I noticed while playing guitar recently that the chords E, D, C#, B, and then E, D, C#, G sound good together although to my knowledge those chords can’t be in the same key because of the E and the C#. What key is this in?
Why does it sound good? In this case, it just does. Perhaps we could construct an argument around the root movement, or around parallel motion between voices creating a pattern the ear can follow. Regardless, as you suspected, it's not in a key. Here's why.
If you lay out all the pitches involved, they constitute nearly the entire chromatic scale, excepting only C and A#/Bb.
Here are all of the chords.
X:0 K:none M:none L:1 "E"[E^GB] "D"[D^FA] "C#"[^C^E^G] "B"[B,^D^F] "G"[GBd]
And here are all of the individual pitches (E# = F), in ascending order.
X:0 K:none M:none L:1 B,^CD^DEF^FG^GA
So that eliminates there being only one (diatonic) scale involved. Nevertheless, the song as a whole, could still be in a key. To be "in a key" means two things:
- There is a tonal center (i.e., the "key")
- There is a primary mode (i.e., scale)
A piece in the key of X need not stay in that key the entire time. It's simply that the construction of the piece leads the ear in the overall direction of X.
In the case of your chord progression, one could make a reasonable case that it's centered around E (the pitch, not the chord), as each half is rooted in that pitch, and you could add an E chord at the end to round things out.
However, it can't be said that there's a primary mode. Essentially every chord changes the pitch center and associated scale, so the ear never creates mode-like ("functional") relationships between the chords. That lack of clear modal commonality between the various chords is what prevents interpreting the piece as being "in a key".
Note-wise, all the roots of all the chords come from E and its parallel key, E minor.
It's quite common to 'borrow' from parallel keys - pop music does it frequently.
It's not a given that every piece has to be in a key. There are many atonal pieces which obviously aren't. That's why they're atonal. However, the vast majority of pieces are in a key - or perhaps two or three keys - modulating between them.
So whilst it's unusual, it happens.
There's no 'rule' that all chords must be diatonic - in fact, in such music, there's often so much expectation from the listener, that it is predictable, with no nice surprises, which are what makes some music exciting.
So, if we were charged with putting a key signature on this tune, where do we go? As already said, the chords are from E - major and minor. It will depend more on what the melody consists of for evidence of a key, and that's denied at the moment. It could easily be a blues type number, which breaks the 'rules' anyway, often written in major, but using minor notes (and that ♭5), so, if pushed, I'd go for writing it in E - E major.
The home chord seems to be E major.
We could talk of 'borrowing'. D and G are not diatonic in E major. They are in E minor. OK, that grants us permission to use them. (But we have 'permission' to use any chord, there's no 'must be diatonic' rule in music. I guess finding a 'borrowing' connection tells us these 'outside' notes aren't TOO outside.)
Why did you choose to use THESE particular non-diatonic chords? Maybe because so much of today's popular music derives from the Blues. You've picked chords rooted on two of the most characteristic notes in the Blues, the ♭3 and ♭7. That could be a lot of why you found those chords attractive.
The C♯ chord is a teaser. That's a bit more 'outside' E major. But a bass line that walks by step is a very strong musical element. And a string of adjacent chords of the same shape (these are all major triads) has unity. Try C♯m instead. That fits too, in a 'keep it diatonic' sort of way.
(See what's happening? We're finding excuses for the G, D and C♯ chords to exist in a basically E major environment. But there's no causation involved. There's no place where one of those SHOULD be the next chord, where we'd be surprised if it wasn't. Not like the expectation of a tonic E after a dominant B7.)
Right. You've written something that sounds good. We've established that E major will do as the overall key and given you a selection of justifications for the non-diatonic chords. Put them in your bag of 'good musical things to do'. I think that's all sorted?
To me it sounded like being "in E", but ultimately, it is whatever key you can make it be in. The first chord is always very important, because it makes the first impression and starts to establish things. The next chord will be compared to what happened before it. The order of things that happen and my immediate reactions to them are, along with scales that are easy to imagine being suitable in each situation:
- E ... Ok, this is in E. Scale: E major.
- D ... oh, it's E Mixolydian style, probably A is the next chord. Scale: E mixolydian (or A major).
- C# ... a little surprise - it didn't go to A, but it's still in E or C#m (relative key), but secondary dominant going to maybe F#m next. Scale: F# melodic minor.
- B ... another surprise, no F#m but it got back to E - this is V chord of E. Scale: E major.
then the next round:
- E ... yeah, we got back to home chord
- D ... ok, nice, I know this already
- C# ... ok, nice, I know this already
- G ... nice surprise! jazz trick, modal interchange, switching between E major and E minor. Scale: G major / E natural minor
All of this goes nicely to an "in E" category in my mind.
Why it's so good is, it has a good combination of familiar and surprising things, which makes it interesting. Staying strictly inside one single key and mode tends to be boring!
Here's a little 80s pop melody over riding over the chords:
The melody jump from the C# chord to the B chord is A - D#, which makes the B chord a B7 really, strongly suggesting that the next chord will be the assumed home chord E. Which it is.
In the end there's an extra surprise - it ends in B. :)
Short answer: E major.
To me this sounds like a rock progression borrowing from Coltrane. ‘Trane’s harmonies lean on chord substitutions based on minor third intervals. So for example, if you liberally substitute chords removed by a minor third, E D C# B, E D C# B is a jazz(ish) substitution for E B E B, E B E E. Simple as anything.
If you’re interested in those minor third substitutions (and who wouldn’t be; they can really spice up an otherwise insipid, tired rock trope), dive way into Giant Steps. It’s a seemingly incoherent progression until you grok Coltrane’s minor third substitutions. With that concept, they’re absolutely sensible.
I have lost track of it but somewhere there’s a book named something like Minor Thirds: the Harmonies of John Coltrane. I heard of it in a jazz seminar at Stanford around 1990. If anybody can find a link to that, please let me know!