November 3, 2008

Wiki proofing new version of AHBP: "Foreword by Peter Brimelow"

This is a comment thread for "Foreword by Peter Brimelow" in the revised version of Steve Sailer's America's Half-Blood Prince: Barack Obama's "Story of Race and Inheritance" that was posted online 11/3/08:

In the comments below, feel free to list typos, errors, and other comments.

The idea with all these web postings is that if, as you are reading my book, you would like to call my attention to something, you can add it as a comment under the appropriate chapter's posting. You can also scan to see if somebody else has already caught it.

This is not a request that you participate in the next round of proofing. I think the book is in pretty good shape now. It's just if you'd like to register any kind of point related to the text, this long series of postings makes it easy to do in an organized fashion.

By the way, if you can't tell which version you are looking at, you can distinguish this revised version by the date on the second page: it says October 24, 2008, while the first says October 23, 2008.

My published articles are archived at -- Steve Sailer


Anonymous said...

page xii - is Peter Brimelow really four time zones away from you?
Also, I'd prefer "Mainstream Media" to "Main Stream Media" but I guess the latter points up the common acronym, MSM, more clearly.

Anonymous said...

Steve --

A problem for your book and postings is that by all accounts, Obama did NOT write "Dreams from My Father: A Story of Race and Inheritance" and Bill Ayers DID.

Someone has run "Fiction Fixer" on the work of both men: see Jack Cashill and the details.

There are so many metrics that show both works written by the same men that it's clear that Obama did not write his first autobiography.

This seriously undermines your work -- that his autobiography shows a clear picture of Obama, in his own words, and in his own writing.

I did warn you, that there was the possibility of someone doing the same sorts of analysis as that of Joe Klein in Primary Colors. It appears now that this is indeed the case ... Obama did NOT write his book.

The other conclusion of course, is that Obama simply is neither disciplined nor smart enough to write that book, seriously undermining your analysis about him.

Anonymous said...

Since I have only commented in the past to make fun of your overwrought concern over the tiny and successful effort of the Bush admin. w.r.t. Somolia, I thought I should say that I am reading and enjoying the PDF of your Obama book, plan to purchase a copy, and think it could be pretty significant.

Anonymous said...

Steve, this format will be painful. Why not use a public hosted wiki like google sites or pbwiki or something and just have people comment there? checking one page may happen, checking twelve pages won't happen

Anonymous said...

Tomorrow promises to be another depressing day for non-leftist. I think it’s as a good time as any for some escapism.

Wouldn’t it be great if Steve could compile a nice list of NON-non-fiction books to recommend your readers, after finishing your book? I added The Coup to my list from there.

Anonymous said...


For what it's worth, I have a PhD in statistics and have done a lot of work in statistical NLP. I thus read the Fiction Fixer analysis ( with some interest.

I remain open to the idea that one of Obama's books was ghostwritten, though I think it is more likely to be the 2006 Audacity of Hope than the 1995 Dreams from my Father; there were many more demands on Obama's time in 2006 than 1995.

However, the evidence presented in that pdf is not sufficient to make the case. You need some control samples to establish that the degree of resemblance is more than would happen by chance (I'd also like to see some scatterplots of the data, as he presents only tables).


1) Run FictionFixer or a similar featurization program on a *large number* of corpora. The result is a set of data structures containing numerical summaries of the corpora. For simplicity, let us assume this is just a single (long) count vector of fixed length for each corpus.

2) Make sure this set of corpora includes as much writing by Ayers and Obama as possible

3) See if features tend to be signatures of a person's writing style. This should be done both internally (by splitting paragraphs or chapters from the same corpus and calculating feature vectors) and externally (by comparing feature vectors computed from different corpora written by the same individual)

4) If this appears promising, calculate all pairwise correlations between feature vectors and apply unsupervised clustering of the resulting diagonal correlation matrix. If all goes well, corpora from the same author should be represented by block diagonals in this matrix.

5) You can boost the rigor by computing two empirical correlation distributions -- one for the distribution of correlations of feature vectors from corpora of the same author, and one for the distribution of feature vector correlations from corpora of different authors.

You can then put the correlation between the feature vectors from "Dreams from my Father" and "Fugitive Days" on this plot -- the number should be near the median correlation between works of the same author.

Anonymous said...

"Steve, this format will be painful. Why not use a public hosted wiki like google sites or pbwiki or something and just have people comment there?"

I strongly agree with this. Wikiproofing without a wiki is a bad idea, especially since you're editing this a day early. You're writing a Superbowl column the day before the game. No you're writing a Superbowl BOOK a day before the game. You must acknowledge the results of the election, he was elected president (or he was not). You can't have sentence after sentence in every section that makes Obama's victory (or defeat if that's how it rolls) in the hypothetical future.

To give two examples-- "This book is too idiosyncratic to have an effect on the election"(xvi) OR "if Obama somehow loses in 2008"(41)

You need to strip everything that references the "future" election. Your lede is simple (assuming McCain doesn't surprised us all):

Barack Hussein Obama was elected President of the United States on the Fourth day of November, in the year of our Lord two thousand eight, and of the Independence of the United States of America the two hundred and thirty-third (as they say in Presidential Proclamations).

That fact should also be the last item on your timeline.

At no time, refer to a future even

Anonymous said...

t99 beat me to it. I especially like the flourish mentioned below by Cashill:

In assessing the postmodern diffidence of the 1995 Introduction and the 2004 Preface to Dreams, Ohio State classics professor, Bruce Heiden makes a fascinating claim: Obama takes no credit for the actual writing process beyond the curiously passive, “What has found its way onto these pages is a record of a personal, interior journey.” Adds Heiden with a smile:

As Obama tells it, his authorship of Dreams was miraculous, because although he lacked the writing skill to be the author of anything, and he didn’t want to be the author of a memoir in particular, and he resisted becoming the author of a memoir, and he tried in vain to become the author of a different kind of book . . . Dreams from My Father, nevertheless somehow it ‘found its way’ onto the page with Barack Obama’s name under the title as the author.

The evidence strongly suggests that Ayers transformed Obama from the struggling literalist of 1994 into the sophisticated postmodernist of 1995, and he did so not by tutoring, but by rewriting, including the very Introduction and Preface.

Anonymous said...

All may not be lost with Steve's book:

Ayers seems to have written only parts of Dreams.

Ayers apparently edited others, and all but ignored long sections. One would expect a higher raw count of any repetitive Ayers’ word in Fugitive Days.

So which sections are Ayers and which are Obama? How does this affect Steve's book.

Anonymous said...

haha wow, you let brimelow write the foreword? I know you don't care about being taken seriously outside of your own kook-osphere but Brimelow is on par with Jared Taylor, at least in the eyes of moderates.

Anonymous said...

Very sassy read. I plan to review it so I leave one adjective to summarize my reading experience for now.
My two minor complaints are:
1. As you suggested it might " " is gone with the wind.

2. There is an unclosed parenthesis on the endnotes introduction page although your copyediting is 99.99999% accurate.

If the book were packaged like a "stupid George Bush" book it would have sold tons.