[autarch]

I came here to pick this nit but you beat me to it. I think calling it Db7 is the best way to think of it, especially since in jazz it's often part of a descending sequence of II-V-I, where the tritone substitution turns it into II-bII-I. So in C major (where G is the dominant), it has to be Db.

[mkbosmans]

Why do you say a tritone substitution turns it in a II-bII-I? Can it be as easily said to be II-#I-I? In that case it would be C#.

[colanderman]

Enharmonic equivalents are usually chosen to match the direction of movement. So a descending passage will prefer to use flats.

[brudgers]

The root is the root.

There can be only I.

[autarch]

I don't think this is correct. I think colanderman's answer is the right one, which is that when we're descending, we generally spell out the pitches with flats. If this were ascending, then we could have I-#I-II. But as I noted, in jazz this particular descending pattern is fairly common.

[brudgers]

spell out the pitches

I, bII, etc. are chords not pitches. II-bII-I is a chord change. Pitches may go up or down and voices may go away or arrive. They may be part of those chords or not. And there may not even be a root pitch.

Or if you want formal theory, chords are scales [1] and there ain't no such thing as a scale with a sharp I. If you sharp the one it’s a different scale starting in a different place.

The b in bII is not an accidental and bII is how to communicate the chord in passing. I/bII is also possible to communicate a modulation (as would be vii/II, etc. if the modulation is elsewhere).

There are “enharmonics” for chords that aren”t the root of course, e.g. #IV and bV depending on intent, convention, or playability. But #I will be marked wrong on the entrance exam and only raise you standing among an avant guard that isnt accepting new members.

[1] see Russell’s Lydian Chromatic Concept.

[bonzini]

> If you sharp the one it’s a different scale starting in a different place.

It depends on what's the function. If you want to use the tritone-substituted G7 to modulate to the sharp fourth, you'd write it C#7 and resolve it to F# (or F#-).

[brudgers]

C#7 is the name of a chord not a pitch and not Roman Numeral Analysis. [0]

The Roman Numeral analysis would be V7/#IV if you are modulating in C. And something else if you aren't in C. Roman Numeral Analysis is independent of key or tonal center. The Roman Numeral Analysis would be bII if you are not modulating because that's how Roman Numeral Analysis communicates.

Thinking about Phrygian mode. The two is always a bII.

{0} As an aside, in The Common Practice the bII is often notated N6 (for Neapolitan 6th) to set up a perfect cadence, i.e. I-N6-V-I.