American Indian Languages: The Historical Linguistics of Native America (Oxford Studies in Anthropological Linguistics, 4)

Rating:
Summary: Where's the Data?
Review: Thoroughly miserable, this book never delivers on its title. Its focus is purely negative: an ad hoc piecemeal attack on the author's apparent nemesis, Joseph Greenberg.

Rather than offering his own positive arguments in support of any genetic relationships between all but the most obvious American language families, Campbell spends the great majority of his time attacking certain portions of the evidence that other classifiers offer. He never addresses the overall context of any circumstance, and he remains silent in the face of any evidence he can't refute. The actual posited cognates in question are almost never cited, and native forms in general are quite rare.

Regardless of whether one proposes six, two-dozen, or the over 50 independent and further un-relatable North American stocks that Campbell clings to, any book that purports to study the "Historical Linguistics of Native America" should at least be chock-full of native words or texts, with grammatical sketches and detailed phonetic transcriptions.

Yet the most we get for the majority of languages or families is a list of phonemes without the context of even one single native word. In a few families case endings or pronouns are given. There is not one sketch, text, or real wordlist. The maps given are available elsewhere. The entirety of the native forms cited could fit on one single leaf.

This book's fatal flaw is its exclusively negative focus. Pages upon pages list references in English to secondary and tertiary sources, but the subject languages themselves are studiously ignored. Never making any positive argument of his own, he never feels obliged to provide the one thing a thinking reader wants, the evidence.

Campbell further embarrasses himself with his uniquely idiosyncratic system of probability analysis. He cites various theories of distant relationships proposed by other scholars. He then (admittedly subjectively) grades the likeliness of these theories, not on a scale of 0% to 100% as is universally accepted, but rather on a scale of positive to negative (!) 100%, with a 0% probability on his scale indicating an actual probability of 50%.

For example, he finds the Tlingit-Eyak-Athabaskan hypothesis to have a +75% probability, by which he means that it is actually 87.5% likely. But to the Na Dene hypothesis (the above family linked to Haida) he gives a 0% probability, by which means an actual 50% likelihood. Any link between Zuni and Penutian (however constituted) he gives a -80% probability. Yes, that's a "negative eighty percent," by which he means an actual possibility of 10%.

Confused? Then don't buy this book. Marianne Mithun's "The Languages of North America" is an excellent general source for north of the Rio Grande, with a conservative classification, a well-specimened typological overview of the documented variation and at least a phonology, sketch, and brief text of each language family. Maps of North America are as good as Campbell's.

Rating:
Summary: Where's the Data?
Review: Thoroughly miserable, this book never delivers on its title. Its focus is purely negative: an ad hoc piecemeal attack on the author's apparent nemesis, Joseph Greenberg.

Rather than offering his own positive arguments in support of any genetic relationships between all but the most obvious American language families, Campbell spends the great majority of his time attacking certain portions of the evidence that other classifiers offer. He never addresses the overall context of any circumstance, and he remains silent in the face of any evidence he can't refute. The actual posited cognates in question are almost never cited, and native forms in general are quite rare.

Regardless of whether one proposes six, two-dozen, or the over 50 independent and further un-relatable North American stocks that Campbell clings to, any book that purports to study the "Historical Linguistics of Native America" should at least be chock-full of native words or texts, with grammatical sketches and detailed phonetic transcriptions.

Yet the most we get for the majority of languages or families is a list of phonemes without the context of even one single native word. In a few families case endings or pronouns are given. There is not one sketch, text, or real wordlist. The maps given are available elsewhere. The entirety of the native forms cited could fit on one single leaf.

This book's fatal flaw is its exclusively negative focus. Pages upon pages list references in English to secondary and tertiary sources, but the subject languages themselves are studiously ignored. Never making any positive argument of his own, he never feels obliged to provide the one thing a thinking reader wants, the evidence.

Campbell further embarrasses himself with his uniquely idiosyncratic system of probability analysis. He cites various theories of distant relationships proposed by other scholars. He then (admittedly subjectively) grades the likeliness of these theories, not on a scale of 0% to 100% as is universally accepted, but rather on a scale of positive to negative (!) 100%, with a 0% probability on his scale indicating an actual probability of 50%.

For example, he finds the Tlingit-Eyak-Athabaskan hypothesis to have a +75% probability, by which he means that it is actually 87.5% likely. But to the Na Dene hypothesis (the above family linked to Haida) he gives a 0% probability, by which means an actual 50% likelihood. Any link between Zuni and Penutian (however constituted) he gives a -80% probability. Yes, that's a "negative eighty percent," by which he means an actual possibility of 10%.

Confused? Then don't buy this book. Marianne Mithun's "The Languages of North America" is an excellent general source for north of the Rio Grande, with a conservative classification, a well-specimened typological overview of the documented variation and at least a phonology, sketch, and brief text of each language family. Maps of North America are as good as Campbell's.