Late in 2010 I published on the blog of Institut Agile an essay in French, the latter half of which was a critique of Steve McConnell's 2008 blog entry "Productivity Variations Among Developers and Teams: The Origin of 10x". The context of this critical essay was a series in which I examined some of the issues in arriving at reliable knowledge in the domain of software engineering.
In January, I posted to my personal GitHub site and to various social networks an English translation of that article, which thanks to the leverage of social networks enjoyed a relatively wide audience. This led to a series of exchanges on the blog of Steve McConnell and others, and some private correspondence. On net, my impression has been that this discussion has left the topic no clearer than it was previously, and I take responsibility for this poor result.
On reflection I have come to realize that my claims, in both the original and the translated version, that McConnell was "cheating with citations" and "abusing the mechanism of scientific citation" were overblown. As a consequence of this carelessness on my part, the discussions that ensued were less fruitful than they might have been if I had presented my findings in a more neutral way, to be improved upon as necessary in the interests of factual accuracy.
Even when dealing with charged issues, my normal mode of operation in debate is to avoid framing it as one person against another, but as both working in pursuit of some truth; in this instance, and on several occasions over a period of time, I did less than my best.
On the bright side, I'm happy with many of the insights I was able to glean from this exchange, frustrating as it may have been to other readers. I am resolved to do a better job, at some future time, of presenting what I found during my investigation of the empirical evidence that McConnell and others have collected regarding programmer productivity. (A draft of this report is available and I will provide it upon request.)