I've been following this thread on-and-off, but after reading a newspaper article this morning (in a Charlotte newspaper, of all
places) about surveys showing good feelings for the US are at an all- time high in Japan, I thought I'd put on my dusty old foreign policy scholar hat and weigh in.

1) Ronald Dore asserts the deterministic rise of China.
I'd just note that his language is almost identical to the 1950s-era descriptions of the Soviet Union. Or depictions of the way North Korea was going to
outcompete the South. He also talks about the
Chinese ability to
"devote the best brains" to new weapons research--did that remind anyone else of the 1950s-era rhetoric about a "Missile
Gap?" The
lesson: long-term power predictions based on straight-line assumptions of current alarmism aren't worth the pixels they're written with.

2) I'm surprised that no one has referred to the extensive IR
literature on great power transition. For much of
recent history
this was the single most important question in the IR
sub-field.
Books on the subject by people like Gilpin or Keohane were on every IR grad student's reading list for at least a decade or
two. The
lesson: simplistic models don't predict very well.

3) I'd also look at the security community literature to better understand the US-Japan alliance--a literature that suggests the US- Japan alliance is resilient in ways that Dore apparently doesn't expect.
That also raises the issue of multilateral security
institutions: no discussion of Japanese security policy is credible without considering the role of the UN at a deeper level than "just pull out of the NPT."

4) Dore's argument about how Japan will Finlandize to a rising China resembles one that David Kang has been making for years about Asia's historically "China-centered" international system. You can read Kang's work for more detail, but I'd point out that Dore doesn't resolve two problems with applying the Kang hypothesis now: 1) Dating to the 12th century, Japan's _political_ participation in that system has often been antagonistic, not as a Chinese client, so why should it predict Japan will become a client now?
and 2) that system was predicated on the absence of a large external power (as in, there weren't long lines to become a Han Chinese client when the Mongols ruled or when the Colonial Empires ruled).

5) The original post made two suggestions: 1) build nukes and 2) leave the US alliance. Maybe I missed someone's post on this, but I don't see why those two have to go together. Why can't Japan build nukes and stay allied with the US? It works for the UK and France (mostly). It works for Israel, too (unofficially).
Dore's argument
only makes sense if you assume some antecedent burning desire to leave
the US alliance. The record high levels of support
for the US I
cited above make me wonder why you think that assumption is
sustainable? I'd also point out that even during the
darkest days of
the Iraq War, polls put support for the US in Japan at significantly above that in most countries.

6) There is a vein in the alliance literature that argues alliances are not just for safety--they're also about information and institutional constraints. If I wanted to argue that Japan will ally
with China, this is where I would start. Note,
though, that this
literature's take on alliances breaks with the binary approach of "either China or the US" and sees a much more complicated world where countries have alliances with multiple (even contradictory) parties, using diverse structures and for different reasons.

One can make a sophisticated argument that Japan (and the US) need to rethink the US-Japan security relationship, as well as Japan's security posture more fundamentally, and Dore is right to raise that
flag. One can even do it without making heroic
assumption that
Japanese policy makers will overcome their nuclear
allergy. I would,
though, suggest that domestic politics specialists who want to make those arguments first familiarize themselves with their foreign policy counterparts'
work.

-----------------------

Robin, I understand your experience with ill-fitting stereotypes and what I suspect is an ethnographer's frustration with Large-N research that lumps individuals into the researcher's categories rather than subjects' own understandings. To ride my own personal hobby horse, I'd say most of what's been written on the Komeito and Soka Gakkai contains more sterotypes than it does scholarship. Still, everyone on this list surely refers to "the Japanese" from time to time (in class, if nowhere else). In Bicycle Citizens you yourself lump millions of diverse individuals into a broad category called "housewives"
and
generalize about them. Are you really saying that
all of us are
expressing our prejudicial stereotypes whenever we use short hand generalizations?

George Ehrhardt