Teaching babies some basic sign language is becoming more and more popular (or, at least, I am more and more aware of it, which might be because I'm a new parent). It seems to be a relative unproven technique; all evidence I've heard is anecdotal, not quantitative. I've recently had a new thought, though, that not only does baby signing promote communication: I think it may actually promote verbalization, specifically.
Quick background: the main idea, as I understand it, behind doing any baby signing is that babies become capable of communication before they are physically capable of speech (verbalization). By teaching a baby some basic signs for 'important' concepts (e.g., "food", "water", "more", "all finished", etc.), parents can help give babies (after 6-9 months) a way to communicate their needs. Without signing, many babies rely on their parents to guess what they need, or often resort to grunting, crying, or screaming. It can be frustrating for the little ones to not be able to tell you what they want: signing can help alleviate some of that frustration.
Based on our experience, and that of a few friends we know who've tried it, signing has been wonderful. The idea is not to raise a child fully fluent in American Sign Language; a few signs are enough. It was wonderful to watch Jessica as she first made the connection between concepts like "milk" and the hand gestures that go with them.
The suggestion has been made from time to time that baby signing might delay verbalization. The concern is that if the baby doesn't "need" to speak because they can already communicate by signing, then the development of speech might be delayed. My instinct has been that this probably doesn't happen, but as far as I know there's no proof that it doesn't. I can never prove that Jessica might not have started speaking earlier without the signing. This suggestion has remained one that I doubt but can never dismiss.
My recent thought is that signing might actually have the opposite effect: it might promote the development of speech, by allowing for more (and more immediate) positive feedback. Here's what I'm thinking: when babies first start saying a new word, it is often a rough approximation at best. I fully admit that there are times when I understand Jessica perfectly well based mostly on context, where other people would probably find her utterances to be gibberish. And there are plenty of occasions when even Dana and I can't be sure what it is she's trying to say.
If she makes a sign, though, while she attempts to say a word, it allows us to quickly repeat her word back to her and give her positive feedback for successful communication. Even though her verbalization was rough, we can encourage her to keep it up. She gets to hear us repeat the word she had in her head, even if her spoken utterance was unintelligible.
I suppose you could argue that this only encourages her to make unintelligible utterances. I don't think it will work that way, but I guess that's another argument I won't ever be able to prove conclusively.