The discussion rages on an Internet group how best to teach sight-reading. Most of the techniques touted as effective are all visual aids, only rarely does anyone mention the need to attach any of the alphabet or number devices to actual sounds.
Today I was describing to a colleague how difficult it was for me to learn to listen without looking at a score for as soon as I look at a score I know what I should hear, rather than having any awareness of what I am actually hearing -- there is a big difference. If I analyze in advance what the student should be playing I deprive the student of my attempt to understand what he or she does in fact hear, so I am stuck with a categorical right note/wrong note approach, rather than with the far more interesting "Ah! I hear how you understand this passage and can now show you what alternatives the composer has suggested."
It was not easy to learn to listen this way but the rewards have been great both for myself and for my students, both in piano and in chamber music.
I went to this trouble in order to demonstrate the validity of Viktor Zuckerkandl's insights into the innate logic of the musical mind and of the unfailing bodily sense of following that logic even into the realm of the conventional wrong note.*
Without that kind of listening I would never have noticed that some tonal works are conceived so as to distort the reliable tuning to the tonic triad or even to the instrument's innate resonance (like the violin's A or the cello's G). When I have heard that happen it has been stunning to identify the event and then be unable to discern it by visually studying the score - so overpoweringly inhibiting is the conceptual world represented by standard music notation.
*Sound and Symbol: Music and the External World, published by Bollingen Press, 1956