There are two common kinds of substitute chords:
i. Tonic substitute
The submediant chord (vi) may function as a PD in I-vi-V-I (as we have seen above) but it may also substitute for T in an interrupted cadence (or deceptive motion). It is able to take the place of the tonic chord because it also contains the tonic note.
ii. Dominant substitute
The leading-note chord (viio), with/without its chordal seventh, and its inversions can substitute for the primary dominant chord. In this case, the crucial common element is the leading-note whose strong voice-leading tendency to move to the tonic enables leading-note chords to function as the dominant. By the same account, iii6 can also substitute for V, hence the variant perfect cadence iii6 - I .
☆ Important theoretical feature: chord identity ≠ chord function
This should begin to be apparent now; the interrupted cadence is a good case in point. We shall see more instances of particular chords being part of different harmonic functional regions, depending on the context, as we move on to our next important theoretical concept.