To assist readers in profiting from this short essay in the form of a harmony lesson, we have embedded the printed score with a borrowed LDS accompaniment on an inexpressive e-piano. (If you feel the need, take 120 seconds to review “Holiday Harmony Assignment” of December 31st last.)
St. George’s Windsor is the ever-popular Anglican hymn for Thanksgiving Day; it was composed in 1858 by Sir George J. Elvey, organist at the chapel of that name. In the score (I have made a very few small changes) it has a regularly repeating rhythm, with all voices moving together — some would call that “square,” but that’s a term that has many other uses in music. The verse meter is 77.77 D. in hymn-book language. There is a nice distribution of secondary-dominant harmony, which adds tonal breadth over just sixteen bars; the only pitch-class in the chromatic scale that is missing is G sharp, which could be the third of V of iii, but there isn’t any iii chord either (A-C-E, not that we regret it). NB. I have left out a few of the roman-numeral analytical markings to save space; but you can figure out what’s omitted if you care about them. (A dominant seventh in third inversion — the seventh of the chord in the bass, as a passing tone — is a smooth means of moving from V in root position to I in first inversion; you can see that in mm. 9-12.)
There are some good illustrations of cadences and their types. The metric pattern of the hymn as a whole easily suggests four phrases of four bars each, but you can also imagine it, maybe at slower tempo, as eight phrases of two bars each. Except for the second of these two-bar phrases, each one ends on an authentic cadence, V-I in F or “as of” a secondary tonic. Thus at mm. 2 and 4 there is an authentic cadence, V-I (but not “perfect” because the upper voice doesn’t end on the tonic degree). Mm. 1 and 3 are identical, and the melody of mm. 2 and 4 are also identical but m. 4 has different harmony: it ends on a half cadence! (? Yes, on the dominant of vi. This V of vi naturally proceeds to vi in the next phrase.) M. 8 ends (“storms begin”) with what some of us would call an “authentic half cadence” in F, jocularly because that is seemingly a contradiction in terms; but if you consider that mm. 7-8 can be heard in C major, then m. 8 is a strong PAC in that key. Still, mm. 9-10 confirm that we really stayed in F major all along; those two bars are entirely tonic and dominant harmony. Mm. 11-12 are exactly the same but transposed (sequentially) to the B-flat major level, and end on an authentic cadence on B flat (IV). Only in the last measure, m. 16, is there a PAC in the main key, ending with the tonic degree in the melody; the “song of harvest-home” is in the “home” key, after all, and it is in the strongest form, the tonic in second inversion (“har-”) resolving to the dominant (“-vest”) and then “home” again — melodically, and harmonically, more final than with the same word in m. 4.
We’ve observed before that the dominant-tonic relationship, V to I, is what moves harmony forward in time; it’s like in-and-out breathing, inspiration literally, and this holds true for secondary dominants and secondary tonics just as much (e.g., V of ii to ii). Moreover, this inspiratory, oxygenative relationship doesn’t have to be in the authentic cadence at the end of a phrase; it can be anywhere in the phrase, and even between phrases, as we see in St. George’s. So this tells us that “cadence” can be harmonic, but isn’t necessarily harmonic. A cadence can also be rhythmic; it can be metric; it can be melodic; it can join two phrases together rather than ending just one; and (perhaps) there are times when it does not even exist. We will continue to pursue this idea.