese at home with his family and Bahasa Indonesia at work, using the language in the proportions 40:60. Which one is the `second language'? A tricky question but representative of a number of other matters which in a book of this size on this subject the author should have discussed quite thoroughly. (The index doesn't contain `second language')
Item 8. The concept of a creole language (ie, a language formed by the merger of several) is certainly pretty relevant to the book's subject. The term is first mentioned briefly on p10 without definition. Next on p292 the text includes the phrase `language mixing, or creolisation' and that is as far as definition goes; it is suggested that British and Irish Celtic have experienced this phenomenon. On p390 with mention of Indo-Portuguese creoles and on p415 with French creoles the author goes right ahead on the assumption that the reader knows exactly what is meant by the term. He doesn't say what the other languages were that mixed with Portuguese and French respectively. I rather doubt that British and Irish Celtic are in fact good examples of creolisation but that isn't the main point, because I'm less concerned with the factual content than with the quality of authorial judgement. Since creolisation is such a relevant concept in the development of the world's languages, the author should have realised that the book needed far more explanation and analysis of it.
Item 9. Lingua franca is another key concept. It has plenty of index entries: no less than 20 languages serving as a lingua franca are separately indexed. I've looked at the five entries that come earliest in the book. None of them give an adequate definition and analysis of the term. The author just assumes that any reader with no prior knowledge of linguistics knows the one and only meaning of that term. In a book of this size he should have done better than that. For example, a lingua franca that is nobody's main language (eg Latin at some periods) is a different thing from one that is the main language of many of its speakers but by no means all (eg English now). A national language like Bahasa Indonesia is another case again. Then there is the matter of the relation between the concepts of lingua franca and creole languages . . The author should have realised when planning the book that it was worth giving the reader a good insight into this concept.
Item 10. Substrate is another of those important concepts in language development that deserved fuller treatment. It doesn't appear in the index at all. However I did notice it on p292 with reference to Celtic. The text there rather gives the impression that `substrate' is a hypothesis that applies only to the particular case of Celtic. The author surely knows very well that this is phenomenon (or at least hypothesis) is applicable to very many cases of linguistic change. So here we have both a clumsily written piece of text and a wrong judgement about the amount of emphasis the concept deserved.
Item 11. Anybody developing a book of this size about this subject ought to notice pretty early on in the project that there is a big strategic design decision to be faced. You want to tell a number of different stories (story of Sanskrit, story of Europe's languages abroad etc). To understand one story the reader needs to grasp a number of general concepts in linguistic change (lingua franca, substrate, spread of language by conquest etc); to understand another story some but not all of the same general concepts come into play again; and so on. So how do you structure your 600-page book?
One approach is to frontload the book with substantial accounts of all the main general concepts, so that the reader gets thoroughly familiar with them. Only after that go on to tell the separate stories, constantly showing which general concepts apply to each story. The alternative approach is to start off telling the stories and introduce each general concept thoroughly as and when it first comes up. Thus many concepts will come in the first story. In the second story you introduce and explain thoroughly any further concepts that are relevant, while just referring without explanation to those concepts already covered in the previous chapter. And so on.
It is a hard thing to say but I suspect that Ostler didn't think carefully enough about the design of this book. What he has produced is a very feeble version of the second option, where instead of introducing concepts thoroughly on their first appearance, he just skids through them as briskly as possible. This is a strategic authorial blunder that results in a fundamentally unsatisfactory book.
The above set of items shows authorial incompetence at several different levels: faulty book strategy; irrelevant material; failure to define things; failure to stress essential distinctions; clumsy sentences. If you happen to know about some of the languages covered, then their parts of the book can be pretty irritating; and the parts about the other languages don't inspire confidence that you are getting any accurate insight.