orpheus wrote:But maybe I'm mistaken. Aren't neurotransmitters chemicals?
The problem isn't that the term "chemical imbalance" is incorrect, but rather the opposite: it is technically correct. However, it is so correct that it is meaningless. In other words, all it is saying is, "The brain is doing stuff". And of course the brain is doing stuff because the brain controls our behaviors and we're describing a problem with our behaviors. So it explains everything, and something that explains everything explains nothing.
For example, what causes addiction? A chemical imbalance. What causes schizophrenia? A chemical imbalance. What causes anger? A chemical imbalance. What causes you to choose cereal over toast in the morning? A chemical imbalance. It doesn't tell us anything and it would almost be more useful to say, "The brain did it". What causes depression? "Stuff in the brain".
Another problem is that the term can be misleading. The finding that antidepressants can be helpful doesn't tell us anything about chemicals in the brain, yet we instinctively want to use it as evidence that there is a biological brain problem with depression. The comparable situation would be to suppose your dog dies and you get a bit sad, and someone gives you some ecstasy. Suddenly you aren't sad any more, but it would be unhelpful to suggest that your sadness was caused by a "chemical imbalance". In a sense, it is a chemical imbalance because the processing of thoughts and emotions is the unbalancing of chemicals in the brain, but we all know that the sadness was caused by the death of the dog. (Note: I'm not trying to say that depression is like "being sad", nor that antidepressants work exactly like ecstasy, instead it's just an analogy of the flaw in simplistic reasoning).
Although on a slightly different topic, I think the issue is dealt with quite well by Neuroskeptic's post here:
Brain Scan Proves that the Brain Does Stuff.