Why are so many programmers arrogant?

Jacquesm's statement in no way suggests a smart person should do his own plumbing. In fact, because a person is smart, he will be more likely to hire a professional. Or, if he does try it himself, he will succeed more often, because he gets the appropriate information and training beforehand. These arguments are actually somewhat circular, because this behavior partly defines what we perceive as 'smart people': people that succeed at what they attempt.

Smart people that know they are smart will inevitably come across as arrogant, because when they think they are right, they know chances are high they actually are right and they will not easily let an opinion drop. They know when they don't know stuff and they won't often end up having to admit they were wrong, because they weren't claiming to be right in the first place. The 'other side' doesn't see this: they only see someone that is pretty sure he is right most of the time and that hardly ever admits he was wrong.

What does this have to do with programmers? Programmers are generally perceived to be smart people and as a result programmers tend to consider themselves to be pretty smart. Unfortunately, this also holds for programmers that actually aren't all that smart and should be more modest with regards to their opinion about matters. The aggregate result of all this is that a relatively large group of programmers think too highly of themselves and that confirmation bias has led some to conclude that 'many programmers are arrogant'.

BTW, this is an explanation. I lack any solid data to back it up. It's an explanatory narrative and it may be wrong. However, if enough people think it sounds right, it's probably right. Another reason why people sometimes seem arrogant, is that they may strongly defend some position, while it never becomes clear during the discussion that they are well aware they may be wrong. If you defend a position very well, it just means that I am wrong, if I am claiming you are wrong. However, that doesn't mean that you are right. We can both be wrong, but, more importantly, we are probably both partially right.

I think a good deal of individuals that draw attention to it do so because they are generally embarrassed by some of the manifestations of this particular problem. You can have an otherwise intelligent person, who's writing can give the indication that there thoughts are somehow less valid, because there intelligence should be questioned due to gramatical errors.

I think stating it, is a way of an individual saying "hey don't discount my thoughts due to the vessel they come it"

I know for me personally, it is a source of major embarrassment and this damned iPhone keyboard only makes it dreadfully worse.

* Black-and-white feedback. Software is very black and white. Your dietitian can tell you to "reduce your sugar intake" instead of telling you you're fat because you drink so much cola. There's no way to sugar coat "you pressed the wrong button, and now you have to start over", unless you want to point out that the programmer who make the stupid program must have been an idiot. Either way, you look arrogant.

* We read a lot. Hackers usually know about the Milgram experiment, how Easter Island's economy collapsed when they cut down all the trees (peak wood), and why wikipedia may or may not be accurate. People who know it all come across as arrogant.

* We speak another language. No, not C. Programmers often speak a dialect that's a bit like English, but evolved into something a little different over IRC and BBs. That language isn't always rude, but it's different, and different can often come across as rude.

* We need to be intellectually assertive. We will find out why the core got dumped, and fix it. It's our brains against the compiler. If we don't win, our programs won't work.

* Arrogant programmers are still employable. There are plenty of arrogant HR officers, secretaries, school teachers, and shop assistants; but none of them act that way on the job. Sales guys are said to be unbelievably arrogant, but you can bet they will be very polite to potential customers.

* Heck, if we get too passive aggressive when we tell the PHB he's wrong, it can end in disaster. "Getting the team onto a version control system is universally agreed to be an essential best practice. It aids communication, reduces bugs, and keeps the source code safe" just isn't strong enough. He's already heard a similar pitch from a guy selling UML code generators, and look how that turned out. A bit more drama might be needed: "We aren't using a VCS. ANYBODY knows that this is as stupid as [your analogy] to control STDs."

Being a good programmer seems every bit as demanding as being a good mathematician, physicist or surgeon.

Or engineer, although I won't consider myself a good one. A "good enough" at best.