tl;dr If you change the way you use data structures a little bit you can come up with parallel algorithms to work on them.

This idea is something that Guy Steele has been promoting well before his 2009 ICFP presentation - one of my favorite papers is this one he coauthored with Danny Hillis in 1986: http://cva.stanford.edu/classes/cs99s/papers/hillis-steele-d...