Following a study by John DeFrancis, chapter one covers six myths about Chinese characters: that they are ideographic, that they represent monosyllables, and claims for their universality, emulatability, indispensability, and successfullness. As an example of "twentieth century chinoiserie" Unger looks at Charles K. Bliss (Semantography and Blissymbolics). And he goes back to Andreas Mueller, a 17th century scholar who claimed to have discovered a key to Chinese characters — and who may have realised that "the overwhelming majority of the characters contain a phonetic component that indicates with fair accuracy the syllable of Chinese for which the character stands".
Again following DeFrancis, chapter two traces the real history of a Yukaghir "love letter" that supposedly exemplified ideographic picture writing. Unger also looks at the classification of writing systems and cartoonist Scott McCloud's model for the relationship between pictures and language.
The first part of chapter three is a scathing demolition of Leonard Schlain's The Alphabet versus the Goddess, which combines egregious errors about the Chinese language with confusions on other topics, ranging from primitive matriarchies to left-right hemispheric differences. The second part presents some evidence from psychological studies of people reading kanji.
Starting with Dave Barry, chapter four describes how leading intellectuals have spread confusion about Chinese characters. The central study is of Ernest Fenellosa and Ezra Pound's pseudo-translation of a Chinese poem, projecting "their own aesthetic preconceptions onto Chinese poetry by means of naive pictorial etymologies". And a survey of some particularly troublesome and commonly misread kanji helps explain why Japanese is hard to read.
In chapter five Unger looks at some of the tricks for memorisation offered by Harry Lorayne and at their application to the learning of kanji. Chapter six is an extended introduction to Gregg shorthand and the "positively uncanny" parallels between that and Japanese kanji; the longest chapter of Ideogram, it is also unusual in remaining focused on the one topic.
Chapter seven ("where do hunches come from?") contains an entertaining introduction to basic probability, along with an explanation of the Monty Hall "paradox"; the tenuous connection to Chinese characters is an attack on Platonism.
Chapter eight treats several topics, none convincingly. Unger presents Searle's Chinese Room thought experiment as if it were totally compelling, though he only mentions two of the weakest counterarguments and ignores the strongest and most obvious one (that the system as a whole understands Chinese). And some statistics Unger offers on academic satisfaction with university resources, in which South Korean and Japanese universities rank badly, seem more straightforwardly explained by generally poor resourcing than by linguistic factors, since the same trend appears in the non- (or less) language-related categories.
The final chapter mixes up philosophy, evolution of language, and linguistics. Unger continues his jeremiad against strong AI, ending Ideogram with a fiat declaration that no computer could ever really have humanity or free will, no matter how indistinguishable from a person it might be.
Some of the chapters in Ideogram are based on previous essays or columns and most of them cover largely independent topics; the volume as a whole feels like a jumble and would have benefited from restructuring. But the result makes for an easy read, though some topics are treated in more depth than others. Ideogram will particularly appeal to students of Japanese, but no knowledge of that language or of Chinese is necessary to follow it; it is recommended to anyone curious about language and linguistics.