The Atoms of Language: The Mind's Hidden Rules of Grammar

Rating:
Summary: You're better off reading the Language Instinct
Review: Despite having a catchy name, this book is a waste of time, you're better off reading Pinker or go for the real juice (Chomsky).
This book is building evidence towards foregone conclusions: namely, that all languages differ by only a certain set of parameters. This was put forth by Chomsky in the 80's, but it takes Baker over 50 pages to spit out.
It starts with "the mystery and paradox" of the Navajo code-talkers. The paradox being, if Navajo and English are so different as to be undecipherable by the Japanese in WWII, how come they're similar enough to translate from one to the other very quickly by the code-talkers? This is maybe a mystery for children: obviously all languages in the world are similar to a degree because they're all spoken by humans, hence at least basic sentences like "Please bomb the island on the right" should be easy to translate. On the other hand, given that the code was used for hardly 2 years, only in very specific tactical situation, the Japanese had very little time to work it: they couldn't very well just go to Utah or wherever and take "Navajo 101". But isn't it clear that given a little more time they would have been able to learn it? Navajo code was obviously a short-term solution, and was efficient mostly in the field as using coding equipment and code-books of the 40's, on a Pacific island, while mortar shells are landing around you, is not exactly easy.
So much for that paradox.
And what is all this crap about chemistry? What connection is there between linguistics and chemistry? The genius of the chemists was to realize that all elements are composed of the same stuff; but we already know that all languages are composed of the same stuff since they're all spoken by humans, namely one of the few million species that inhabit the earth. The next genius in chemistry discovered that there is a relationship between the different elements; but obviously we know about big groups of related languages, like Indo-European languages. So there's hardly any surprise there.
Baker claims that the next task is to connect all the parameters together in a sort of periodic table, like Mendeleyev did. Dream on.

As for principles and parameters: Baker himself says that for every proposed parameter, there appear to be languages that don't behave well with it. He hides behind the fact that these languages are less frequent- but what's the point of having this parameter if some languages, even a single one, violates it? You will end up with as many parameters as there are languages, and that is no explanation at all.

Rating:
Summary: Excellent book, but not for everyone
Review: The Principle and Parameters approach in linguistics is one of Chomsky's most profound theoretical insights, and its elegance and depth in explaining linguistic phenomena across languages is one of the most impressive achievements in linguistics. Baker's book is the best popular introduction to the approach that I have read. It is not as fun and entertaining as Pinker, but it is certainly as understandable, and it does not "dumb down" quite as much as Pinker. The book is a quick read, and contains an impressive chapter on Mohawk. Baker takes the theoretical approaches that he introduces earlier in the book, applies it to the case of Mohawk, formulates a novel explanation, and shows how we can get a deep understanding of the structure of Mohawk from a few, easily understood and elegant principles.

All in all, the book is an excellent introduction to how linguistics is done, and the models through which linguists currently think about languages and linguistic phenomena. It gives the best, most understandable explanation of central theoretical concepts such as "parameter" and "I-language" that I have seen, and gives a brief overview of "optimality theory" and other hypotheses in competition to Chomsky's version of P&P.

There is much to learn from this book, but I think that only those with a genuine interest in and sympathy to generative linguistics will find this book illuminating. To appreciate the depth and insight of the Principles and Parameters approach, you need some mastery of the technicalities and constructions, and mastery of the technicalities requires patience. To understand the problems and solutions that arise, you have to be willing to sit, think, and go over words and sentences in exotic languages slowly, including their inflections, affixes, and word order. Baker provides enough so that anyone can understand them; but you will still need to spend some time on these sections. I recommend this book to educated readers with some competence in linguistics, or to those who have a genuine interest in learning about generative grammar by looking at specific exotic languages.

Rating:
Summary: The Excitement of Dry Categorization
Review: The title of the book comes from the realization that the syntax of languages may be composed of true elements, like atoms which can normally combine only in particular ways so that certain kinds of langauges will not occur, or will do so for only for a short time before decomposing into a more stable type of language.

Linguists are still in the process of identifying these atoms and Baker is giving a popular account of the current state of investigation.

Mark C. Baker explains modern attempts to break down and categorize language by its syntax and by binary parameters that work thoughout each language providing rules that people following unconsciously in generating new utterances within any particular language.

He demonstrates that languages can be catagorized according to particular parameters which don't appear to have ANY relationship to the culture of the people speaking the language. For example, in building phrases within phrases most languages consistantly add new elements to phrases to create a larger phrase either always at the begnning of the smaller phrase or always at the end.

This seems to refute beliefs that differences in languages indicate fundamental differences in world views. Factually people of almost identical culture live side by side speaking languages that differ drastically syntactically.

So languages seemingly do NOT vary from each other in unlimited ways. Therefore there MUST be rules about what does and does not NORMALLY happen and presumably rules to the exceptions and to the exceptions to the exceptions.

These rules would be innate in human consciousness and would provide the foundations on which the actual syntax of a languages is based.

Languages can be classified syntactically according to type and sub-type and so forth entirely independantly of any genelogical relationships between them.

Baker's writing is lucid and transparent and he lets his subject matter and the puzzles it presents carry the excitement in the book.