You're oversimplifying.

The medical and therapeutic community has a host of terms for various clusters of conditions and symptoms associated with depression. For some, the approach to treatment is pretty clear: If the person also has severe social anxiety for example, treating the anxiety should also alleviate the depression. For others, the standard is to try therapies until something works. Since even moderate depression shaves off years of life expectancy, it is worth trying a fifth kind of therapy if the previously four didn't do it.

Intense grief over a lost partner is normal, as is anxiety over unsatisfactory work. Severe depression can be a lot worse than either.

Just the fact that very different people with very different genetics in very different life circumstances experience a similar mental phenomenon called "depression", doesn't necessarily imply they can all be treated the same way. Phenomenological descriptions may be similar, but causally are they similar? A trivial example: a person who's depressed because her husband died would require a different treatment than a PhD student who's depressed because her research isn't progressing.

For a given person, some treatments will work while others won't. So yes, there seem to be multiple possible causes. But

• Sometimes somebody will respond to his or her respond to circumstances differently from how other people would respond in the same circumstances, suggesting the existence of an underlying factor (whether biological or dispositional) that's circumstance independent.
• Therapy can be customized so as to address a diversity of possibilities.

This is why I'm not too enthusiastic about anti-depressants; they seem akin to painkillers.

A variety of factors may cause undesirable brain chemistry, but sometimes undesirable brain chemistry might be a part of the default physiology of the individual.