For example: the following is received: "dit dah dit dit" i.e. an (L). For the first dit, select the left column at the top (E), next for the dah, select the right column, the (A) box, on the row below the (E), follow this by selecting the left column (R) below the (A) for the next dit and finally, the left column below the (R) for the last dit giving (L) as the decoded character. It's a bit cumbersome to describe in words but just interpret each letter in the chart as being its own background plus those directly above it read in sequence downward from the top.

The (S) has the (I) and (E) above it, so the (S) is dit-dit-dit as interpreted from the backgrounds of light-light-light read from the top, downward.

The (O) has the (M) and (T) above it, so the (O) is dah-dah-dah as interpreted from the backgrounds of dark-dark-dark, read from the top, downward.

The (C) has the (K), (N) and (T) above it, so the (C) is dah-dit-dah-dit as interpreted from the backgrounds of dark-light-dark-light, read from the top, downward.

The (Q) has the (G), (M) and (T) above it, so the (Q) is dah-dah-dit-dah as interpreted from the backgrounds of dark-dark-light-dark, read from the top, downward.

And so we see the (?) as being light-light-dark-dark-light-light or dit-dit-dah-dah-dit-dit.