PHP is popular because it compliments the optimistic coding style employed by most programmers.

Much of PHP's popularity stems from the easy deployment (just copy up the .php files) and that every $1/mo cheapo hosting service supports it.

I'd be interested to see what the correlation between experience in each language and satisfaction with that language looks like. PHP is, by far, the most widely-used web development language, so I would expect it to be rated highly simply by virtue of greater familiarity - if you can create a new site in PHP that uses all the latest whiz-bangs without breaking a sweat, but have used Python so little that every line of Python has you beating your head on the wall trying to figure it out, then of course you'll report higher satisfaction with PHP.

I suspect it would be much more informative to see whether satisfaction increases more quickly as you gain experience in one language than another - or whether there are any for which increased experience typically leads to decreased satisfaction.

The problem with a survey of this kind is finding developers who have extensive and roughly equal experience with all the languages. Unless this was done (the article does not specify) the results are rather meaningless. Obviously a developer will usually prefer the language thay are most familiar with, and developers working for money in commercial environments are seldom able to choose which language to use based on personal preference.
What would be interesting si to get a group of people who have never programmed in any of the target languages to participate in a trial of those languages, coding a spcific test case application, and see which they find easier to get productive with.

I'd think that the people doing this survey would be smart enough to survey developers that have had a decent amount of experience with each language.

Honestly I don't find it all that shocking that PHP would still win. In basic syntax its still the closest out of the bunch to the mainstreams like C/C++ and Java except Perl but then has more web oriented goodies than Perl. Now some of you may be thinking about comparing frameworks like Rails and Django but this is about the language itself. If you track the Tiobe index you'll see that none of the other languages in question have come close to the popularity of PHP.

Maybe as developers that were taught with C/C++ give way to more self-taught guys or those that learned only OOP based languages you'll see surveys like this change. The difference between count(x) and x.count() are only huge to the newer dev crowd and thats what you'll see most of them say when asked for a comparison.

Dave, in response to your question about the relationship between experience and satisfaction, I have found that often developer satisfaction with PHP increases with experience, whereas many developers who are new to the language start out with some trepidation, as they are getting used to some of the quirks. However, once the quirks are worked out, and PHP developers "get it", they start making a lot of progress very quickly. With PHP's adoption still growing, I would venture to predict that PHP will retain this position for some time as developers get more experience with it outside of the "el-cheapo website hosting company" realm.