From Tchaikovsky's harmony textbook, "Guide to the Practical Study of Harmony", page 98:

In the image below, why are the bars contained in the red box of example 258 considered good? The paragraph (Section 88) before example 257 in the image seems to imply that these would be "not good" examples, as the blue highlighted notes in the red box show thirds and fifths violating the above paragraph.

Tchaikovski "Guide to the Practical Study of Harmony", page 98

  • 1
    Oh, how interesting! The third measure of 257, apparently intended to be "not good," is basically exactly the same as the fifth measure of 258, which is clearly labeled "good." Both are first-inversion major triads moving to a 42 chord. The only difference is that the tenor is an octave higher in 258 and the seventh chord is dominant as opposed to minor. Otherwise, they're literally just transpositions of each other. I see no reason why one is "good" and the other not.
    – Richard
    Commented Feb 5, 2021 at 22:20
  • Good question! The G/G♯ and E/E♭ combinations certainly seem to fall foul of the first part of the rule. I see some confusion from Tchaikovsky (or perhaps his editor/translator) here.
    – Laurence
    Commented Feb 5, 2021 at 22:26
  • Ditto @Richard's comment. The only difference between the two examples he mentions are the voicing of the alto and tenor. But the "bad" sound is present in both and for the same reason. It's also very strange that 257.3 has G moving to D-7. That would be a retrograde progression in C major and very odd sounding. It makes more sense if the F is changed to F#, making the D chord a D7.
    – Aaron
    Commented Feb 5, 2021 at 23:29
  • @Aaron And with that F-sharp added, the second example is even closer to the first example! This is really very odd...
    – Richard
    Commented Feb 5, 2021 at 23:49

2 Answers 2


(I deleted my initial answer. Sorry, I missed your whole point!)

IMO lot of old theory books have this kind of unclear language.

One way to make sense of it is the general wording "we should take care not to..." is illustrated with #257 - just to show what we are to be concerned with - then the more specific point is made about bass and upper voices and #258 illustrates good and bad.

Bars 4 and 5 of #258 seem to contradict the "take care not to employ" direction of section 88. The only sense I can make of it is it doesn't say "never use..." It's not an outright prohibition. If I try to make sense of it, I imagine it means something like "take care when combining a pitch in one voice with its chromatic alteration in another voice, don't do it indiscriminately, avoid altering the third or fifth, avoid placing the alteration in the bass." That's my sense of the words in combination with the musical illustrations.

If the wording was...

In order to avoid disagreeable tone-combinations we should take care when employing simultaneously in different voices of a chord a tone with one of its chromatic alterations, especially where that tone forms the third or the fifth of the chord. Specifically a chromatic passing note should not be used in the Bass when the Bass tone is doubled in an upper voice.

...and if #258 bars 4 and 5 were labelled acceptable, I think it would all make sense. #257 merely shows the cross relationship of concern and #258 are specific example of good, acceptable, and bad. (Italics for the wording I changed, which isn't much.)

Keep in mind it's a 1900 translation from German into English. We probably should consider some finer points of wording were not perfectly captured.

  • I think the question isn't about why the 258 examples are good. It's "why are the 258 examples good, while the seemingly equivalent 257 examples are bad?"
    – Aaron
    Commented Feb 5, 2021 at 21:52
  • I agree with @Aaron, they both are really similar, and the only actual difference is in the bass. Commented Feb 5, 2021 at 23:42
  • 1
    I completely missed the point, but wrote an entirely new answer. Thanks for setting me straight! Commented Feb 5, 2021 at 23:54
  • 1
    Considering that it's a very old edition of an old translation from German (and, possibly, a translation itself), i would not exclude errors in the explanation or even in the notation (or its annotations). I have a couple of "old" methods for percussions, they are still used and reprinted nowadays, but they still have a couple of errors. Commented Feb 6, 2021 at 0:26

One reason given for avoiding, not eliminating every instance, of cross relations is that such a relation emphasizes the chromatic tone. Example 258 is pretty clear (what Tchaikovsky thought, anyway); the chromatic tone occurs in the bass in the unboxed measure; in the red boxes, the chromatic tone is in the soprano.

In example 257 (which I answer second as it's taking more thought for me to figure out what's going on), the chromatic tones are in the third or the fifth of the implied harmony. I suppose that having the third as a chromatic passing tone may blur the major or minor quality of that harmony (of course, one may wish to do that.) I'm not so sure about the fifth. Maybe having G-G#-A in the soprano with a C-X-F in the bass would work by making the C a sort of secondary dominant though being an augmented chord, or maybe not.

There are two (at least) cases where chromatic passing tones don't cause problems. One is between sections; the other is between the ascending and descending forms of a minor scale. In the minor case, the two forms are not really chromatic but mutable.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.