Modern harmonies are made theoretically much more difficult than they need be by our system of notation, which grew up in the Middle Ages. The old modes knew no modulation in our sense, and in the seventeenth century, when the tempered system came into vogue, making every kind of modulation possible, the old notation was retained. How unsuited it is for modern music appears from the drastic contradictions which it involves. It is quite a common thing to see the same note simultaneously written as F sharp in one part and as G flat in another. This is what makes modern harmony seem so much more difficult than it really is, for when the music comes to be heard, these formidable-looking intervals resolve themselves into something quite natural and generally not difficult of apprehension by a musical ear. Unfortunately we are compelled to learn music through the medium of a keyed instrument, generally through the most unmusical of instruments, the piano, and we learn theory largely through the eye and the reason instead of through the ear. The problems of harmony will seem much simpler if we remember that its basis is the interval--music does not know "notes" as such, but only intervals--that the number of possible intervals is very small and their relations quite simple, and that everything which is not reducible to a very simple vulgar fraction is heard, not as a harmony, but as a passing note, an inflection of a note of a chord. In fact the advance made in chord combinations since the introduction of the tempered system is not very great. All, or nearly all, the chords used by Wagner are to be found in the works of Bach. The suggestion to explain Wagner's harmonies by assuming a "chromatic scale" rests upon a misapprehension of the nature of a scale. Every scale implies a tonality, i.e. a tonic note, to which all the other notes bear some definite numerical relation. There cannot be a chromatic scale in the scientific sense in music; what we call by that name in a keyed instrument is merely a diatonic scale with the intervals filled in; it always belongs to a definite key, and the accidentals are only passing notes. It is in passing notes that we must seek the key to Wagner's harmonies. With Wagner more than with any other composer since Bach the parts must be read horizontally as well as vertically. As long as we look upon harmonic progressions as vertical columns of chords following one upon the other we may indeed explain, but we shall never understand them. Each chord must be viewed as the result of the confluence of all the separate voices moving harmoniously together. This, too, will help us to grasp the character of "altered" chords, so lavishly employed by Wagner, and of "inflection," by which term I mean to denote all kinds of passing notes, appoggiaturas, suspensions, changing notes, and the like. All are phenomena of harmonic notes striving melodically onwards, either upwards or downwards.
Although little has been done in the invention of new combinations, the character of the harmonic structure has changed considerably since the eighteenth and early nineteenth centuries. This is evident at first glance on comparing a score of Haydn or Mozart with one of Wagner or Liszt. There, although chromatic harmonies are not unfrequent, they occur only sporadically, the general structure being diatonic, whereas with the later masters the whole tissue is chromatic; the score fairly bristles with accidentals, and a simple major or minor triad is the exception. Very different too is the periodic structure. The phrases no longer fall naturally into eight-bar periods interpunctuated with cadences, but are determined by the text, and although the eight-bar scheme is generally maintained--much disguised, it is true, but still recognizable--it is determined not by half-closes at the sections, but by the eight beats of the two-line metre, while the periods follow each other in even flow without any indication of cadence. In other words the musical form is governed by the declamation.
Theory of harmony is one thing, living music quite another. The musical hearer of a work like Tristan und Isolde will understand its harmonic structure, though he know nothing of the theoretical progression of the chords, provided the performance be good, i.e. correct, just as a man ignorant of grammar will understand a sentence which is clearly enunciated. The composer needs no theory of harmony; his ear is his only guide, as the eye of the artist is a sufficient guide for his colouring without any theory of colour. There is only one thing which the composer must keep before him and which the hearer must consciously be able to recognize--the Tonality. The problem of harmony therefore in practice reduces itself to that of modulation. To recognize the tonality quickly and certainly, look for the cadences. They are as it were landmarks, placed along the melodic road, indicating from time to time where we are.
I cannot dismiss the subject of harmony without mentioning the chord which from its employment at decisive moments and its extraordinary mystic expressiveness has been called the soul of the Tristan music. Its direct form is
as it occurs in the beginning of the Prelude.
The instrumentation of Tristan does not present any special features different from that of Wagner's other works. It is less heavily scored than the Ring, and at the same time the instrumentation is more concentrated. Wagner usually employs his wind in groups of at least three in each colour--e.g. three flutes, two oboi and one English horn, two clarinets and a bass clarinet, etc.--and so is able to keep his colours pure. It is partly to this that the extraordinary purity of his tone in the tutti is due, partly also to the sonority imparted to the brass by means of the bass tuba, and still more to the consummate skill of the composer in the distribution of his parts.
There is an interesting note at the beginning of the score in which the composer seems to be trying to excuse himself for using valve instruments in the horns. While admitting the degradation of tone and loss of the power of soft binding resulting from the use of valves, he thinks that the innovation (which I need scarcely observe is not his) is justified by the advantage gained in greater freedom of movement. In such matters one must be allowed to form one's own judgment, and though it may seem like trying to teach a fish to swim, a humble amateur may be permitted to wish that Wagner had here resisted the tide of progress. It is not only that the tone and power of binding are injured, but the whole character of horns and trumpets is altered when they are expected to sing chromatic passages like the violin and the clarinet. As the point is of some interest, I should like to bring it before the reader with some examples. The essential character of the horns is nowhere more truly conveyed than in the soft passage near the beginning of the overture to Der Freischütz, and it is the contrast between the two nature scales on the C horn and the F horn which gives the character to this lovely idyll. The trumpets are capable of even less variety of expression than the horns, as their individuality is even more strongly marked. How entirely that character is conditioned by the mechanism of the instrument may be illustrated by an example. The third movement of Beethoven's seventh symphony contains an interlude molto meno mosso. The choral theme is accompanied by a continuous A, sustained in octave in the violins, which in the intervals between the verses descends to G sharp and returns
The repeat at the end enters ff. after a strong crescendo, and at this point the sustained A is taken over from the violins by the trumpets and given forth with piercing distinctness above the tutti of the orchestra, the effect being one of extraordinary brilliancy. Now comes the point with which we are concerned. In the intervals the trumpet cannot descend to G sharp, because it has not got the note in its natural scale, and is therefore obliged to repeat