Seeing that song, then, is to be the vehicle for emotion, even though that emotion be so tenuous as almost to defy verbal expression, for the most part we ally words and music. The timbre of a voice, singing tones without words, might carry a message to the sensitive, just as the inflection of a voice may be exquisite joy or suffering to a lover: but it would be insufficient to move the average hearer to any response. The reason is that there is always a dual process at work in mind: there is the sense-perception of the actual sound, and a brain-recognition of its meaning. This latter must be supplied by the hearer himself from his own imagination or experience. The non-musical multitude has neither, and is therefore unable to complete this second process of recognition. Thus the hearer hears, but does not understand. It is probably for some such reason as this that we resort to words to make the message clear. Herein lies the importance of the words themselves, and of the diction of the singer. Quite notoriously, many singers entirely fail to make their words intelligible to the listener, and in the majority of cases this is due to insufficient stressing of the consonants. Vowel tones carry, while consonants do not. If we want to shout to anyone we call out "Hi" or "Hey": never by any chance do we try to reach them with a "P-p-p-p-p" or a "T-t-t-t-t," and for precisely this reason. If, therefore, a singer wishes his words to carry to the end of the hall he must needs exaggerate his consonants to allow for this loss in transit: the vowels will look after themselves. Then, although the balance of the words as they are uttered may be a trifle distorted, they will nevertheless reach the hearers in due proportion. Comfort in listening is greatly increased when this sense-perception is clear and unambiguous, and the brain-recognition is easy by reason of a certain familiarity. When the sense-perception is blurred, as in faulty diction, extra work is thrown on to the brain: listening then becomes a strain, and the brain is fatigued with supplying the details which it supposes the singer to have intended. The listener has, as it were, to put in his consonants for him, to dot his "i's" and cross his "t's."