I was wondering if a song's key can be declared as just being ambiguous, where you just don't know if it's in major or minor? I think this concept applies only to relative keys as parallel keys have different notes.

I was listening to a podcast in relation to this where they said it is the musical version of an ambiguous image that is open to two interpretations.

I first saw the concept in an article Tonal Pairing and the Relative-Key Paradox in the Music of Elliott Smith. Where the author describes tonal pairing as based "not on one stable sonority, but on the tension between two tonal centers" that is discussed in this article. An example they give in this article is the opening four measures of Johannes Brahms’s Intermezzo, op. 119, no. 1 that cycles between the relative keys of Bm/DMaj.

enter image description here

  • 7
    To say that "a song is in a key" is a shorthand way of saying that we can think about it sensibly from the perspective of that key. The key isn't really a part of song itself. Commented Mar 8, 2020 at 3:34
  • Isn't this the case with all songs that use a power chord structure?
    – Strawberry
    Commented Mar 8, 2020 at 12:44
  • I regularly compose music without a key listed - technically setting everything to C major so GarageBand doesn't adjust the frequency of any notes automatically. Commented Mar 8, 2020 at 13:11
  • The first few measures of Schubert's G major quartet alternate between G major and g minor. I'm not sure you really can tell the key until after the opening. Commented Mar 9, 2020 at 5:06
  • 1
    If I understand the detail of the question, this isn't two keys at the same time it's alternating keys. Commented Mar 10, 2020 at 19:19

7 Answers 7


A 'key' is a generalisation of the notes used in any piece. It's a guide to what will appear mostly. If a piece has two sharps in its key signature, they will most likely be F♯ and C♯. We then acknowledge that it is in the key of D major and/or B minor. Perhaps a guide for musicians.

Occasionally, that is taken further, and it could just be that it's in E Dorian, F♯ Phrygian or A Mixolydian, for example.

There are many times when the key of a particular piece changes - modulates - to its parallel key or the key of its dominant, or even subdominant. That's when the word key gets messed about. What we can't say is that a piece with those two sharps is suddenly in key Em for three or four bars in the middle somewhere.

My students are given the impression that home is a key. All journeys start from there (bit like real life!) and only when returned there more or less permanently is the journey over. We may have visited other places on the way, and stayed there temporarily (a holiday?) but generally speaking, we return to the static residence - home. Occasionally, we may decide somewhere else is a better place to be permanently, so we move house and that becomes the new home - thus a new key.

Moving to and from relative keys is commonplace - there's often a Bm section in those pieces which start in key D. Moving to parallel keys is rarer, but will still happen, quite easily as the same root note is familiar to the listener. Much was made of that a couple of centuries ago with the Tierce de Picardie - although one wouldn't call that a key change!

Thing is, the harmonies for both - relatively speaking - are the same, basically. Triads in said D (Bm) key are D maj, G maj. and A maj., and Bm, Em , F♯m. Not forgetting C♯o. All of which get used in both D major and Bm. Inextricably linked,.

The question reminds me of - 'Can you be a father and a son at the same time?'

  • 5
    Your reference to 'Can you be a father and a son at the same time?' reminds me of the riddle that generally goes something like this: "2 fathers & 2 sons fish up 3 fish and split them evenly between themselves without cutting any fish apart. How does that happen?" The answer is that there's a grandfather, his son, and his son's son, and this takes advantage of people generally assuming that a father and a son won't be the same person.
    – Dekkadeci
    Commented Mar 8, 2020 at 9:33
  • @Dekkadeci - that's where it came from. OP seems to be of the opinion that music can't be major and minor in the same piece, I believe.
    – Tim
    Commented Mar 8, 2020 at 9:59
  • I always used the comparison from starting at home and leaving away and coming back home. Only much later I found out that this feeling of coming home is very common spread. Commented Mar 9, 2020 at 13:58

What does it mean to be "in a key"? That's really at the heart of this question.

The first question I'd ask in response is, "Why do you care? What are you trying to determine by saying piece X is in key Y?"

Some people are looking for a simple answer, often determined by the key signature and what local key/chord progression a piece ends on. They will just declare that to be the key and be done with it.

Other people are interested in something more nuanced. One might ask what local key a piece is in at a particular point in time. Some of those people might look at the local context, including accidentals, cadences, etc. and again just declare on the basis of the score that "In measure X, the piece is in key Y." If there are some conflicting factors, they will weigh various parameters and make a determination based on a global view of the score, as if they were solving a puzzle. This is the type of "analysis" frequently done in an introductory music theory course, where it seems like the goal is often to label a particular place with a Roman numeral in a particular key.

Of course, sometimes the ambiguity gets stronger. Some pieces aren't even tonal at all. But assuming you're working within the general framework of tonal harmony, we might ask as a listener, "How do I know what key I'm in at point X?"

And that's a much harder question. It depends partly on your priorities in analysis. Some people may just look at the local scale and declare an answer based on that. Certainly if you're trying to determine whether you're in D major or B minor in a classical piece, the use of A-sharp as leading tone will often determine the key (i.e., B minor if A-sharp is frequently present and used as leading tone; D major if not). Local cadences, points of rest, and emphasized harmonies might also give clues.

But that may still not be enough. To go to the Brahms example in the question, the beginnings of pieces of music in particular may often be quite ambiguous. Suppose you place a single pitch D, maybe in multiple octaves. What key are you in? D major? B minor? What about G major, if D is the dominant? Or what if the notes that coalesce around D are actually B-flat and then F and A-flat, and suddenly you resolve to E-flat, which is your key? Retrospectively, you realize that D you started on was the leading tone.

But you can't know any of these things with a single pitch D sounded. In that sense, you might spend a few notes (or a few bars or even longer) at the beginning of a piece not actually knowing as a listener what key you're going to end up in. Many composers actually seem to enjoy creating sort of "fake-outs" at the beginning of pieces, and this became more common in the 19th century.

In that sense, it's very easy to say that a piece can be in an ambiguous key for a while. If you have a score-based perspective on analysis, you may look ahead and say, "Oh, obviously that D is a leading tone, which you can see four bars later." But as a listener you don't yet have that information.

I'd compare the beginning of a piece of music to the creation of a "world." The composer can give you clues about the rules of that world -- what it looks like, what places you can go, what places you are unlikely to go, what the rules are about how this world operates (is there gravity? which way does it pull?), etc. These are all expectations set up not only by key, but also by instrumentation, texture, dynamics, articulations, etc. To use an analogy, it can be like being dropped down by a video game into a new world with a slow graphics card. In some simple cases, everything appears at once, and it's very clear what the world looks like. In other cases, you drop in with a field of white around you -- and then gradually out there you see a splotch of green, and then there's a tree. Later you hear a trickle of something, and a stream appears. As the field of view fills in, you get a sense of what the landscape is. And the you can begin (as Tim says) your journey in this musical world.

To be "in a key" as a listener in an emerging world usually means that there are signposts or pointers that designate a particular note as having greater importance or "home." A stereotypical gesture is a leading tone resolution or a V-I cadence: if you have one of those, it's like a bright red flashing arrow pointing at a note in your world and saying, "There! That's home!"

But then you look at a piece like the opening of the Brahms Intermezzo. The first note is F-sharp. (The splotch of green.) Are we in F-sharp? No, a few notes later we've outlined a descending B minor triad. (A tree appears.) Are we in B minor? Well, the next notes might be an E minor triad, which could be iv. But in the next bar we get an A7-D implication. (The stream has emerged.) We might make a new guess: perhaps the piece opened on vi, and we just heard a circle-of-fifths sequence (vi-ii-V-I in D) that points us to D major. But rather than a bright red arrow cadence pointing us to D major, we don't get that strong chordal bass motion of a 5-1 leap, just a hint at V7-I in D major.

And the next bar starts doing some weird stuff: F-sharp minor? In D major, that's an unusual iii chord. In B minor, that would be an even more unusual minor v chord (for classical harmony). What the heck is that? It's like our world opened with some trees and a babbling brook, and suddenly a giant jellyfish dropped out of the sky and landed with a plop in this forest landscape. What's going on?

The only way to know is to "stay tuned" and try to make sense of the developing tonal landscape. At this point, as a listener, if you asked me what key I'm in, I'd say, "I don't know. Could be B minor. Could be D major. Definitely not clear." In essence, ambiguous. Key choices need not only be ambiguous between relative major and minor. As noted above, a limited set of notes may not make clear what key you are in for a while. The opening B minor triad by itself could be a i in B minor or a vi in D major or a ii in A minor or a iv in F-sharp minor or even something more exotic. Only further context can determine that. And sometimes the context can be ambiguous for quite some time, or even for an entire piece. Brahms continues the play with listener expectations for quite some time in this particular piece, introducing various chromatic bits in the subsequent measures that twist and turn and hint at unseen destinations within the dark forest.

Some people resist this idea of ambiguity, generally because some early experience with music theory has taught them that there's supposed to be "one right answer" for an analysis. Introductory exercises in theory often are chosen specifically because they lack ambiguity and are easy to understand (and easy to grade). Thus, students get a false sense that a piece must be in a particular key, and a particular chord must have a particular function in a particular key. Some people might therefore insist that this Brahms opening is NOT "ambiguous," but rather oscillating between keys or something. Maybe. That's one interpretation. But if you believe in the concept of tonicization (i.e., a temporary move toward a non-tonic note, without changing key) or even just an idea of applied/secondary dominants, then one could easily analyze a lot of this piece in at least two different keys. Which one is the "true" governing tonic? Does one have to oscillate or move between keys? The real listening experience of music, particularly for a piece like this Brahms Intermezzo is not always that determined. The ground in the landscape may never seem firm and always be shifting under your feet.

So yes, in my opinion, it's very possible for segments of a piece and even an entire piece to have an ambiguous sense of key. For other analysts who want a "one true answer" and a label to put under every chord to "solve the crossword puzzle" (as I generally think of Roman numeral "analysis"), they might want to weigh the various options and argue for a particular tonic at any given moment. If you forced me to, I could make arguments for what Roman numeral to put under every chord in this piece and what local key you're in. But personally, I'm often happier just to live in the shifting tonal landscapes as the stream babbles on and leaves blow, wafting wistful strains in various directions accompanying me on a journey that may or may not ultimately take me to a "home"....


Any song may begin in a major key or a minor key or in a mode such as the Dorian mode. If song starts in a major key it can modulate to a minor key.A pivot chord can also be used to change from a major key to a minor key. Some keys do not have a leading note one semi-tone below the tonic note , they have a sub-tonic note a full step below the tonic and may sound more ambiguous. The natural minor keys;for example F minor (natural form ) has a sub-tonic chord E flat = VII , not vii° triad. Many of J.S. Bach's pieces modulate through closely related keys, but they usually end with an Authentic Cadence in the Home key (where it began ). Example is Menuet in E major by J.S. Bach (from French Suite No.6 in E Major , BWV 817.)


It is possible for a song or portion thereof to be in an ambiguous key. Consider the chorus to the song "Hotel California" by the Eagles. The verse is rather clearly in B minor, with the chord sequence Bm F# A E G D Em F# (i V VII IV VI III iv V). Most non-tonic chords are major, but that's actually fairly common when using minor keys, and nothing in the melody would suggest a change of key. Once the chorus arises, however, things become ambiguous.

The first two chords of the chorus are G and D, which would work equally well in B minor (VI III) and D major (IV I), and the melody sounds like it shifts to D major. Indeed, all but the last two notes of the chorus melody would work just fine in D major. The third and fourth chords are a V-i cadence in Bm which establish that the song never really left that key, but if they were changed to A D they would instead establish that the song had modulated to D major. If the ending of the chorus were then left as it was, it would be clear that the chorus had shifted from D major to B minor, but it wouldn't be clear where that would have happened.

  • The one that gets me is' Unforgettable'.From memory, starts in key C but ends in key G. And I can't find where the modulation occurs. And it all sounds right - as if it's really in key C at the start, and really in key G at the end.
    – Tim
    Commented Mar 9, 2020 at 7:34
  • @Tim: El Shaddai sounds like it starts in E minor, but somehow ends up in D major, and the Theme from Arthur has basically the same chord sequence, but ends up elsewhere.
    – supercat
    Commented Mar 9, 2020 at 14:48
  • 'Fly me to the Moon' - Am (1st chord) or C maj. (last chord)? And does anyone really care..?
    – Tim
    Commented Mar 9, 2020 at 16:56

I'd say yes, but beware: the idea of a key center is a very human element of music. In real life, there's often no empirical way to say "yep, that song's definitely in C♯ minor, not E major". While in some cases it can be black-and-white, the key of a piece can also be highly subjective and up for interpretation.

Interestingly, one consequence of this observation is that the composer and the audience may have different interpretations of the key of a piece, just as they may differ in their other interpretations of elements of the music. Personally, when I look back at some of my pieces I wrote before I had a very solid grasp on the idea of what a key actually means, I find that I now hear the music as being in an entirely different key than what my hastily-chosen key signature would tell me.

But ultimately, the key or tonal center of a piece is just a construct. It represents a state of stability or common return, and this can be interpreted many ways. Many pieces are not content to stay in one clear set of seven notes, and when they depart from diatonicism, it can be useful to conceptualise a piece as having multiple or fluid tonal centers.

This is apparent from the very idea of a modulation itself. When you change from one key to another over time, it makes sense for the listener to redefine their idea of the tonal center in their mind. If I start a pop song in C major, then pull a truck-driver-style modulation up to D♭ major, no listener will hear the second half of the song as being all relative to the previous tonal center of C. We musicians use the idea of a key center as a way to orient ourselves in an otherwise potentially confusing harmonic landscape, and the best of us can easily redefine this coordinate system on the fly.

The whole genre of the blues is often cited as an excellent example of fluid tonality in music: Elements of major and minor run parallel to each other in the blues - take a look at the common 12-bar blues format. The IV7 chord is a dominant seventh chord, but its seventh is directly contradicted by the V7 chord's third a half-step away. Now, personally, I like to think of blues as its own tonality, but it can be very useful to recognise the elements of major and minor represented in this harmonic landscape.

A lot of pop music is written in such a way as to spark debate on its key. We all know the I-V-vi-IV chord progression, but some songs seem to use the same chords but base themselves around the vi chord. Should we then redefine our coordinate system to emphasize minor? Sometimes it makes sense to do so sometimes it doesn't, but a large amount of pop music exists somewhere between those two extremes. "Viva la Vida" by Coldplay is often quoted as being in A♭ major, but also as in F minor. There have been some interesting academic arguments on this dual sense of the key center in popular music: all-around musical smart guy Adam Neely (I reference his videos a lot, actually, and he's great with theory) has a video essay on this in his discussion of "Sweet Home Alabama" and its potentially disorienting tonality where he gets into a lot of what I've already discussed.

And of course, since we're talking about keys, everyone's probably been waiting for someone to mention "Giant Steps" and "4:33". Think about what the idea of a key center is in the context of those two extreme examples. What's amazing to me is that skilled composers can use chromaticism (in this case, any departures from diatonicism) to set up a continual oscillation between two or more key centers often so subtly as to escape detection by a casual listener.


I remember that this question araised several times in the last months around a song e.g. Am F C G ... or a similar progression: Is this iv - VI - I - V? or is it i IV iii bVII???

This shows the ambiguity of the function and you can here both Do or La as tonal center. That's why they are called related keys and related chords. In German we use the term "Parallel Key" and "Parallel chord" *) ... this also demonstrates the parallelism of the two kind of harmonic centers and its ambiguity.

*) I know the parallel key in English is what we call the Variante in German: Cm and C, Am and A etc.

A la Turca or für Elise are just 2 other examples of famous pieces that are modulating in both "modes" (minor and major). Also many Sonatas and Symphonies begin in minor and the secondary theme (subject group) is in major (or in contrary: primary theme major and secondary theme minor - if it is not modulating to the dominant) like e.g. the 5th Symphony by Beethoven.

The point is: we define the key from the opening **) bars and the final chord - and also from the dominating passage with the longer durance in a certain key.

**) while there is to say that many songs even don't begin on the 1st degree - as Tim says in his comment Fly me to the moon: vi ii V I) there are other examples that begin with ii-V7-I (and this is like an "up beat" bars motif, but then in the first bars the melody is "cadencing" to the tonic.


Best example I can think of is Benjamin Britten's Fanfare for St Edmundsbury for three trumpets. It's not a "song" but it demonstrates the point in your question.

Each of the three trumpet parts is in a different key and time signature. Each trumpet plays a section on its own at the start, and then all three play together, still in three different keys. Such is Britten's genius that he makes this work. Clever stuff.

Your Answer

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