Nonchord tone (Wikpedia) is one of sources I've learned it from.

I'm analyzing nonchord tones on Charlie Puth's song, Attention, to get used to them. However, some of them, I call "unknown" nonchord tone, are strange to me, they make no link from what I read online. For example:

  • the 1st looks like a passing tone, but it doesn't resolve stepwise to the next chordtone
  • The 2sd seems to be an escape tone, but it resolves to a nonchord tone instead of a chordtone
  • With the 4th I have no idea
  • The 5th could be understandable if there's an Anticipation or the first phase of Suspension happen right before it, but nothing goes as I expect it to
  • The 6th looks like a decorated suspension or anticipation, but it doesn't have a preparation phase
  • the 7th might be a passing tone, but after it there must've been a chord tone, not a nonchord tone like the 8th

At first, I thought some of these nonchord tone are used to form a new chord, but then in other measures the composer use another new nonchord tone. Then I asked myself if the chord progression continuously change to harmonize those "unknnown" nonchord tone. [For example, the first bar (of some other songs) starts with Am triad and it has an added "unknown" nonchord tone D in it; then the 5th bar which repeat Am chord doesn't have D but has an added "unknown" G in it]. Well, the changing-chord-progression stuff like this doesn't sound true, right?

Is there any technique for writing these "unknown" nonchord tones?

And Could you guys explain every single one of these strange nonchord tones? How does it work? Why are they used? and Do they have any relationship with the chord?

Charlie Puth Attention Score

  • 5
    There's more to music than harmony. There is e.g. also melody. These notes may be there for melodic reasons. Commented May 4, 2019 at 15:25
  • To expand on @YourUncleBob's comment, don't think a composition is a music theory exam. In a composition, if it sounds good, you can keep it Commented May 4, 2019 at 22:05
  • 2
    These are not non- chord tones, these are non-tone chords. But they sound great. Commented May 5, 2019 at 4:19
  • 1
    The notes are not advanced or mostly not even "non-chord", for example the first one in practice makes the minor chord a minor 7th. Any note that you play will contribute to the harmony in some way. Chord symbols are used as hints for accompaniment, they are not a complete description of harmony. I have a bit of a hard time seeing what kind of a practical purpose would need such analysis of such a very simple pop song. You should spend 10x more time playing and practicing than theoretic analysis. Practice is under-rated. :) Commented May 5, 2019 at 15:42
  • 1
    Unrelated: Measure 15's beaming is absolutely revolting. Sight-reading that would be horrible! If OP is the one who transcribed this, I would advise changing that.
    – user45266
    Commented May 6, 2019 at 1:15

4 Answers 4


It's not uncommon for melodies to include notes that aren't in the chord, especially on weak beats of the bar. The article linked is not really a good way to understand popular music or jazz music (and my common practice music theory isn't really good enough to give an appraisal of it's usefulness to analyse classical music in an explanatory way). The thing is 20th century popular music (especially post 1950s) has incorporated a whole new harmonic and melodic vocabulary that comes largely from the blues, but then also integrated that into the existing western harmonic framework, and so while more conventional "classical" explanations of chord functions are relevant in many cases, there is also a whole other set of vocabulary that completely breaks that down.

But the melody you linked contains nothing particularly weird (other than being in a horrible key to read). It's really normal in modern music to have melodies that have non-chord tones on a weak beat, and which resolves to chord tones. In classical terminology, a "passing note" has to resolve to the next closest chord tone, but in real music, melodies often jump around to a different chord tone. If you transcribe a bunch of pop/rock/blues/funk/soul melodies you will find this happens all the time. As a non-classical theorist it's easy to forget that this was ever considered unorthodox.

Let's answer your specifics before making a more general point though:

1) So the first bar goes 1 b7 5. That's a really really common melodic shape*, you'll hear that everywhere. You can think of it as forming part of a minor 7th arpeggio if you want (or dom 7 in a major 3rd context), but it's really not necessary to think of it like that. I guess that melodic pattern probably has it's origins in the blues, but now it's just been assimilated into pop vocabulary. Similarly you often find 5 b7 1 on the way up (think "my momma said to leave you alone" in "chain of fools" for a memorable example). So you can think of the b7 as a "blue note" if you want, and in a major key it really functions like that (just like the b5 or b3 also do in blues-influenced major key songs)

In a minor key though you can also think of it harmonically, as the 7th is also just valid chord extension for a resolved minor chord. In jazz music you'll hear even more "spicy" chord extensions being sung/played over minor chords, too, the 9th for example is a "clash" or "tension" in classical theory, but in jazz (and therefore by extension, sometimes in pop), the 9th is often just considered an "interesting colour" over a minor chord, and you can just leave it there to hang out. Because of the way this particulary b7 is integrated into the melody, I would interpret this as a "melodic note" rather than analysing it as a "chord extention", but in another context, you could equally hear a flat 7 (or a 9 even) being sung in a strong position over a minor chord, and there's nothing wrong with that.

https://instaud.io/3DFU singing 1s, 7ths, and 9ths over a tonic to show you what I mean.

3) that isn't a really an accurate transcription of the melody, it's really just a "fall", a low semi-spoken note. But the transcriber had a to choose a note so they went with a the tonic. I'd have probably written it as a C# if pressed to give it a note, but honestly it'd be best written with an X notehead with a fall after it.

4 & 5 ) What we have here is a "recontextualisation" of the original melody. You see this a lot in pop music, the same melody repeated over a new chord to give the notes new context, it's a useful device. Since none of the melodies notes clash too much with the new chord, it works well. As it happens, it's not a bad fit for the new chord anyway, the D# just becomes a weak tension that then resolves stepwise to the C#; a chord tone. But even if that wasn't the case, our ear is much more forgiving of non chord tones when they occur in a repeated melody from previously, because the "echo" of that melody is still in our head and so we sort of enjoy the tension between the familiar melody and its new context. This is particularly effective when the melody implies the tonic chord, because we then sort of hear the melody "pulling us home", by implying the home chord while another one is being played.

A good example of this the use of the "I wanna run away" and "You and I" melodies in the Galantis "Runaway", they serve to essentially "pull us home" to the tonic while the b6 and b7 chord are being blasted.

6) See 1), minor 7th is fair game as a chord extention in jazz, soul, R&B etc. That said, is does actually resolve up to the A#, so I don't see the issue. The drop to the 6 is just a bit of "ornamentation" on the 7th, just adds a bit of interest to give a bit of movement. You could equally sing the the F# as a G# or ornament the G# by going up to an A# instead of down to the F#. That said, I think it's better as the F#, the melodic shape is better with "you were mine" as a the highest note, it gives it more impact.

7 & 8) that little melodic figure just anticipates going to the tonic. It sort of implies like "F# G# A# (D#)" but you don't need to actually sing the D, it's kinda "implied". In classical terms it sort of functions like an appogiatura, or maybe a suspension, I forget which one is which, the one where a note from the previous chord is played before the chord gets there (I'll look this up and edit the answer in a minute ok).

Similar little riffs can do the same job. Here it was "F# G# A#", but it could have been (as straight 16ths so I don't have to write rhythm) "E# F# G# A#" "B A# G# F#" or "C# A# G# F#" or even "G#/A\G# F#" for a more bluesy feel (where the / indicates a bend, or microtonal inflection). It's just a little piece of "musical punctuation" that says "hey we're about to go back home".

As a more general endnote, because blues, rock, and soul music has been less formally studied and written about by academic music theorists, and because there's so much variety in how much different songs and musicians draw from each tradition, and because, unlike with classical musicians, most of the most innovative and significant musicians did their work outside the academy and so weren't writing or lecturing about it (and mostly weren't using the same musical language as classical musicians, a lot of the time didn't read or write music for example) there is much less of an accepted way to talk about it. Classical theorists sometimes have the annoying habit of simply saying "well pop music breaks the rules" and leaving it at that, which is essentially just giving up understanding most relevant music written after about 1920. And a lot of musicians/songwriters come at it from the angle of "what sounds good, is good, theory is for nerds, don't think about it too hard, use your ears", which I have a lot of respect for (well, especially the "use your ears" part), but if you're someone like me (and a fair few others on this stack) who likes to really understand how the music they like works, then you're going to have to come up with your own syncretic framework that kind of incorporates classical and jazz theory knowledge, but also incorporates your knowledge of how harmony and melody function in the songs that you like (which probably has to include the blues if you're really going to "get it" imo), and bits and pieces you've heard about and read about on the internet. And then you take that, and mix it into 1 big soup of "post 20th century music theory" I guess.

For an example, see this question here , analysing how bVII chords function in rock music. That is an example of functional harmony, (none of this "oh it's nonfunctional dw about it" nonsense) that nevertheless exists completely outside the classical or even jazz framework, but is a really really common device in rock and other popular music. Any working musician who plays a lot of repertoire (especially keys and guitarists) will have some ideas in their head of a sort of "practical music theory" that explains the music they encounter, but there's not yet a completely formal framework (as far as I know) that lumps it all together in a coherent and well-defined way. In a way, maybe that's for the best, since modern music is constantly changing and shifting all the time.

  • 1
    If you change "... notes that aren't in the chord" to "... notes that aren't in the accompaniment chord indicated by the chord symbol", then I think it might clarify some of the confusion that the OP has (or had). Namely, the difference between chord symbols written as accompaniment guides, and actual sounding chords produced by playing the written notes together with accompaniment. What exactly the actual sounding chord is, is somewhat subjective and depends on tempo and dynamic emphasis of the played notes, and even expectations coming from the listener's cultural background. Commented May 6, 2019 at 14:38
  • @piiperi That's true in a sense, and maybe I should add that as an additional point to the answer, (i.e. that written chords may be oversimplifications of the actual underlying harmony) but that's actually not what I meant there. I meant that melodies will include passing notes etc. especially in weak beats, that aren't part of the chord e.g. in {2} and {6}. Even for {1}, while a m7 would be a valid voicing in the first chord, I don't think that "1 ♭7 5..." in a melody necessarily implies a 7th in the harmony, it's just a valid bluesy melodic shape (in both major and minor contexts).
    – Some_Guy
    Commented May 7, 2019 at 17:54
  • It's true that sometimes you sing "non-chord tones" on strong beats to essentially "extend the chord" to a more interesting one than the raw accompaniment by adding another note to it (like I did in my Dm9 to Am9 singing example), but mostly that isn't what's happening here, it's just passing notes etc. On the linked recording when the tutti accompaniment comes in the chords are voiced with some 7ths, so I suppose you could write them with m7 and maj7 chords instead of the basic chords if you wanted to, but honestly I prefer not. Let the players decide their own voicings.
    – Some_Guy
    Commented May 7, 2019 at 17:56
  • If it was my chart, I'd write "E♭- D♭ B♭- C♭Δ" and leave it at that. I might write D♭(6) just to avoid people putting a 7th at the top (because I personally hate ugly out of place 7ths which both ma7 and ♭7 would be here, whereas 6/9 etc. would be fine), but really I'd probably just leave it as D♭, neither major nor dominant, and let the player fill it out (or not fill it out) however they want. In the recording it's always played as a plain unadorned major chord, so yeah, fine (unless you count the 9th in that arpeggio, which case ok it's an add9 I suppoooose, but eh, I don't think so)
    – Some_Guy
    Commented May 7, 2019 at 17:58

This passage of music is more modal than tonal.

What this means is that the harmony is directional. Traditionally, tonal music travels from tonic to dominant and back. This section of four chords repeated is harmonically static. For instance, tonal music will often use a major V chord even in minor, but this piece uses the minor v chord, which doesn't have as much harmonic weight.

In modal music, the melody will usually be based more on the scale (aka mode) than on the harmonic movement of the accompaniment. So this passage definitely fits into the standard modal framework; the accompaniment lays down harmonically-static vamp, while the melody floats on top, sticking to the main mode (in this case D# aeolian).

So overall, you're sort of using the wrong tool for the job. You're using the terms of tonal music to describe a piece of music that isn't conforming to the rules of tonal music.

  • I don't feel any need to use modes for explaining what's happening in the song. :) It sounds like a bog-standard pop song of today. Commented May 5, 2019 at 15:31
  • 1
    @piiperi those two aren't mutually exclusive, what Peter means is that it makes more sense to think of the melody in relation to a "home scale" rather than in terms of functional harmony. In that sense it's "modal". In that sense, even a song in the "ionian mode" (aka major scale) can be "modal" if it basically treats that scale as a set of notes to just enjoy playing, regardless of what small amount of harmonic movement there might be. A lot of Irish music is like this for example, and many Irish folk melodies fit equally well over a drone or sparse chord progression.
    – Some_Guy
    Commented May 5, 2019 at 15:58
  • Hello mr Peter, just wanted to say that I enjoyed this answer, in mine I analysed it more functionally as the melody relates to the chords, but the explanation of "music like this often cares more about the home key than what's going on in the chords, especially if those chords are diatonic to the home mode" is also a perfectly good way of looking at it :) The duality of simultaneous "modality" and skeletal "tonality" is also characteristic of the blues too, and I think that you can draw a line from the blues through soul to pop songs like the one linked by OP
    – Some_Guy
    Commented May 5, 2019 at 16:01
  • 1
    Peter's explanation about modal vs tonal is right in principle - and your comment, explaining this as layering a scale-oriented lead melody on accompaniment chords that bring a more functional feeling to the mix, is actually very insightful. But taking the actual song into account, I think all this is overkill. The OP's trying to see e.g. the C# notes in the first bar as "non-chord tones" or "passing tones" feels like he might assume that the accompaniment chord symbols are a complete harmonic description or something. More practice and toying around with music is in order here, not theory. Commented May 5, 2019 at 16:32
  • @piiperi ¿Por qué no los dos?
    – Some_Guy
    Commented May 7, 2019 at 18:33

It would help if you explain a bit about the notation you provided. Where did it come from what exactly is it transcribing from the song? There aren't any lyrics, it it an instrumental part? Etc, etc, etc. Why make us guess at all these points?

I assume this is your transcription of the main melody sans lyrics. If that is the case, one point up front: I don't think the note you labeled #3 - the D4 - is the best notation. That note is a kind of unpitched slide.

Second point: the chord symbols should probably have extension figures like min7, add6, etc. as appropriate. In pop and jazz those chords are very common. This isn't a minor or quibbling detail. It's absolutely essential to understanding the characteristic pop/jazz harmony style and therefore notions of chord tones versus non-chord tones. Depending on how complex the chords are in a pop/jazz song you can get to the point of almost all tones being chord tones. The notation should probably be in Eb minor instead of D# minor and the minor chords should be min7 minor seventh chords.

(Side note, read this article: Temperly, The melodic-harmonic 'divorce' in rock. The style of music covered by the article is different than this song, but closely related, and I think the concept will apply to this song generally. The essential idea is some passages in rock music use melody that doesn't follow chord tones.)

Finally, it's probably best to think of non-chord tone definitions as idealized formulae. Even if a melody doesn't fit a formula, that formula can still be used to describe the music. If something behaves like an escape tone but doesn't resolve according to the formula, it could be reasonable to describe it as an unresolved escape tone. The sense of propriety in music to follow formulae will depend on style. Temperley's point in the article above is rock discards these formulae at times. By contrast classical music is very circumscribed in following them closely.

Now let's review your highlighted notes...

  • #1 chord tone, the seventh of a minor seventh chord
  • #2 escape tone, unresolved
  • #3 to #4 to #5 could just be considered an unresolved appoggiatura both over the C# chord, then a resolving appoggiatura on the A#m chord. This is kind of murky because of how #3 isn't really clearly sung. Regardless, if the note is D# it is a dissonance that resolved by stepping down to C#. This is a good example of melody not following a non-chord tone formula, but definitely displaying a more fundamental concept: resolving a dissonance by step-wise motion to a consonance.
  • #6 the G# is a chord tone of a minor seventh chord and the F# is it's neighbor tone.
  • #7 and #8 if you consider the B chord in mm. 4 and 8 as a B6 chord then in both measures the G is the chord tone and the A is a suspension in m. 4 and an escape tone in m. 8.

Be sure to analyze chord properly before detailing non-chord tones.

Don't overlook appoggiatura, it's a very important non-chord tone. It's similar to a suspension in that they normally resolve by a step down. You might think of it as a unprepared suspension. Technically it isn't, but the resolutions are basically the same.

Label things as unresolved if that seem appropriate.

With these considerations in mind I think you are on the right track with how you are analyzing the melody. I disagree with the comments from others that your are using the wrong analysis tool by labeling non-chord tones. This music is clearly homophonic in which case non-chord tones seems like a perfectly legitimate concept!


This is the progression i-V-iii-IV. It is something similar as the often used phrygian cadence. As we have this progression in the ear you can indeed improvise anything above in the modal scale with no chromatics but diatonic and pentatonic.

My first suspicion was that the notated key is wrong.

But here we have the full accompaniment: Attention - Charlie Puth (Musescore)

I think the chord notation as uncomplete. You will understand the non-chords better if you write the chords as: D#m7-C#9-A#m11-B6,Bmaj7

It won't help a lot to specify the non chord tones. But the wiki-link you've postet with the list of non-chords is very good!

listening to the video we can hear a upper voice in 4ths - there will be even more non chord tones.

  • 1
    I don't think it's really all that necessary to write the chords as extended chords to encompass the melodic notes here, and in many cases it might be overkill. But even if you want to do that, C#9 is not right at all. Maybe C#add9 (if you really want to include that 1 D# passing note as part of the chord, but really, I think plain old C# is actually better here).
    – Some_Guy
    Commented May 7, 2019 at 17:09

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.