Imagined Communities is a transdisciplinary classic now read widely across the social sciences. Anderson's aim is to correct weaknesses in both liberal and Marxist approaches in political science to understand nationalism. This is a worthy goal, because nationalism is the water in which we swim, a mostly invisible field in which we enact Americanness, or Indonesianness, or any of the 195 countries that Google claim exists. Anderson's goal is to denaturalize the nation, to show it as particular kind of modern imaginary that only became possible towards the end of the 18th century, and only really omnipresent in the 20th.

Anderson's nations are sovereign limited communities, a specific kind of belonging such that everyone on Earth belongs to one, perhaps two nations. The nation is historically transcendent, France may fall, but La France is eternal. And despite this transcendence, the nation is also bounded by historical consciousnesses, with days of independence and founding citizens.

Anderson locates the origins of this style of thought in the late 18th century, and in the specific mode of thinking engendered by newspapers and novels that time is continuous series of causes and effects. All days are the same kind of thing, but each individual day is different. A sense of historical consciousness is required to separate the present from antiquity, while the rising class of bourgeois administrators and businessmen, who started the day reading newspapers, develop standardized beliefs based around the importance of metropolitan life in the capitol.

This shift of consciousness is one of the hardest things to write about. It's the primary topic that Foucault spent his career groping towards. It is also the least evidenced part of the book, as Anderson compares two classics of Filipino literature, the modern nationalistic Noli Me Tangere and the older Pinagdaanang Buhay ni Florante at ni Laura sa Cahariang Albania, a medieval fantasy epic poem set in Albania. The form of the novel, that events begin at a precise moment with an assumed past, that time can pass in detailed dialog or be elided in weeks depending on the needs of the plot, rather than the smeared out Now of pre-modernity is interesting, but difficult to prove.

A second area that Anderson covers is the relationship of the nation to the state. The state, defined as the administration of taxes, justice, warfare, infrastructure, education, and welfare, is far older than nations. A key issue for nationalism was the need for educated administrative officials in the colonies, but a firm bar on their advancement beyond their home region. From Spanish South America to French Indochina, frustrated educated youths served as key agents in nationalistic revolutions.

Anderson's book is fascinating, erudite--if cursed with a tendency to assume that the reader is also fluent in French, and useful for sweeping out the cobwebs of conventional wisdom. I can see why it became such a success, and is widely taught. Yet it reminds me of two other theoretical paradigms, Jasanoff's sociotechnical imaginaries (clearly borrowing from Anderson) and Winner's politics of artifacts, in that it is seductive but ultimate explains less than it conceals. And for a junior scholar, you are simply not good enough to deploy this paradigm in your own research. Anderson offers a potent theory, yet one that doesn't not go far enough to explain the potent lure of nationalism in the 20th and 21st century. If nations are imaginary communities, they are ones which millions of people have died to preserve. Few other imaginations have that potency, only religion comes close, and if "imaginary" deconstructs the mythology that underlies a nation (George Washington, William Tell, anthems, the iconography of ruins, etc), it doesn't come close to explaining the vital relationship between the individual and that imaginary.