Define x∙=x, (λx.M)∙=λx.M∙, and develop an application by contracting its root when that root is syntactically a beta-redex: ((λx.M)N)∙=M∙[x:=N∙]. Otherwise (MN)∙=M∙N∙. Newly created root redexes need not be contracted. Induction with the substitution lemma proves that if M⇒N, then N⇒M∙. Thus any two parallel reducts share this reduct.